UNIVERSITY OF CALGARY. In-home PLC to DSL Interference Characterization and Mitigation. Khaled Ali A THESIS

Size: px
Start display at page:

Download "UNIVERSITY OF CALGARY. In-home PLC to DSL Interference Characterization and Mitigation. Khaled Ali A THESIS"

Transcription

1 UNIVERSITY OF CALGARY In-home PLC to DSL Interference Characterization and Mitigation by Khaled Ali A THESIS SUBMITTED TO THE FACULTY OF GRADUATE STUDIES IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF ELECTRICAL AND COMPUTER ENGINEERING CALGARY, ALBERTA January, 2015 c Khaled Ali 2015

2 Abstract It is often advocated that a solution to the problem of mutual interference between digital subscriber line (DSL) networks and in-home power line communications (PLC) networks is to prevent the PLC networks from utilizing the DSL spectrum. However, this solution will render PLC networks inoperable with the introduction of wide-band DSL technologies like G.fast. As an alternative solution, this thesis proposes to use the common mode (CM) signal, which contains information about the electromagnetic interference (EMI), to estimate and subtract the differential mode (DM) PLC-to-DSL interference from the DM DSL signal. Since the PLC-to-DSL interference environment within a residential setting has neither been characterized via measurements nor by a model, a measurement campaign is conducted. A set of 480 measurements are collected, within two residential test-sites, to characterize the PLC-to-DSL interference environment for two DSL modem installation scenarios. This campaign shows that the PLC-to-DSL interference channels are frequency selective and vary significantly from room to room within the house. Two interference mitigation solutions are proposed in this thesis. The first solution relies on scheduling the PLC channel access; while, the second solution pre-multiplies the PLC symbol by the inverse of the DM cross-coupling channel before transmission. Both solutions utilize adaptive frequency domain interference cancellers (FDICs) that are insensitive to the non-stationarity of the PLC channel and the frequency selectivity of the coupling channels. The performances of the proposed solutions are evaluated, and their effectiveness in mitigating the PLC-to-DSL wide-band EMI is demonstrated using analysis that incorporates the measured PLC-to-DSL coupling channels. i

3 Acknowledgements First and foremost, I would like to sincerely thank my supervisor, Dr. Geoffrey Messier, for his patience, guidance, and support. Dr. Messier taught me how to conduct high calibre research that has practical value. Dr. Messier s commitment to excellence and his ability to advice, guide, and encourage his students to reach their potential is admirable. Working under his supervision made me a better researcher and significantly improved my communication skills. Also, I would like to thank my co-supervisor, Dr. Stephen Lai, for his advice and support. I have been fortunate to have a co-supervisor who cared about my work and gave me valuable feedback. I also thank the Lai and Messier families for welcoming me into their homes to perform channel measurements. I would like to thank my thesis committee, Dr. John Nielsen, Dr. Abu Sesay, Dr. Kyle O Keefe, and Dr. Michael McGuire for their valuable time and advice. Thanks to Lincoln Zhao, Mohamed Gaafar, Michael Wasson and all my colleagues at the FISH Lab for all their help and support. Many thanks to Mohamed Ammar Al Masri for the various stimulating discussions and coffee runs. Thanks to Ms. Ella Lok and the staff at the ECE department. Many thanks to Dr. Rainer Iraschko and the staff at TRTech. Words cannot express my gratitude to my parents, who have been very supportive through out my life. I would have not been able to make it this far, if it were not for their kindness, patience, forgiveness, and advice. In addition, I would like to thank my brothers, Mohamed and Sameh, for their understanding and continuous support. Finally, special thanks to my wife, Noha, and my father-in-law and mother-in law. This work was financially supported by TRTech and the Natural Sciences and Engineering Research Council (NSERC) of Canada. ii

4 Dedication To my loving parents: Mustafa and Magda. iii

5 Table of Contents Abstract i Acknowledgements ii Dedication iii Table of Contents iv List of Tables vii List of Figures viii List of Symbols x 1 INTRODUCTION Broadband Residential Internet Access Technology HFC Broadband Access DSL Broadband Access PLC Broadband Access PLC In-Home Networking DSL and PLC Coexistence Related Work on DSL and PLC Coexistence DM PLC-to-DSL Channel Measurements DSL Electromagnetic Interference Mitigation Solutions Narrow-band EMI Mitigation Solutions Wide-band EMI Mitigation Solutions Thesis Contribution Cross-Coupling Channel Characterization Interference Cancellation Thesis Outline THE DSL and PLC TECHNOLOGIES DSL Technology The Physical Network Fiber to the Exchange Fiber to the Cabinet Fiber to the Premise The DSL Families Basic Rate Interface The High Bit-rate DSL Family The Asymmetric DSL Family The Very High Speed DSL Family Fast Access to Subscriber Terminal DSL Families Comparison DSL Signalling and Frame Structure Discrete Multi-Tone Modulation DSL Signalling DSL Frame Structure The DSL Interference Environment Intrinsic Interference iv

6 Extrinsic Interference PLC Technology PLC In-home Network PLC Signalling and Channel Access PLC Signalling and Modulation MAC for Broadband PLC PLC Noise Environment CROSS-COUPLING CHANNEL MEASUREMENTS DSL Modem Installation Scenarios DSL and PLC Interference Environment Measurement Methodology Measurement Hardware Calibration Measurement Setup Measurement Campaign Test-Sites Case Study A Case Study B Results Cross-Coupling Channel Frequency Responses Stationarity of Cross-Coupling Channels Effect of Spatial Separation INTERFERENCE SYSTEM MODEL Current System Model Effect of Mutual DSL and PLC Interference on Bit Rates Effect of PLC Interference on DSL Bit Rates Effect of DSL Interference on PLC Bit Rates INTERFERENCE MITIGATION SOLUTIONS Modified System Model Proposed System Model Variations in the DM to CM Estimated Ratio Integration of the FDIC Interference Mitigation Block Diagram Scheduling-Based Interference Mitigation Solution Medium Access Cancellation Algorithm C-DSL Training C-PLC Training Performance Evaluation Mean Square Error Analysis Improvement in Bit Rate Pre-Distortion-Based Interference Mitigation Solution PLC Symbol Pre-distortion Cancellation Algorithm C-DSL Training v

7 5.4.3 Performance Analysis Training Phase Transmission Phase Comparison with Spectral Management Solutions CONCLUSION Measurement Campaign Findings Interference Mitigation Solutions Recommendation for Future Research Interference Channel Characterization PLC-to-DSL Cross-Coupling Channels DM and CM DSL Direct Channels Interference Cancellation Bibliography vi

8 List of Tables 2.1 Comparison of DSL Families Average coherence bandwidth in khz Latency: Scheduling-based solution Training phase available bit rates Transmission phase available bit rates Latency: Scheduling-based versus pre-distortion-based solutions vii

9 List of Figures and Illustrations 2.1 Conventional DSL network Sinc functions at various sub-carrier indexes DMT transmitter block diagram DMT receiver block diagram DM and CM signal at the receiver DSL super-frame structure Near end crosstalk among twisted pairs Far end crosstalk among twisted pairs Electrical wiring in north America [1] In-home PLC network Simplified PLC coupling circuit PLC frame structure for a 3-user PLC network Desk Modem Scenario Entry Point Scenario Co-located DSL and PLC networks interference environment North Hills 0320BF Balun Northern Microdesign PLC Coupler Insertion loss of the balun and the PLC coupler before calibration Modified Balun Modified PLC Coupler Right to Left: Open, Short, and Load (100 Ω) Through calibration Setup for PLC-to-DSL coupling GPIB controlled measurement setup Case Study A: test-site floor plan Case Study B: test-site floor plan Coupling in Differential Mode Coupling in Common Mode Common mode to differential mode transfer function (C2DTF) Variation in DM cross-coupling channels from one power outlet to another Entry Point Scenario coupling gain probability density function Desk Modem Scenario coupling gain probability density function Mean DM and CM PLC-to-DSL coupling Desk Modem Scenario.: DM PLC-to-DSL coupling for Room A Plug 1 over measurement interval Entry Point Scenario.: DM PLC-to-DSL coupling for Room A Plug 1 over measurement interval Desk Modem Scenario: Room A Plug 1 DM vs. CM PLC-to-DSL coupling Entry Point Scenario: Room A Plug 1 DM vs. CM PLC-to-DSL coupling. 71 viii

10 3.26 Effect of relative distance between DSL and PLC modems on PLC-to- DSL coupling Effect of relative distance between DSL and PLC modems on crosscoupling channels coherence bandwidth Current system model Received DSL Signal versus PLC interference US and DS frequencies for band plan 998E30 [2] Available DSL bit rates Degradation in DSL bit rates Received PLC Signal versus DS DSL interference Received PLC Signal versus US DSL interference Available PLC bit rates Degradation in PLC bit rates Proposed system model Interference coupling on twisted-pair Integration of the FDIC into a DMT transceiver Interference cancelling scheme block diagram Desk Modem Scenario: MSE of the proposed scheme versus the relative distance between the DSL and PLC modems Entry Point Scenario: MSE of the proposed scheme versus the relative distance between the DSL and PLC modems Desk Modem Scenario: Achieved improvement in bit rates versus the Euclidean distance between the DSL and PLC modems Entry Point Scenario: Achieved improvement in bit rates versus the Euclidean distance between the DSL and PLC modems Achieved total DSL bit rates for DS transmission, for various DSL cable run length Desk Modem Scenario: Achieved bit rates for DSL-DS Entry Point Scenario: Achieved bit rates for DSL-DS Desk Modem Scenario: Achieved bit rates for PLC Entry Point Scenario: Achieved bit rates for PLC Desk Modem Scenario: Achieved bit rates for DSL-US Entry Point Scenario: Achieved bit rates for DSL-US ix

11 List of Symbols and Abbreviations η c,dsl DSL CM AWGN matrix η c,plc PLC CM AWGN matrix η d,dsl DSL DM AWGN matrix η d,plc DSL DM AWGN matrix γ 0 SNR in presence of AWGN only γ F SINR after utilizing C-DSL γ W SINR after utilizing Weiner filter γ W/O SINR before interference mitigation φ MSE of FDIC φ F MSE of C-DSL φ Tr MSE of C-DSL during training period of the pre-distortion solution φ Tx MSE of C-DSL during transmission period of the pre-distortion solution φ W MSE of Weiner filter f Sub-carrier spacing f DSL DSL sub-carrier spacing f PLC PLC sub-carrier spacing Γ SINR gap x

12 γ(i) λ SINR for frequency bin i Normalization factor v d,dsl Estimated of DM PLC interference matrix v d,plc Estimated of DM DSL interference matrix C DSL DSL FDIC coefficient matrix C W Wiener filter coefficient matrix C PLC PLC FDIC coefficient matrix H DSL Direct DSL channel frequency response matrix H PLC Direct PLC channel frequency response matrix H f DSL c,dtx PRx CM interference channel frequency response matrix between the DSL receiver and a PLC transmitter, sampled at integer multiples of f DSL H f DSL c,ptx DRx CM interference channel frequency response matrix between a PLC transmitter and the DSL receiver, sampled at integer multiples of f DSL H f DSL d,dtx PRx DM interference channel frequency response matrix between the DSL receiver and a PLC transmitter, sampled at integer multiples of f DSL H f DSL d,ptx DRx DM interference channel frequency response matrix between a PLC transmitter and the DSL receiver, sampled at integer multiples of f DSL H f PLC d,ptx DRx DM interference channel frequency response matrix between a PLC transmitter and the DSL receiver, sampled at integer multiples of f PLC xi

13 q Transmitted PLC signal matrix r c,dsl Received CM DSL signal matrix r c,plc Received CM PLC signal matrix r d,dsl Received DM DSL signal matrix r d,plc Received DM PLC signal matrix u Generated PLC symbol v c,dsl CM DSL interference matrix v c,plc CM PLC interference matrix v d,dsl DM DSL interference matrix v d,plc DM PLC interference matrix x y z Transmitted DSL signal matrix Desired DSL signal matrix Desired PLC signal matrix (...) Average of a variable ε D DSL maximum transmission power ε P PLC maximum transmission power ŷ ẑ (...) Estimate of desired DSL signal matrix Estimate of desired PLC signal matrix Estimate of a variable xii

14 b(i) Number of bits for frequency bin i b max Maximum number of bits allowed per frequency bin M N Number of DSL sub-carriers Number of PLC sub-carriers N 0,c PSD of CM AWGN N 0,d PSD of DM AWGN N o PSD of AWGN R Total number of bits R 0 Total bit rates in presence of AWGN only R Bound Total number of bits in presence of background noise R F Total bit rates after utilizing C-DSL R Interference Total number of bits in presence of interference plus background noise R W/O Total bit rates before interference cancellation R W Total bit rates after utilizing Weiner filter t i Start time of Slot i T P PLC frame duration T SF DSL super-frame duration T S DSL frame duration (...) Complex conjugate of a variable xiii

15 E[...] Loss % a H(i, j) x(i) A/D ADSL Expectation of a variable Bit rates loss percentage Normalization factor i th row and j th column of the matrix H i th row of the vector x Digital to analog Asymmetric DSL ADSL2 Asymmetric DSL 2 ADSL2+ AM AMR AWGN BNC BRI C-DSL C-PLC C2DTF CB CM Asymmetric DSL 2 plus Amplitude modulation Automatic meter reading Additive white Gaussian noise Baby N Connector Basic rate interface DSL FDIC PLC FDIC CM to DM transfer function Coheren Bandwidth Common mode xiv

16 D/A DFT DM DMT DOCSIS DS DSL EMI FDD FDIC FFT FT FTTB FTTC FTTCab FTTDp FTTEx FTTH FTTP Digital to analog Discrete Fourier Transform Differential mode Discrete multi-tone Data over cable service Interface specification Downstream Digital subscriber line Electromagnetic interference Frequency division duplexing Frequency domain interference canceller Fast Fourier transform Fourier transform Fiber to the building Fiber to the curb Fiber to the cabinet Fiber to the distribution point Fiber to the exchange Fiber to the home Fiber to the premise xv

17 G.fast GPIB HDSL HFC HFTP IDFT IFFT ISP ITU-T Fast access to subscriber terminals General purpose interface bus High bit-rate DSL Hybrid fibre-coaxial cable Hybrid fiber twisted pair Inverse discrete Fourier Transform Inverse fast Fourier Transform Internet service provider International telecommunication union-telecommunication standardization sector LAN LPF OFDM ON PAM PDF PLC PLC-DC PSD PSTN Local area network Low pass filter Orthogonal frequency division multiplexing Optical node Pulse amplitude modulation Probability density function Power line communications PLC domain controller Power spectral density Public service telephone network xvi

18 QAM QAM RJ SHDSL SMA TB TDD US VDSL Quadrature amplitude modulated Quadrature amplitude modulation Registered jack Single-pair high-speed DSL Sub-miniature version A Terminal block Time division duplexing Upstream Very high speed DSL VDSL2 Very high speed DSL 2 VNA xtalk Vector network analyzer Crosstalk xvii

19 Chapter 1 INTRODUCTION Recent advances in power line communications (PLC) have made it popular for inhome networking. This makes PLC an increasingly relevant source of interference for digital subscriber line (DSL) networks within the home environment. This thesis presents two measurement case studies that characterize the PLC-to-DSL coupling channels, within a residential setting. In addition, this thesis proposes two interference mitigation solutions that enhance the coexistence of in-home PLC and DSL networks. The rest of this chapter is organized as follows. Broadband access technology for residential Internet is introduced in Section 1.1, while home networking solutions via PLC technology is discussed in Section 1.2. The coexistence environment between in-home PLC networks and DSL broadband access networks are discussed in Section 1.3, while background information on topics relevant to the PLC and DSL interference characterization and mitigation are discussed in Section 1.4. The thesis contributions are highlighted in Section 1.5, while the outline of the thesis is presented in Section Broadband Residential Internet Access Technology Broadband access technology is an integral part of the world wide communication network. The term broadband refers to any access technology that is always-on and can support multiple services at high data rates. Residential broadband access has become a pillar of the world culture. Throughout the developed world, and in some developing countries, access to the Internet is considered one of the household essentials due to the various services that can be provided through it, along with the 1

20 reduction in cost of having it. The ultimate in broadband access technology, in terms of data rates and reliability, is optical fiber. However, due to the cost of installation and replacement of existing access technologies, creating a world-wide communications network via optical fiber is not feasible. Currently, hybrid optical fiber connections, where the link to any particular subscriber runs partly on optical fiber cables and then partly on copperbased wires, are utilized to form hybrid networks that serve as the backbone of the world wide communications network. Residential broadband access technologies can be classified into two categories: wired (or fixed line) and wireless broadband access technologies. Fixed line broadband access technologies utilize a physical connection between the service provider and the customer; thus, a physical wired network has to be in place before the nodes in the network can communicate. Wireless broadband access, on the other hand, utilize air as the transmission medium. Wireless broadband solutions provide, along with freedom of mobility, instantaneous wide area coverage that is especially beneficial for remote areas where the infrastructure for fixed line technologies do not reach. However, wired broadband access technologies have several advantages over the wireless technologies such as higher data rates and better reliability. The most commonly used copper-based wired broadband access technologies are digital subscriber lines (DSL) and hybrid fiber-coaxial cable (HFC). DSL utilizes the copper twisted-pairs of the public service telephone network (PSTN) as the transmission medium, while HFC utilizes the coaxial cables used by the digital cable television (CATV) network as the transmission medium. Both DSL and cable modems technologies incorporate hybrid fibre optic connections, where portions of the twisted-pairs or the coaxial cables within the networks are replaced with fiber optic cables. Another copper-based fixed line broadband access technology is power line communications 2

21 (PLC). PLC utilizes the power lines that are used to carry electricity to subscriber houses as a transmission medium. The aforementioned copper-based fixed line broadband access technologies are discussed in Sections to HFC Broadband Access The cable television (CATV) network was created for broadband unidirectional transmission. Various optical nodes (ONs) are connected to the head unit via fiber cables, creating HFC connections. Out of each ON, branches of shared coaxial copper cable connect various customers. Thus a tree is formed, where the root of the tree is the head-end [3]. Note that up to 500 customers can be serviced by a single ON. To enable broadband access to the Internet, the CATV network is upgraded to support bidirectional traffic. HFC relies on dividing the bandwidth of the shared coaxial cable into non overlapping channels, where one or more of these channels are dedicated for upstream transmission while the rest of the channels are reserved for downstream transmission and cable TV. The most recent version of the cable modems protocol Data Over Cable Service Interface Specification (DOCSIS), which is the most widely used protocol for HFC broadband access, provides data rates of up to 400 Mbits/s [3]. However, the major drawback of HFC broadband access is that the cable TV network is a shared tree network, and the bandwidth per customer is limited. The available bandwidth per customer is dependent on the number of active users and poses a significant challenge for upstream data rates when a large number of users are active. In addition, due to the structure of the tree, security and scalability are always a challenge DSL Broadband Access DSL technology utilizes dedicated link for each customer, where data is transmitted over the public service telephone network (PSTN). A PSTN utilizes twisted-pairs 3

22 of copper wires that were originally designed to serve as a medium for transmitting speech signals. Since human speech is in the range of 300 Hz to 3400 Hz, higher frequencies can be used for transmitting data over existing PSTNs [4]. The ubiquity of the PSTN has motivated new DSL technologies that increase throughput using a combination of advanced communications techniques and replacing the twisted-pairs in portions of the PSTN by fibre optic cables [5]. These hybrid connections shorten the distance the DSL signal has to travel over passive copper twisted-pairs between the central office and the end users, and results in a wider bandwidth channel. An advanced standardized DSL technology is the very high bit rate digital subscriber line 2 (VDSL2). VDSL2 utilizes up to 30 MHz bandwidth [2] and vectoring [6] to achieve transmission rates up to or exceeding 100 Mbits/s. Currently, a standard is being developed for a new DSL technology, called G.fast (fast access to subscriber terminal) [7], that has the potential of achieving data rates of 1 Gbits/s, over a spectrum that spans 106 or 212 MHz [8]. These fast data rates being dedicated to each customer, unlike HFC, have made DSL a popular choice among end users for residential broadband access. The DSL technology is discussed in more detail in Section PLC Broadband Access Extensive research on utilizing PLC networks to deliver Internet to customer houses has been performed. However, one of the main obstacles that prevented utilizing power lines by Internet service providers (ISPs) is the need for a repeater at each transformer, since data signal cannot pass through the transformers [3]. Installing repeaters at each transformer, especially in North America where each transformer serves only few houses, was proved costly. This high cost and the presence of other telecommunication infrastructure prevented the concept of providing Internet over power lines to materialize. 4

23 1.2 PLC In-Home Networking Utilizing legacy wires, such as power lines, telephone cables, and coaxial cables, to distribute data within the home environment has no rewiring requirement. Various standards have been developed for the three aforementioned legacy wires to be used for residential data distribution. Among these standards, the international telecommunication union-telecommunication standardization sector (ITU-T) home networking standard G.hn [9] has specifications for each of the three legacy wires networks, along with multi-domain specifications. For the multi-domain specifications, two or more of the legacy wires networks can be utilized simultaneously to deliver data. Among the three legacy wires, power lines are the most extensively used in houses. This coupled with the recent advances in the area of networking within the home using legacy wires make PLC networks increasingly common within the home. The power lines can be utilized to form a network that transform the house into a smart home, where appliances within the house are connected. Creating an in-home local area network (LAN) via PLC technology has many advantages over other in-home networking solutions such as Ethernet and wireless. The first advantage is the low implementation cost which is due to the existence of a pre-installed power line network in each house. Thus, no new wiring or physical installation is required, which is not the case with Ethernet LANs. A second advantage that in-home PLC LANs has over Ethernet LANs is the presence of multiple access nodes (in form of electrical outlets), which are spread throughout the house. Currently, the average cost of a PLC network card is the same as the cost of a wireless network card; however, as in-home networking via PLC gains popularity with end users, mass production of PLC networks cards will cost about half the cost of the wireless network card because PLC network cards do not require an RF component [10]. Additionally, data rates over PLC networks can reach up to 200 Mbits/s, with 5

24 the ability of integrating multiple functions over the same network. For instance, automatic meter reading (AMR), home automation, and triple play services (Internet, television, and telephone services) can be integrated over an in-home PLC network simultaneously [11]. All the aforementioned advantages make in-home PLC networks a suitable choice for in-home networking [3] and, in cases where one medium is not sufficient, is a cost-effective complement to other in-home technologies [12]. Further detail on the PLC technology is presented in Section DSL and PLC Coexistence In-home PLC networks operate over the same spectrum as DSL networks. This increases the likelihood of crosstalk between PLC and DSL communications systems. For instance, two home networks that operate at the same frequency range, one over copper twisted-pairs (138 khz - 30 MHz [2]) and the other over power lines (1.8 MHz - 30 MHz [9]), would interfere with each other. The DSL and PLC interference environment is discussed in more detail in Section 3.2. Communication standards such as ITU G.hn [9], have been developed with mechanisms that prevent any interference between various systems within the home environment. However, the interference from PLC-to-DSL is usually prevented by forcing the PLC network to notch frequencies that affect the DSL signal. While this is a viable solution, denying PLC networks access to the DSL spectrum will render in-home PLC networks inoperable especially if VDSL2 or G.fast technology is employed. Further detail on mitigating the PLC-to-DSL interference via spectral notching is discussed in Section Other PLC-to-DSL interference reduction solutions, such as reducing the PLC transmit power (which is known as spectral management) or increasing the distance 6

25 between the DSL and PLC modems (which is referred to as spatial separation), have been proposed in the literature. As will be discussed in Section , spectral management degrades the performance of the PLC network, and with the increase in the usable DSL bandwidth, spectral management solutions will render the PLC networks inoperable. From the measurement case studies presented in this thesis and in [13, 14], it is shown that the interference levels between the PLC and DSL modems do not depend on the distance separating the modems. Thus, spatial separation, which is discussed in more detail in Section , does not mitigate the PLC-to- DSL interference. The goal of this thesis is to enhance the co-existence environment between DSL and PLC networks in a residential environment. The objective is to mitigate the PLCto-DSL interference without hindering the performance of the in-home PLC network. Ultimately, this will lead to the increase of data transmission rates in both DSL and PLC networks. Complementary signals are inserted onto each of the wires of the twisted-pair by the transmitter. At the receiver, the difference between the complementary signals is the differential mode (DM) signal, while the common mode (CM) signal is the arithmetic mean of the complementary signals. The DM signal contains the desired signal, while the CM signal is mainly composed of the interference. Further detail on the DM and CM signalling in DSL systems is discussed in Section The main hypothesis of this thesis is that the common mode signal contains information about the differential mode interference. By utilizing the common mode signal, the differential mode interference can be estimated and subtracted from the differential mode signal. Removing the differential mode interference increases transmission rates in DSL networks without limiting the capacity of the in-home PLC network. The common mode and differential mode signalling are discussed in more 7

26 detail in Section 2.1.3, while estimating the DM interference from the CM signal is discussed in Chapter 5. A thorough literature survey indicates that the PLC-to-DSL cross-coupling channel measurements in a residential setting that included both DM and CM reception methods have never been performed [13]. Only DM PLC-to-DSL channel measurements in lab settings have been performed in the literature; further detail on work done to date with regards to PLC-to-DSL cross-coupling channel measurements can be found in Section Characterizing the PLC-to-DSL cross-coupling channels is central to the main hypothesis and the interference mitigation solutions proposed in this thesis. Thus, a measurement campaign that studies the cross-coupling channels between DSL and in-home PLC networks has to be performed. The focus of this thesis is divided into two areas. The first focus area is PLC-to- DSL cross-coupling channel characterization in a residential setting. A measurement campaign that studies the PLC-to-DSL interference environment within residential test sites is performed. The second focus area is interference cancellation. Two interference mitigation solutions based on adaptive filter theory are proposed. The effectiveness of the proposed solutions in mitigating wide-band EMI from an in-home PLC network on to the DSL system, based on field measurements, is evaluated. 1.4 Related Work on DSL and PLC Coexistence In this section, work related to the two focus areas of the thesis is discussed. Related work to the PLC-to-DSL coupling channel measurements is presented in Section 1.4.1, while work related to the PLC-to-DSL interference mitigation is given in Section

27 1.4.1 DM PLC-to-DSL Channel Measurements Measurement case studies that investigate the differential mode PLC-to-DSL coupling channel in a laboratory environment were performed in [15, 16, 17, 18]. These studies investigated the effects of various factors, such as the distance separating the PLC and DSL cables and the shared length between the PLC and DSL cables, on the PLCto-DSL coupling over the VDSL2 spectrum. The main finding of these studies is the potential for significant crosstalk between PLC and DSL systems. However, these studies do not characterize the PLC-to-DSL coupling channel in an actual residence, nor do they consider the CM PLC-to-DSL coupling channel. The coupling between PLC and DSL systems over the frequency range of 138 khz to MHz (band plan 998ADE17 ) was measured in [19]. The objective of this study was to investigate the quality of service for Internet protocol television (IPTV). Two case studies, [20] and [21], investigated the effect of PLC on the throughput of a VDSL modem. The DM PLC-to-DSL coupling channels were neither measured in [20] nor in [21]. Rather the effect of the PLC system on the VDSL modem was simulated by inducing a PLC signal into a co-axial cable with variable attenuators. The main finding of [20] and [21] is that the VDSL throughput experiences degradation when the ratio of the received VDSL signal power to the PLC signal power is less than 20 db. In other words, as long as the DSL cable run length is less than 300 m, the VDSL throughput will not be hindered. Note that in [19] the DM PLC-to-DSL coupling channels up to only MHz was measured, while in [20] and [21] no channel measurements were performed. In addition, the CM PLC-to-DSL coupling channel was not studied in [19, 20, 21]. It is important to note that the studies discussed above did not measure the PLC-to-DSL coupling channels, within a residential setting. Knowledge about the coupling channels between in-home PLC and DSL networks is essential in determining 9

28 an interference mitigation solution. Since there is neither a model nor previously conducted field measurements that characterize the interference environment between in-home PLC networks and a DSL modem, further field measurements of the DM and CM PLC-to-DSL coupling channels are required DSL Electromagnetic Interference Mitigation Solutions The copper twisted-pairs, utilized as the transmission medium by the DSL networks, were not initially designed to transmit broadband signals over a wide spectrum. As with all copper based wires, the twisted-pairs turn into an antenna at high frequencies [22]. This antenna radiates a portion of the transmitted signal as electromagnetic waves which cause interference to other systems operating at same frequencies. This radiated electromagnetic interference (EMI) is known as EMI egress. The copper twisted-pairs also pick up the radiated electromagnetic waves from other systems, and this causes what is known as EMI ingress. EMI to DSL can be classified as narrow-band EMI or wide-band EMI interference. Narrow-band EMI, as the name suggests, typically affects few of the DSL sub-channels. The most common narrow-band EMI to DSL systems are from amplitude modulation (AM) radio and amateur radio. Wide-band EMI, however, causes interference to a large portion of the DSL spectrum, which is usually from co-located systems that operate over the same spectrum as DSL. The most common wide-band EMI to DSL systems is from a co-located in-home PLC network [12, 23]. Techniques used to mitigate narrow-band and wide-band EMI to DSL systems are discussed in Sections and respectively Narrow-band EMI Mitigation Solutions Narrow-band EMI mitigation can be performed at the analog front of the DSL modem or within the digital structure. No optimal solution exists, but rather the choice 10

29 depends on each situation [22]. Digital narrow-band EMI mitigation solutions are usually preferred to analog solutions, since digital solutions are less expensive and more flexible than analog ones. However, in certain scenarios an analog solution would be necessary. For instance, if the EMI levels are high enough to saturate the analog to digital converter, EMI mitigation has to be performed at the analog front of the DSL modem. Note that a mixture of analog and digital narrow-band EMI mitigation solutions are utilized in most scenarios [22]. Various techniques are utilized in the literature to mitigate the narrow-band EMI effect on DSL signal. These techniques are either active or passive. Note that both analog and digital narrow-band EMI mitigation solutions can utilize either active or passive techniques. Passive techniques rely on information about the frequency of the interference and utilize filters (usually, notch filters) to eliminate the narrow-band EMI egress. Active techniques, on the other hand, rely on finding a reference signal or on a priori knowledge of the EMI to mitigate the differential mode narrow-band EMI egress. Usually, a reference signal is obtained via satellite or antenna in wireless systems; however, for DSL this is not required since the common mode signal can be utilized. Studies that take advantage of the common mode signal to mitigate the effect of narrow-band EMI were performed in [24, 25]. In [24], the authors mitigate the effect of impulsive noise on the DSL signal by utilizing the CM signal. Similarly, in [25], the CM signal was used to reduce the impact of EMI on the DM DSL signal Wide-band EMI Mitigation Solutions Wide-band EMI mitigation solutions fall into one of the following four categories: spectral notching, spectral management, interference cancellation, and spatial separation. In Section to , the work done to date in each of the four aforementioned wide-band EMI mitigation solution categories is discussed. 11

30 Spectral Notching Home networking standards, such as ITU G.hn [9], ensure that home networks that operate over the same spectrum as any of the DSL technologies do not interfere with the functionality of the DSL system. This is achieved by preventing home networks from utilizing frequencies occupied by the DSL system. Even though preventing home networks from operating over the DSL spectrum ensures successful operation of the DSL system, it denies home networks from benefiting from a wide range of frequencies. In addition, with the emerging of new DSL technologies that utilize wider bandwidth, such as G.fast (up to 212 MHz), denying PLC networks access to the spectrum that overlaps with DSL networks will render the PLC network inoperable [26] Spectral Management As an alternative to spectral notching, spectral management is proposed in [27] and [28] to alleviate the effects of the PLC interference by reducing the transmit power for PLC sub-carriers that cause interference to the received DSL signal. In [27], the authors assume a flat PLC-to-DSL cross-coupling channel and study the achieved DSL bit rates at various PLC transmit power levels. In [28], the authors propose reducing the transmit power for PLC sub-carriers that interfere with the downstream frequencies of VDSL2. As will be shown in Chapter 5, the proposed solutions achieve higher DSL bit rates than both solutions proposed in [27] and [28]. In addition, it will be shown that, mitigating the effects of PLC interference through spectral management negatively affects the PLC bit rates, while the proposed solution does not hinder the PLC bit rates Cross-Coupling Channel Equalization An interference cancellation solution was proposed in [28], which entails utilizing adaptive filters to estimate and mitigate the effects of the DM PLC interference on the DM DSL signal. In [28], the authors propose connecting the PLC network and 12

31 the DSL network, by utilizing a coupler that carries the PLC signal transmitted on the power lines to an adaptive filter that is connected to the DSL line. A fundamental assumption in [28] is that the coupling channel between the PLC transmitter and DSL receiver is the same as between the DSL transmitter and PLC receiver. As shown by the measurement campaign presented in Chapter 3, the PLCto-DSL coupling channel varies significantly from one outlet to the other. And the coupling channel experienced at the PLC receiver outlet will be very different than the coupling channel at the PLC transmitter outlet. In addition, the authors in [28] assume that the retraining of the adaptive filter is only required to accommodate the time-variation of the PLC channel, and they do not account for the outlet-to-outlet variation experienced by different users. Another wide-band EMI mitigation solution that utilizes an adaptive filter to equalize the cross-coupling channel was proposed in [29]. In this study, a time domain adaptive filter utilizes the CM signal to estimate the DM EMI over the asymmetric DSL (ADSL) spectrum ( MHz [30]). Because of the nature of the wide-band interference, this filter requires a long training time [29]. Note that this time domain filter requires a long training time for a DSL network operating over the ADSL spectrum, which is much smaller than the spectrum utilized by VDSL, VDSL2, or G.fast technologies. Thus, utilizing a time domain adaptive filter is not feasible for DSL technologies that operate over a spectrum that span tens of MHz. In addition, a core assumption made in [29] is that the relationship between the differential and common mode cross-coupling channels is a smooth function of frequency. As will be shown in Chapter 3, the ratio of the PLC-to-DSL DM to CM cross-coupling channels varies significantly from one sub-channel to the other Spatial Separation The authors in [31] propose that changing the location of the PLC transmitter or 13

32 the DSL receiver might be utilized to mitigate the interference between PLC and DSL; since the coupling levels vary from one power outlet to the other, the locations with the lowest coupling levels should be chosen. However, as will be shown in Chapter 3, the spatial separation between the DSL and the PLC modems does not have a significant impact on the coupling channels captured in the measurements presented in this paper. The average coupling levels are relatively the same for all outlets. 1.5 Thesis Contribution The goal of this thesis is to mitigate the mutual wide-band interference between co-located DSL and PLC networks, within the home environment. To achieve this goal, the focus of the thesis is on two areas: cross-coupling channel characterization and interference cancellation. The thesis contribution to the area of cross-coupling channel characterization is summarized in Section 1.5.1, while the thesis contribution to the area of interference cancellation is presented in Section Cross-Coupling Channel Characterization Cross-coupling channel characterization is essential to understand the various interference sources affecting the DSL networks. In this thesis, the cross-coupling channels between DSL and in-home PLC networks are studied. Residential field measurements of PLC-to-DSL cross-coupling channels have never been performed. A measurement campaign, which is discussed in detail in Chapter 3, has been designed and conducted to characterize the PLC-to-DSL cross-coupling channels. Both the DM and CM PLC-to-DSL cross-coupling channels are studied within two residential houses. For each of the two houses, the DM and CM PLC-to- DSL cross-coupling channels are measured for various rooms. A set of 480 DM and CM cross-coupling channel measurements between the DSL 14

33 and PLC networks has been collected for the two DSL modem installation scenarios discussed in Chapter 2. The main findings of the measurement campaign are: A strong DM PLC-to-DSL cross-coupling channel exists, with cross-coupling channel coupling gains ranging between 40 db and 80 db. The PLC-to-DSL cross-coupling channels are frequency selective and independent of the proximity between the PLC modem and the DSL modem. The PLC-to-DSL cross-coupling channels vary significantly from one outlet to the other. The CM to DM PLC-to-DSL interference ratio is not a smooth function of frequency, and varies from one outlet to the other Interference Cancellation The interference cancellation portion of this thesis is concerned with how to mitigate the mutual interference between a DSL modem and a co-located in-home PLC network, without negatively impacting the performance of the PLC network. Two PLC-to-DSL interference mitigation solutions are presented in Chapter 5. Both solutions utilize adaptive filters to estimate the differential mode interference on a tone by tone basis using the common mode signal. Only adaptive time domain cancellers that attempt to mitigate the wide-band EMI suffered by the DM DSL signal have been researched [29]. Due to the frequency selective nature of the PLC interference, these time domain adaptive filters cannot mitigate DM PLC interference. The first interference mitigation solution uses an adaptive frequency domain canceller that utilizes the CM DSL signal to extract information about the DM PLC interference and relies on restricting channel access to one PLC user per DSL super-frame. This scheduling is performed to combat the variation in cross-coupling channels from 15

34 one power outlet to the other. While this will increase the latency experienced by the PLC users, it will be shown that this additional delay is acceptable for multi-media applications. The second interference mitigation solution eliminates the need to restrict channel access to one PLC user per DSL super-frame by pre-multiplying the PLC symbol with the inverse of the cross-coupling channel. This eliminates the latency caused by the first solution but will be more expensive to implement due to the additional signal processing required. Similar to the first solution, the second solution also utilizes an adaptive frequency domain canceller that estimates the DM PLC-to-DSL interference from the CM DSL signal. As will be shown in Chapter 5, both proposed solutions successfully mitigate the PLC-interference on the DSL network, without degrading the performance of the PLC network. This is one of the main advantages that the proposed interference mitigation solutions have over the interference mitigation solutions discussed in Section Thesis Outline This thesis is organized as follows. In Chapter 2, a brief background on the DSL and PLC technologies is provided, along with description of the PLC-to-DSL interference environment within a residential setting. In Chapter 3, a measurement campaign that characterizes the PLC-to-DSL crosscoupling channels within two test sites is presented. Both the differential and the common mode cross-coupling channels were measured in two test sites. In Chapter 4, the current system model is presented. Additionally, the effects of the mutual wide-band interference on the achieved bit rates for both DSL and PLC systems are discussed. In Chapter 5, two interference cancelling solutions are presented. While both 16

35 solutions utilize adaptive filters to extract an estimate of the DM interference from the usually ignored CM signal, each solution has its advantages and disadvantages. The first solution while having the advantage of simplicity, adds restriction on the PLC channel access per DSL super-frame. This restriction introduces latency, albeit within the acceptable range for multimedia services. The second solution removes the restriction on PLC channel access per DSL super-frame; thus, eliminating the added latency introduced by the first solution. However, the second solution encompasses pre-shaping the PLC symbols before transmission. This adds to the complexity of the PLC transmitter, which is translated in the implementation cost. Finally, in Chapter 6, the thesis is concluded where the main findings of the thesis are summarized and future research direction is introduced. 17

36 Chapter 2 THE DSL and PLC TECHNOLOGIES The objective of this thesis is to enhance the coexistence between DSL and PLC networks in residential settings. DSL, despite its maturity, is still an active area of research. Recent advancement in DSL research lead to the development of standards that promise to provide data rates of up to 1 Gbits/s, which makes DSL the leading technology in copper-based wired broadband access. Similarly, PLC, because of its existence within every structure, has generated a lot of research advancement which has made it a popular choice for in-home networks. DSL and PLC utilize overlapping spectra; as a result, interference between both systems occurs. In order to mitigate the interference between co-located PLC and DSL networks, a better understanding of both networks is required. In this chapter, the DSL and PLC technologies are briefly introduced in Sections 2.1 and 2.2 respectively. 2.1 DSL Technology In this section, we introduce DSL technology. The layout of the DSL physical network is presented in Section The various DSL families are discussed in Section 2.1.2, and the DSL modulation, frame structure, and signal transmission techniques are presented in Section Finally, the DSL interference environment is presented in Section

37 From Core Network Central Office (Local Exchange) FTTEx FTTCab FTTB FTTH Street Cabinet (Pedestal) Street Cabinet (Pedestal) Building (remote DSLAM) USER USER USER Figure 2.1: Conventional DSL network. Twisted-pairs Fiber The Physical Network In a DSL network, as shown in Fig. 2.1, cables run from the core network to central offices (COs) (or local exchanges or access nodes (ANs)). Historically, the cables between the core network and the COs were twisted-pairs. In each of these cables, there is about 1000 twisted-pairs. From each CO, up to 100 twisted-pairs are connected to pedestals that service a number of houses. Finally, each house is connected to its respective pedestal by two twisted-pairs. However, only one of these twisted-pairs is used in conventional DSL networks. Thus, a house is connected by one passive copper twisted-pair to the CO. As a result, the DSL signal travels a very long distance which considerably reduces the transmission bit rate. Presently, however, sections of the twisted-pairs are being replaced by fiber optic cables. This connection is referred to as a hybrid fiber twisted pair (HFTP) connec- 19

38 tion [5]. These hybrid connections shorten the distance the DSL signal has to travel over passive copper twisted-pairs. HFTP connections have many deployment scenarios, depending on the connection from the CO to the end user. The main HFTP deployment scenarios are discussed in Sections to [32, 33] Fiber to the Exchange In the fiber to the exchange (FTTEx) deployment, copper twisted-pairs connect the CO to cabinets (or pedestals) that service a number of houses. Each house is connected to its respective cabinet by two twisted-pairs. However, only one of these twisted-pairs is usually active. Thus, in an FTTEx deployment, a house is connected by one passive copper twisted-pair to the CO. As a result, the DSL signal travels a very long distance which considerably reduces the transmission rate Fiber to the Cabinet In fiber to the cabinet (FTTCab) deployment, fiber cable is utilized to connect the CO to a remote optical unit (ONU) within the cabinet. Thus, DSL transmission over copper twisted pairs occurs only between the ONU and the customer premise. The FTTCab deployment has many variants, such as fiber to the curb (FTTC) or fiber to the distribution point (FTTDp). The main difference among these variants is the reach of the fiber connection within the DSL network Fiber to the Premise Fiber to the premise (FTTP) deployment has two variants. First is fiber to the building (FTTB), where a fiber cable connects the CO with a multi-dwelling building. Once in the building, a remote DSL access multiplexer (DSLAM) manages the DSL transmission over very short runs of twisted-pairs. The second variant is fiber to the home (FTTH), where the CO, and consequently the core network, is connected to the user by a fiber optic cable. FTTH is considered ideal and future proof [5], since a need 20

39 for data rates higher than the rates achieved by this deployment cannot be envisioned. However, installation of FTTH is both time consuming and very expensive, which is why it is not widely deployed The DSL Families DSL technologies have a very broad range, and as a result, they can be grouped by modulation scheme, data rate transmission direction, or by any other common factor between the technologies. In this section, the DSL technologies are classified into families, where each family share similar modulation schemes and data rate transmission directions. In Sections to , the major DSL families are introduced, while in Section the spectrum, data rate, and maximum reach of the various DSL technologies are compared Basic Rate Interface The basic rate interface (BRI), which is considered as the original DSL, is based on the integrated service digital network (ISDN) [34] technology. BRI provides the ability to transmit both data and voice signals over the PSTN, with data rates up to 160 kbits/s over a single twisted-pair. The maximum length of a single run of the twistedpair cannot exceed 18 kft; however, with the use of repeaters, operation over longer distances can be achieved. BRI is offered as a replacement for plain old telephone service (POTS). However, where the demand for POTS exists, BRI and POTS can coexist. Based on the geographical region of deployment, various modulation schemes for BRI are used. In North America, a four level pulse amplitude (4-PAM) modulation is used [35]. BRI provides a symmetric transmission rate, i.e., the entire bandwidth is utilized for both upstream (US) and downstream (DS) transmission. This is achieved by the use of echo cancellation. 21

40 The High Bit-rate DSL Family The high bit-rate DSL (HDSL) family of DSL technology is the most mature DSL technology. The HDSL family has witnessed a number of technologies, all of which have not been standardized. The common feature among theses variants is the ability to deliver symmetric high data rates over long distances, which is suitable for business customers. The two most common technologies in this family are the high bit rate DSL (HDSL) and the single-pair high-speed DSL (SHDSL), which are standardized by the international telecommunication union-telecommunication standardization sector (ITU-T) in G [36] and G [37] respectively. HDSL offers symmetric data rates of Mbits/s over two twisted-pairs. Each of the twisted-pairs carry a 784 kbits/s, over a single run of 12 kft. To increase the reach of a single run, more twisted-pairs can be used, in addition to utilizing repeaters. Similar to BRI, HDSL utilizes a 4-PAM modulation with echo cancellation. HDSL utilizes low frequencies to increase its reach; thus, HDSL cannot coexist with either POTS or BRI. SHDSL, on the other hand, provides data rates that range from 192 kbits/s to 2.3 Mbits/s, in increments of 8 kbits/s, over a single twisted pair. However, there is an option to utilize a second pair to improve the reach. SHDSL provides a multi-rate transmission, and relies on coding gain to achieve its high data rates. The modulation utilized is a 16-PAM modulation with trellis coding (TC) and echo cancellation. Various regional operational conditions are specified in G.991.2, with an annex dedicated to achieving data rates up to Mbits/s. Similar to HDSL, SHDSL is not compatible with either POTS or BRI The Asymmetric DSL Family Residential customers, unlike business customers, require higher data rates in the DS transmission direction. As a result, the asymmetric DSL (ADSL) technology emerged, 22

41 which is compatible with both POTS and BRI and deliver high data rates over a long reach. The three main technologies in the ADSL family are the asymmetric DSL (ADSL), the asymmetric DSL 2 (ADSL2), and the asymmetric DSL 2 plus (ADSL2+), which are standardized by the ITU-T in G [30], G [38], and G [39], respectively. ADSL offers asymmetric data rates up to 8 Mbits/s in the DS direction (from the CO to the customer) and up to 896 kbits/s in the US direction. The maximum reach of an ADSL line is 18 kft; however, at this reach the available DS data rate is Mbits/s. ADSL was the first of the DSL technologies to utilize DMT modulation. The frequency spectrum of ADSL starts from 25 khz up to 1.1 MHz; however, as an option, frequencies up to 80 khz can be avoided in presence of BRI. Frequency division duplexing (FDD) of US and DS transmission is achieved in ADSL by utilizing the frequencies from 25 khz to 138 khz for US transmission, while the frequencies from 138 khz to 1.1 MHz are reserved for DS transmission. ADSL2 is an extension of ADSL. The same frequency spectrum is utilized in both; however, ADSL2 offers asymmetric rates of up to 12 Mbits/s (DS transmission) and 1 Mbits/s (US transmission). The enhancement in the data rates is achieved by utilizing trellis coding, which was optional in ADSL, and the ability to use one-bit constellations, such as binary PAM [4]. In addition, an all-digital mode is present in ADSL2, where the entire bandwidth can be utilized, in absence of POTS and BRI. ADSL2+ is an enhanced version of ADSL2, where the DS bandwidth is extended to 2.2 MHz. As a result, the achieved DS data rates are increased to 20 Mbits/s, especially for customers close to the CO. For various regional operational conditions, the DS transmit PSD can be shaped to meet specific requirements. 23

42 The Very High Speed DSL Family The main feature of the very high speed DSL (VDSL) family is that it provides both symmetric and asymmetric transmission, which is suitable for both business and residential customers respectively. While the ADSL family successfully offer high bit rate over long distances, the demand for higher rates kept on rising. The VDSL family of DSL technologies offers the highest data rates achieved over copper twistedpairs. However, these high data rates are only available over short distances. The two DSL technologies within the VDSL family are the very high speed DSL (VDSL) and the very high speed DSL 2 (VDSL2), which are standardized by the ITU-T in G [40] and G [2] respectively. VDSL offers asymmetric data rates that can reach 52 Mbits/s for DS transmission and 1.5 Mbits/s for US transmission, for customers within a 1 kft radius from the ONU. Symmetric data rates of 10 Mbits/s can be achieved over distances of 4.5 kft. DMT modulation, along with FDD of US and DS transmission is used in VDSL. VDSL uses frequency ranges from MHz to 12 MHz, with these frequencies divided into two DS and two US band plans. In addition to being compatible with both POTS and BRI, VDSL is compatible with the entire ADSL family. VDSL2 offers data rates that can reach up to 200 Mbits/s (asymmetric) and 100 Mbits/s (symmetric), over short distances. VDSL2 uses a wider bandwidth (from 25 khz up to 30 MHz) and has a longer reach than VDSL. DMT modulation, along with FDD of US and DS transmission is used in VDSL2. G standard specifies four major band plans, with eight profiles, where each of these band plans is suitable for a specific HFTP deployment [41]. Band plan 8 (4 profiles: 8a, 8b, 8c, and 8d) utilizes a bandwidth of 8.6 MHz with data rates up to 50 Mbits/s. Band plan 12 (2 profiles: 12a and 12b) utilize a bandwidth of 12 MHz and has a maximum data rate of 68 Mbits/s. Both band plan 8 and 12 are suitable for FTTEx deployments, 24

43 with an 8 kft maximum length of twisted-pair run. Band plan 17 (1 profile: 17a) utilizes a bandwidth of 17.7 MHz and provides a maximum data rate of 100 Mbits/s. This band plan is suitable FTTCab deployments, where the maximum length of the twisted-pairs is approximately 5 kft. Finally, the Band plan 30 (1 profile: 30a) utilizes the entire VDSL2 bandwidth and offer data rates up to 200 Mbits/s. This band plan is suitable for HFTP deployments where the twisted-pair length does not exceed 1 kft, such as FTTB deployments. Similar to VDSL, VDSL2 is fully compatible with POTS, BRI and the entire ADSL family Fast Access to Subscriber Terminal Fast access to subscriber terminal (G.fast) technology is the most advanced among the DSL technologies. Currently, G.fast is being standardized by ITU-T in G.9700 [7]. G.fast, which has the potential of delivering aggregate data rates of up to 1 Gbits/s, is suitable for FTTP deployment, where the maximum length of the twisted-pairs is approximately 0.82 kft (250 m) [8]. This high data rate is achieved by utilizing a wider bandwidth. G.fast is being studied for two band plans: from 2.2 MHz to 106 MHz and to 212 MHz. Instead of FDD, G.fast is expected to utilize time division duplexing (TDD) of US and DS transmission. It is expected that the US to DS ratio will be flexible, but constant among lines served by the same distribution point. The US to DS ratio has to be constant to mitigate the NEXT effect. Finally, it is expected that G.fast will be compatible with all the aforementioned DSL technologies DSL Families Comparison Table 2.1 compares the utilized bandwidth, the maximum data rates, the maximum reach over twisted-pairs, and the ITU-T standards for the DSL technologies mentioned above. In this thesis, the focus is on profile 30a of the VDSL2 technology. Thus, any 25

44 Table 2.1: Comparison of DSL Families. Family Technology Spectrum Max. Data Rates Reach ITU-T MHz Mbits/s kft Standard BRA ISDN up to G.961 HDSL up to G HDSL on two pairs SHDSL up to G on a single pairs ADSL DS: G US: 0.64 ADSL ADSL DS: G US: 1 ADSL DS: G US: 1 VDSL (net) 1.0 G VDSL 10 (net) 4.5 VDSL (net) 1.0 G (net) 5.0 G.fast G.fast (212) 1000 (net) 0.82 G.9700 referral to the DSL spectrum, indicates frequency ranges up to 30 MHz. Note that the solutions to interference mitigation, introduced in this thesis, are not frequency dependent. As will be shown in Chapter 5, the interference mitigation filters proposed in this proposal can be utilized by any DSL technology that is based on DMT modulation DSL Signalling and Frame Structure Currently, DSL systems utilize discrete multi-tone (DMT) modulation. DMT modulation, which utilizes a discrete Fourier transform (DFT) based block transceiver, is a variant of orthogonal frequency division multiplexing (OFDM). One of the main differences between DMT and OFDM is that the output of the DFT block in DMT is real valued samples, while the output of the DFT block samples in OFDM are complex. The rest of this section is organized as follows. DMT modulation and DFT-based 26

45 transceivers are discussed in Section DSL signalling and frame structure are discussed in Section , while the DSL frame structure is discussed in Section Discrete Multi-Tone Modulation In DMT, the channel bandwidth is partitioned, utilizing the inverse discrete Fourier transform (IDFT), into orthogonal sub-channels (or tones). This division simplifies the channel equalization needed to negate the channel dispersive effect. Moreover, DFT-based transceivers are favoured in multi-channel modulation due to the availability of computationally efficient implementation methods for the DFT, such as the fast Fourier transform (FFT). Fig. 2.2 shows three sub-channels (sub-channels 10, 11, and 12), where the subchannels remain orthogonal at the sub-carrier frequency. Note that, at the sub-carrier indexes, i.e., the center frequencies of the sub-channels, the contribution from other sub-channels is zero. Thus, from the receiver s point of view, sampling the received signal at the sub-carrier frequencies is equivalent to having multiple parallel nonoverlapping sub-channels. The number of sub-channels depends on the frequency separation between subcarriers (i.e., the sub-carrier spacing f) which varies from one system to another. The main factor that affects the number of sub-carriers is the coherence time of the channel. The coherence time is defined as the duration during which there is no variation in the channel impulse response. The number of sub-carriers is usually chosen such that the number of sub-carriers is maximized and the symbol duration is less than the coherence time. Variation in the channel during the DMT symbol duration degrades the orthogonality of the sub-channels. Since the DMT symbol consists of multiple tones, it inherits the robustness of a tone, i.e., a sinusoid signal, to dispersion. This fact is exploited to simplify the channel 27

46 1 0.8 Carrier no. 10 Carrier no. 11 Carrier no Sub carrier index Figure 2.2: Sinc functions at various sub-carrier indexes. equalization for DMT-based systems. The insensitivity of a tone to the dispersion of the channel can be easily seen in the frequency domain, where the Fourier transform (FT) of an infinite duration tone is a delta function. No matter how dispersive the channel is, distortion of the tone by the channel is limited to amplitude and phase changes, which can be negated by a single-tap equalizer. For a signal that is composed of multiple tones at various frequencies, the effect of the dispersive channel will be also limited to changes in the amplitudes and the phases of each of the delta functions corresponding to the various sinusoids. The same concept can also be applied to discrete-time tones, where the DFT of a discrete-time tone is also a delta function in the discrete-frequency domain. However, during the truncated DMT symbol, the received signal is composed of the summation of different replicas each of which is delayed differently due to the dispersive channel. Thus, the received multiple tones are no longer orthogonal during the DMT symbol time. This, along with causing inter-symbol interference (ISI), complicates the equalization of the channel effect. To overcome ISI and simplify the channel equalization process, two things are 28

47 performed. First, a guard interval is utilized between consecutive symbols such that the ISI occurs during this guard interval. Additionally, few samples are taken from the end of the DMT symbol and appended to the beginning of said symbol. By doing so, the received multiple tones are guaranteed to have suffered the same dispersion, and thus, the DFT of these multiple tones are delta functions and the effect of the dispersive channel will be limited to changes in the amplitudes and the phases of each of the delta functions. This is achieved, as will be explained below, via adding a cyclic prefix to the input sequence before applying it to the channel. A block diagram of the DMT transmitter is shown in Fig The bit stream is converted from a serial stream of bits into N parallel sets of bits via a serial to parallel (S/P) converter. The number of parallel sets corresponds to the number of sub-channels, and the number of bits in each of the parallel sets is based on each subchannel signal to noise ratio (SNR). Each set of bits is modulated via a quadrature amplitude modulation (QAM) encoder, such that the output of the QAM encoder is the complex N 1 vector X define by X = [X 0, X 1,..., X N 1 ] T, (2.1) where X i is the modulated QAM symbol for sub-channel i. X 1 x 1 Bit Stream S/P QAM Encoder X 2 x 2 X N 2N-point complex-to- real IFFT P/S x 1, x 2,..., x 2N Add Cyclic Prefix D/A and Filter x 2N Figure 2.3: DMT transmitter block diagram. The IDFT of the output of the QAM encoder is obtained via a 2N-point complex 29

48 to real IFFT operation, which results in the 2N 1 real vector x defined by x = [x 0, x 1,..., x 2N 1 ] T. (2.2) A real output is required because, in DMT, the output of the IFFT is applied directly to the channel after digital to analog conversion. To obtain the real vector x from the complex vector X via the IFFT operation, a 2N 1 vector X H with Hermitian symmetry property has to be formed from the N 1 complex vector X. This can be easily performed by X H = [ Re{X N 1 }X 0, X 1,..., X N 3, X N 2, Im{X N 1 }, X N 2, X N 3,..., X 2, X 1, ] T, (2.3) where Re{X i } and Im{X i } indicate the real and the imaginary parts of X i respectively, and X i is the complex conjugate of X i. The elements of x, i.e., the IDFT of X H are calculated by x n = 1 2N 1 X H,k e j(2π/2n)kn, n [0, 2N 1]. (2.4) 2N k=0 The output of the IFFT block, i.e., the time domain input samples defined by vector x, is passed through a parallel to serial (P/S) converter, which converts the parallel sub-symbols to a series of sub-symbols as shown in Fig Before the inserting the DMT symbol into the channel, the cyclic prefix is added to the beginning of the DMT symbol. The cyclic prefix is a copy of the last v 1 sub-symbols of x 2N, which is appended to the beginning of time domain input sequence as shown in Fig The length of v is chosen such that it equals the length of the channel impulse response. In that manner, the ISI between consecutive symbols will be confined to the cyclic prefix. Note that instead of appending the last v 1 sub-symbols to the DMT symbol before transmitting it over the channel, one could append v 1 zeros, to confine the ISI to the cyclic prefix. However, the cyclic prefix serves another purpose. 30

49 As mentioned earlier, adding the cyclic prefix serves the purpose of simplifying the channel equalization. By appending the last v 1 samples of the time sequence representing the DMT symbol to the beginning of the sequence, the aperiodic input time sequence seems periodic over the length of the convolution. Thus, the linear convolution can be represented by a circular convolution, once the cyclic prefix is removed. Modelling the linear convolution as a circular convolution is important because it allows the usage of DFT. This means that, after removing the cyclic prefix, the DFT of the received sequence equals the product of the DFTs of the transmitted sequence and the channel impulse response. Consequently, the channel effect can be negated using a simple one-tap equalizer for each sub-channel. Finally after adding the cycling prefix to the input sequence, the input sequence is converted to an analog signal via digital to analog (D/A) block as shown in Fig Before the analog signal is applied to the channel, a transmit filter, shown in Fig. 2.4, may be utilized to eliminate any out of band power leakage and to ensure that the transmitted signal power spectral density (PSD) remains in a specific range. Note that the PSD levels can be controlled digitally, but sometimes it is easier to control it via the analog front end. [42] Fig.2.4 shows the DMT receiver block diagram. Once the signal is received, it passes through the receiver filter which minimizes the out of band noise, after which it is converted from analog to digital via the analog to digital (A/D) block. The cyclic prefix is then stripped from the received sequence y. Through a P/S converter the received sequence is converted to 2N sub-symbols. These sub-symbols are fed to an FFT block which performs and 2N-point real-toimaginary conversion. The FFT block takes the DFT of the received sequence and reverses the Hermitian symmetry. Recall, the cyclic prefix is added to the input sequence to force periodicity over the convolution interval, which results in a linear 31

50 y 1 Y 1 ^ X 1 y 2 Y 2 FEQ ^ X 2 QAM Decoder P/S Bit Stream Filter and A/D Add Cyclic Prefix y 1, y 2,..., y 2N S/P N-point complex-to- real IFFT Y N ^ X N y 2N Figure 2.4: DMT receiver block diagram. convolution being equivalent to a circular convolution. Thus, for a noiseless channel, the received sequence after the FFT block is a multiplication of the input sequence and the channel frequency response, which can be written in vector form as Y 0 X 0.H 0 Y 1 X 1.H 1 Y = =, (2.5).. Y N 1 X N 1.H N 1 where H i is the frequency response of the i th sub-carrier defined in (2.6). H k = 1 N 1 h n e j(2π/2n)kn, k [0, N 1]. (2.6) N k=0 To extract the X i from Y i, the received sequence is passed through a single-tap filter per sub-channel called frequency-domain equalizer (FEQ) as shown in Fig The coefficients of the FEQ is the inverse of the channel frequency response. Note that division is not usually preferred from an implementation perspective. Thus, the single-tap FEQ is implemented via a complex operation that involves scaling and rotating Y i to mitigate the effect of channel impairment [22]. In presence of noise, the outputs of the FEQ are not exactly equal to X but rather ˆX, which is an estimate of X. This estimate of X is then passed through a QAM encoder followed by a P/S converter, as shown in Fig. 2.4, to reconstruct the 32

51 transmitted bit stream. Note that in absence of coding, ˆX is passed through a simple symbol by symbol decision process to recover the constellation points before utilizing the QAM decoder. However, if trellis encoding (i.e., convolution codes) is utilized, then a trellis decoder is required to decode the constellation points DSL Signalling DSL signals are transmitted in differential mode (DM) where complementary signals are transmitted over a twisted-pair of wires. DM signalling provides resilience to electromagnetic interference (EMI) as any external interference will couple identically on both of the pairs, and thus, any interference will be eliminated at the receiver. However, in practical situations, external interference affects each wire in a pair differently. The common mode (CM) signal, which is the arithmetic mean of the signals, can be determined at the receiver, with little extra cost. The CM signal contains information about the external EMI; thus, the CM signal can be utilized to estimate the DM EMI. That estimate can then be used to cancel the interference. Balun DM Signal CM Signal Figure 2.5: DM and CM signal at the receiver. Fig. 2.5 shows a balun that is connected to a twisted-pair, and outputs two signals: the DM and the CM signal. A balun is a resistance transformer that converts the balanced DSL signal to a differential mode signal, at the receiver end. From the center tap of the balun, the common mode signal, which is usually ignored since it does not contain relevant information about the desired signal, is obtained. Ideally, a balun will output a DM signal that is only made up of the desired signal and a 33

52 CM signal that only contains noise and electromagnetic interference. However, due to imperfections in the balun, and in the twisted-pairs, the DM portion of the EMI is found in the DM signal DSL Frame Structure The DSL network utilizes a dedicated channel for each user. Data is transmitted over the DSL channel in super-frames. Since each DSL user has its own dedicated channel, all super-frames on the DSL channel belong to a single user. The super-frame for VDSL2 is divided into 257 frames. The last frame within a super-frame is the synchronization (sync) frame [2]. Sync symbols are usually utilized for only synchronization purposes, where bits are modulated via 4-QAM constellation; however, pilot signals can be transmitted during the sync frame for channel state estimation purposes [6]. To facilitate parallel sub-channel estimation, orthogonal pilot sequences are again used [43]. Note that the pilot sequences are vendor specific [44]. DSL Super-frame t Synch T S T SF Data T D Figure 2.6: DSL super-frame structure. A super-frame, shown in Fig. 2.6, has duration of T SF. Note that the value for T SF varies from one standard to the other; for VDSL2 T SF =64.25 ms [2], while for G.fast T SF < 10 ms [8]. Recall, the DSL technology utilized in this thesis is VDSL2. For simplicity, and without loss of generality, let us assume that the first frame in each 34

53 super-frame is a synchronization frame and there are 256 frames per super-frame. Thus, as shown in Fig. 2.6 a DSL super-frame constitutes a synchronization frame followed by 255 data frames, each with a duration of T S =0.25 ms. Thus, in this thesis, the duration of VDSL2 super-frame is assumes to be 64 ms The DSL Interference Environment Interference in DSL can be classified as intrinsic, such as thermal noise and crosstalk, or extrinsic, such as impulsive noise and PLC interference [45]. In this section, the various sources of interference in DSL networks and the interference mitigating techniques utilized to combat them are discussed. The rest of this section is organized as follows. In Section , DSL intrinsic interference sources are discussed, while DSL extrinsic interference sources are presented in Section Intrinsic Interference Crosstalk (xtalk), which occurs when a signal power leaks from one twisted-pair to another, is a wide-band interference and is the main cause of errors in a DSL network [22]. There are two types of crosstalk: near end crosstalk (NEXT) and far end crosstalk (FEXT) [46]. Tx/Rx Signal NEXT Tx/Rx Figure 2.7: Near end crosstalk among twisted pairs. NEXT occurs among twisted-pairs when an interfering signal is transmitted from the same end of the cable as the receiver, as shown in Figure 2.7. FDD is utilized to 35

54 prevent NEXT among DSL lines served by the same distribution point [22]. However, since TDD is used instead of FDD in G.fast, the US to DS ratio should be constant among twisted-pairs served by the same distribution point. By fixing the US to DS ratio, it is ensured that all transceivers at the customers premises (or at the CO) are transmitting at the same time, which eliminates NEXT. Tx/Rx Signal FEXT Tx/Rx Figure 2.8: Far end crosstalk among twisted pairs. FEXT, on the other hand, occurs when an interfering signal is transmitted from the end of the cable that is opposite to the receiver as shown in Figure 2.8. Various mitigation techniques are utilized to mitigate the FEXT, the most prominent of which is vectoring [6]. Vectoring is discussed in more detail Section The concept of interference mitigation is essential to any communication systems. Various techniques over the years have been developed to estimate and cancel DSL intrinsic interference. In Sections to , interference mitigation techniques that are utilized in mitigating DSL intrinsic interference are highlighted Adaptive Filters Mitigating the effects of NEXT by utilizing time domain adaptive filters was proposed in [47] [48]. For a cable containing m+1 twisted pairs, m filters are required to mitigate the NEXT effect on each twisted-pair. Thus, a total of (m+1)m adaptive filters are needed to eliminate NEXT in said cable. In addition, access to the disturbing signals is required for this approach to work. The computational complexity of this 36

55 approach makes it impractical, especially at the customers premises. A variation of this approach was proposed in [49], where the highest n NEXT sources are cancelled using n adaptive filter. However, the same limitations on this variation still exist, along with the task of detecting the highest n disturbers Frequency Division Duplexing Frequency division duplexing (FDD) of US and DS transmission in DSL networks is used to mitigate the effects of NEXT [22]. By utilizing FDD, transceivers at the same end of the cable are transmitting and receiving at different frequency bands. This creates a zipper pattern where the available spectrum is divided into smaller sub-spectra. The sub-spectra are allocated in an alternating manner to US and DS transmission; thus, eliminating NEXT. However, as a result of utilizing FDD, the available VDSL2 bandwidth is restricted to specific non-overlapping frequency ranges, which affects the overall transmission rate. In addition, in presence of multiple DSL systems that employ different duplexing schemes, utilizing FDD is rendered ineffective in mitigating the effects of NEXT [50] Vectored Transmission For a single DSL system, the utilization of FDD results in the elimination of NEXT. Thus, the major source of intrinsic interference for DSL is FEXT. In 2001 a new transmission technique called vectored transmission (vectoring) was proposed in [51], which can be utilized to eliminate FEXT from a DSL network for all transceivers within the same cable binder co-located at the DSLAM. As long as all the twisted-pairs in the cable are connected to the same CO and are managed by the same service provider [52], FEXT among twisted-pairs can be virtually eliminated, which results in transmission rates up to 100 Mbits/s [44]. However, if not all the twisted-pairs within the cable binder are considered in the joint signal processing (i.e, presence of uncontrolled lines), the resulting FEXT will negatively 37

56 impact the vectoring process [53]. In 2010, vectoring was standardized in [6], and is being adopted by a number of service providers as the next-generation broadband technology [54]. While vectoring shows promising improvement in data rates, especially over short lines, the presence of uncontrolled lines significantly deteriorates this improvement. This is a severe limitation of vectoring, especially when local loop unbundling (LLU). In LLU the incumbent local exchange carriers are forced to share their infrastructure with other operates. Since multiple operators share a cable binder, and since channel state information is not shared among the multiple operators, mitigating the effect of FEXT via vectoring will not be possible. In [55], it was shown that data rates dropped from 100 Mbits/s to 70 Mbits/s in presence of a single uncontrolled line. Another limitation of vectoring is its complexity in situations where a relatively large number of twisted-pairs served by the CO [56] Extrinsic Interference Extrinsic interference can be classified as narrow-band or wide-band interference. Radio frequency interference from AM and amateur radio, which usually affects few tones at a time, is considered narrow-band interference. These types of narrow-band extrinsic interference usually are geographically variable [45]. On the other hand, wide-band extrinsic interference is usually from systems co-located with the DSL system, and utilizes the same spectrum as DSL. One of the most prominent of these technologies is power line communications (PLC) [12, 23]. Techniques that mitigate the effect of EMI on DSL, both narrow-band and wide-band EMI, have been discussed in Section Note that this thesis focuses on wide-band extrinsic interference from PLC on DSL in a residential environment. Recall, the objective of the thesis is to propose solutions that mitigate the interference between the DSL modem and a co-located 38

57 PLC network, within a residential setting. 2.2 PLC Technology Although PLC as a technology has been proposed since the beginning of the last century [57], it is only recently that PLC networks are seen as a viable option for home networking. Communications over power lines utilized a narrow-band single carrier at its early stage [58]; currently, broadband PLC utilize a wide spectrum that reaches up to 100 MHz and has the potential of delivering data rates of up to 500 Mbit/s [3]. As discussed in Chapter 1, delivering the Internet to customers over the power lines faced many challenges due to the cost and availability of cheaper alternatives. However, PLC technology saw success in various areas such as vehicular networks, smart grid applications, municipal applications, and local area networks (LANs). Because of the existing extensive power line infrastructure and the continuous advancement in LAN technology, PLC is becoming an attractive solution to create in-home networks among devices that might benefit from access to the Internet. In this section, broadband PLC technology and its application in home networking are introduced. The architecture of a typical in-home PLC network is presented in Section 2.2.1, while broadband PLC signalling and frame structure is discussed in Section PLC In-home Network Electric power is carried to neighbourhoods over high voltage power lines in the range of 6 to 16 kv. For residential usage, this high voltage is reduced via step-down distribution transformers to 240 V [1]. Fig. 2.9, shows a distribution transformer with two leads. The two leads of the transformer are connected to Line 1 and Line 2. 39

58 The potential between Line 1 and Line 2 is 240 V [58]. In north America, the 240 V is split into two phases via a central tap that is connected to the neutral of the house [59]. The potential between the neutral and each of Line 1 and Line 2 is 120 V, as shown in Fig Thus, within the panel, there are two lines, the neutral, and the earth ground. Note that the earth ground connection is not shown in Fig Distribution Transformer Panel Line 1 16 kv Neutral 120 V 240 V Line V Figure 2.9: Electrical wiring in north America [1]. The power lines within the house are terminated at bus bars with the panel (also known as the circuit breaker panel). These power lines are either single-phase or two-phase. The single-phase power lines are connected to outlets within the house to supply light fixtures and small appliances such as TV, PLC couplers, etc, with electricity. Large appliances such electric stoves, water heaters, etc, are connected to the two-phase since they require the usage of the 240 V between Line 1 and Line 2. In-home PLC networks are utilized for many purposes. An in-home PLC LAN can have multiple functions, and support communications on various spectra. There are two categories of PLC networks: narrow-band and broadband. Narrow-band PLC (NB-PLC) is used for AMR and home automation; broadband PLC (BB-PLC) is mainly used for distribution of triple play services within the house. Fig shows a PLC network within the home environment. The unshielded power lines run throughout the house. PLC transceivers are connected via the mains outlets to form 40

59 the PLC network. To these transceivers, devices such as computers and television, are connected. Power Lines Broadband Router PLC Transceiver PLC-DC PLC Transceiver PLC Transceiver Figure 2.10: In-home PLC network. Note that cross-phase coupling, i.e., the ability of the PLC signal to couple from one phase to the other, is an issue for NB-PLC. The low operating frequency used by NB-PLC prevented the signal from coupling across phases because the impedance between phases caused the signal to attenuate significantly. A capacitive coupler is required to enhance the coupling between phases [59]. However, since BB-PLC utilize high frequencies, the bus bars in the panel act as a capacitor which enables the cross-phase coupling. Additionally, BB-PLC signal couples across phases at large appliances, which are connected to both phases. Thus, cross-phase coupling is not an issue for BB-PLC [1]. 41

60 As will be discussed in Section 2.2.2, the shared PLC channel access by PLC transceivers (users) is performed in one of two ways, depending on the type of traffic. Contention based channel access does not require a controlling entity to organize the channel access among users. Contention free channel access, as with other broadband access technologies with shared mediums, requires a domain controller to facilitate access among PLC users based on each user s requirement. Note that this is the case when a minimum quality of service is required. In contention free channel access, one of the PLC transceivers takes over as the PLC domain controller (PLC-DC). The PLC-DC regulates access to the PLC channel and is connected to the Internet via a broadband router, as shown in From Fig In addition to accessing the Internet via a computer, services such as IPTV or voice over Internet protocol (VOIP) are provided via the PLC transceivers PLC Signalling and Channel Access In this section, the aspects of the physical and MAC layer of in-home PLC networks relevant to this thesis are discussed. In Section , PLC signal transmission and modulation are discussed, while the MAC standards for broadband PLC networks are introduced in Section PLC Signalling and Modulation Power Lines PLC Coupler DM Signal CM Signal Figure 2.11: Simplified PLC coupling circuit. 42

61 Similar to DSL, differential mode signalling is utilized in PLC. Fig shows how a capacitive PLC coupling circuit used to transmit broadband signals over inhome power lines [60]. The objective of this coupling circuit is to filter out the 60 Hz high voltage waveform and transmit a broadband signal over the power lines. The PLC modems also have fuses before the capacitors and diodes after the transformer for over voltage protection. The fuses and diodes are not shown in Fig What is relevant to this thesis is the functionality of the transformer. Similar to the balun, the coupler acts as an impedance transformer, where two balanced signals are converted to two unbalanced signals. The DM signal, which is the difference between the two balanced complementary signals, is utilized to carry the desired Signal. The CM signal, is usually ignored, since it mainly contains EMI. PLC utilizes DFT-based transceivers for signal modulation/demodulation [58], which allows the partitioning of the broadband channel into narrower sub-channels. Among the variants of DFT-based modulation techniques, DMT modulation is the most suitable for PLC communications. This is because base-band transmission is utilized in PLC networks. DMT modulation was discussed in detail in Section Note that the PLC sub-carrier spacing f PLC is larger than the DSL sub-carrier spacing, since the PLC channel state varies more rapidly than the DSL channel. Recall, the number of sub-carriers (or sub-channels) is inversely proportion with the coherence time of the channel. Since the DMT symbol duration has to be shorter than the coherence time of the channel to prevent degrading the orthogonality of the sub-channels, and since the coherence time of a PLC channel is shorter than the coherence time of a DSL channel, the DMT symbol duration for a PLC network is shorter than that of the DSL network. And due to the inverse relationship between time and frequency, a shorter PLC symbol duration means a wider PLC sub-carrier spacing. 43

62 MAC for Broadband PLC As discussed in Section , the DSL network utilizes a dedicated channel for each user. The PLC channel, on the other hand, is shared among all the users in the PLC network. As will be discussed in Chapter 3, the cross-coupling channel varies significantly from one PLC user to the other (i.e., from one power outlet to the other). Thus, to propose solutions that overcome the outlet-to-outlet variations in the PLC-to-DSL cross-coupling channels, it is essential to understand the channel access mechanism utilized by in-home broadband PLC networks, for multimedia applications. Various standard bodies have developed standards for the MAC layer of a broadband PLC network. Among the most wide-spread standards are HomePlugAv [61], IEEE P1901 [62], and G.hn [9], which are standardized by the HomePlug Consortium, the IEEE, and the ITU-T standard bodies, respectively. In all the aforementioned three standards, PLC nodes are divided into two classes, PLC domain controllers (PLC-DCs) and PLC users. In [62], the PLC-DC is called the local administrator, while in [61] and [9], the domain controller is referred to as the connection manager and the domain manager, respectively. PLC users are referred to slave stations or PLC nodes. PLC-DCs have various functions such as transmitting beacons to provide info on the contention periods and contention free periods, broadcasting information about channel, and assigning time-slots to PLC users. Usually, the first node that joins the network takes over as a PLC-DC [9]; however, as will be discussed in Chapter 5, the PLC user closest to the DSL modem is the PLC-DC in this thesis. All the above standards employ one of two channel access mechanisms, either carrier sense multiple access with collision avoidance (CSMA/CA) or time division multiple access TDMA [63],[64]. CSMA/CA is reserved for best effort traffic, where channel access is contention-based. Best effort traffic, such as web-based or 44

63 applications, is a classification of traffic that does not have a quality of service (QoS) requirement, such as packet loss, latency, etc. Sensitive traffic, such as VOIP and video conferencing, is traffic that requires its packets to be delivered on time with a certain error threshold. For this type of sensitive traffic, TDMA is used as an access mechanism where the PLC-DC assigns time-slots to the various PLC users according to their QoS requirements. In this thesis, it is assumed that the in-home PLC LAN is utilized for QoS traffic; thus, the PLC network employs TDMA as a channel access mechanism. The TDMA channel access for a three-user PLC network (User A, User B, and User C) is depicted in Fig The vertical axis indicates the names of the users and the combined channel, while the horizontal axis indicates the time, which is divided into time-slots. The user s packets only represent when traffic becomes ready for transmission at each node. Thus, Fig shows how the time-slots are allocated to the various PLC users. PLC t 1 t 2 t 3 t 4 t 5 t 6 t 7 t 8 User A A A A A A User B B B B B B User C C C C C Channel Slot 1 Slot 2 Slot 3 Slot 4 Slot 5 Slot 6 Slot 7 - B A C A C B T P Figure 2.12: PLC frame structure for a 3-user PLC network. t Assuming that all three PLC users have the same QoS requirement, the PLC-DC allocates the next available time-slot to the PLC user with the earliest frame arrival 45

64 time, (i.e., the PLC user whose frame was queued for transmission before all other PLC frames in the network). In Fig. 2.12, the first frame arrival is between t 1 and t 2 from User B followed by User A then User C, where t i is the start time of Slot i. Thus, the PLC-DC assigns the next available time slot, Slot 2, to User B; after which, the PLC-DC allocates Slot 3 and Slot 4 for User A and User C, respectively. Note that Slot 1 is not used because the first frame arrival occurs after the beginning of Slot 1. Note that it is assumed that the PLC transceivers follow the recommendations of the ITU-T G.hn home networking standard. While the interference mitigation solutions discussed in Chapter 5 propose some modification to MAC layer of the PLC network, these modifications comply with the recommendations of the G.hn standard. 2.3 PLC Noise Environment PLC utilizes unshielded power-lines which are more sensitive to EMI than the twistedpairs utilized by DSL. The power lines within the home pick up EMI from various noise sources such as AM and amateur radio, appliances within the house, etc. In the literature, there are three main classes of noise present on the PLC channel. However, in this thesis, a fourth class is introduced. The first class of noise is the coloured background noise, which is the sum of various noise sources and is considered time-invariant since the noise level does not change for various consecutive AC cycles [58]. Typically, the coloured background noise has a PSD of -145 dbm Hz [65]. The second class of noise is the impulsive noise which is characterized with shorter durations and higher amplitudes than the background noise. Impulsive noise within PLC system is either synchronous or asynchronous to the mains frequency, or in form of isolated impulses. Synchronous impulsive noise occurs at the frequency of the AC 46

65 mains and is usually due to silicon controlled rectifiers and from electronic circuits. Asynchronous impulsive noise occurs at frequencies much higher than the AC mains, and a typical source of asynchronous impulsive noise is switching regulators. In the literature, the PSD of impulsive noise is in the range of -105 dbm db [65]. Hz The third class of noise to PLC systems is the narrow-band interference from AM and amateur. As was the case with narrow-band DSL interference, this type of interference affects few sub-channels at a time and is usually mitigated via notch filters. Finally, the fourth class of noise to PLC systems is wide-band interference from co-located DSL systems, which is introduced in this thesis in Chapter 4. This class of noise typically affects a number of sub-channels. For PLC, the dominant wide-band interference source occurs due to the US-DSL transmission. Further detail on the effect of DSL on the performance of a co-located PLC system is presented in Section

66 Chapter 3 CROSS-COUPLING CHANNEL MEASUREMENTS The main goal of this thesis is to mitigate the interference between co-located PLC and DSL networks, within the home environment. The interference mitigation solutions proposed in this thesis utilize adaptive filters to extract an estimate of the DM interference from the CM signal. As will be discussed in Chapter 5, both the DM and CM PLC-to-DSL cross-coupling channel frequency response matrices are used in the training of the adaptive filters, and in the performance evaluation of the proposed interference mitigations solutions. Field measurements of the interference environment between a DSL modem and a co-located in-home PLC network have never been performed. In addition, a model that describes the PLC-to-DSL cross-coupling channels within a residential house does not exist. Thus, a field measurement campaign is required to study the interference environment between a DSL modem and an in-home PLC network. The measurement campaign, introduced in this chapter, characterizes both the DM and CM PLC-to-DSL cross-coupling channels in two test-sites. A vector network analyzer (VNA) based measurement system is developed to measure the complex PLC-to-DSL cross-coupling channels. Measurements were performed in residential houses to study the effect of the co-located DSL and PLC systems on each other. Among other findings, analysis of the measurement data reveals that the PLC-to-DSL cross-coupling channels are frequency selective and the spatial separation between the PLC and DSL modems have no significant impact on the interference levels. The rest of this chapter is organized as follows. In Section 3.1, the DSL modem 48

67 installation scenarios are discussed. The PLC-to-DSL interference environment is discussed in Section 3.2; the methodology used to measure the DM and CM PLCto-DSL cross-coupling channels is discussed in Section 3.3. Finally, the measurement campaign, which is composed of two case studies, is presented in Section DSL Modem Installation Scenarios The DSL modem installation scenarios used in the measurement campaign are presented in this section. Within a house, the DSL copper twisted-pair is terminated by a modem using one of two DSL modem installation scenarios: the Desk Modem Scenario and the Entry Point Scenario. The PLC-to-DSL cross-coupling channels, presented in Section 3.4, are measured for both the Desk Modem Scenario and the Entry point Scenario. Additionally, the performance of the interference mitigations solutions, proposed in Chapter 5, is evaluated for both scenarios. TB LPF Telephone Lines LPF DSL Transceiver Broadband Router Figure 3.1: Desk Modem Scenario. In the Desk Modem Scenario, shown in Fig. 3.1, the DSL modem is located on a desk within the house. The twisted-pair carrying the DSL signal to and from the 49

68 house is connected to a terminal block (TB). The TB is connected to the house s interior telephone wiring, over which the broadband DSL signal is carried through the house to a DSL modem. Low pass filters (LPFs) are installed on each telephone to prevent audio distortion from the DSL signal. TB LPF DSL Transceiver Telephone Lines Broadband Router Figure 3.2: Entry Point Scenario. The Entry Point Scenario, on the other hand, considers the situation where the DSL modem is installed where the telephone cable enters the house, as shown in Fig The signal from the DSL modem is distributed throughout the house either via Wi-Fi, Ethernet, or over the house interior coaxial TV cable to one or more set top boxes. In addition, an in-line LPF is installed after the TB for the household telephone cable so that the DSL signal is blocked from travelling within the house. This makes installing LPFs on each telephone unnecessary. Both scenarios have their own advantages and disadvantages. The Desk Modem Scenario carries the DSL signal deep into the house, which is beneficial if the Internet is carried throughout the house via a WLAN. The Entry Point Scenario, on the other hand, minimizes the EMI suffered by the DSL signal, because the LPF that prevents the DSL signal from travelling within the house over the house s interior telephone 50

69 wiring also blocks the EMI that coupled on the house s interior telephone wiring from affecting the DSL signal. Thus, to reduce the cross-coupling between DSL and in-home PLC networks, the Entry Point scenario is preferred. In addition, installing the DSL transceiver within the house requires the installation of extra LPFs for each telephone outlet to prevent interference to the voice band. However, since most customers utilize wireless routers to carry the Internet throughout the house, terminating the modem within the house, i.e., the Desk Modem Scenario, allows for a better coverage. 3.2 DSL and PLC Interference Environment Power lines are not designed to communicate data at high frequencies; rather, the main purpose of power lines is to supply alternating high voltages at very low frequency. Since power lines are not shielded, portions of the signals transmitted over the power lines will radiate. Fig. 3.3 shows the interference environment between co-located DSL and PLC networks, within the home environment. In this figure, a DSL modem and an inhome PLC-network that operate over the same spectrum co-exist. Note that the DSL modem is installed via the Entry Point Scenario discussed in Section 3.1 and it supplies the house with Internet, while the PLC network forms a LAN that carries the Internet within the house. The power lines within the house form a tree with various branches that connect the various rooms of the house. Any signal transmitted over the power lines travel through all the branches of the power line tree; as the signal travels down the power line tree, portion of it is radiated. Due to the radiation from the various branches of the power line tree, the house is transformed into a large antenna. This radiation is picked up by the copper twisted-pair. As will be confirmed by the measurements 51

70 TB LPF DSL Transceiver Broadband Router Power Lines PLC Transceiver Telephone Lines PLC-DC PLC Transceiver PLC Transceiver Figure 3.3: Co-located DSL and PLC networks interference environment. presented in Section 3.4, this PLC radiation causes significant interference to the DSL network. Similarly, due to imperfections in the twisted-pairs, a portion of the DSL signal is radiated. This radiation is picked up by the power lines; however, since the DSL signal travels from and to the house, the level of DSL-to-PLC interference varies significantly between upstream and downstream DSL transmission. Further detail on the levels of DSL interference on a co-located PLC network is discussed in Section

71 3.3 Measurement Methodology In this section, the methodology behind the channel measurement campaign is described. Recall, the main purpose of the measurements is to characterize the DM and CM PLC-to-DSL cross-coupling channels. Since both DSL and PLC systems utilize multi-carrier FFT based modulation, as discussed in Chapter 2, measuring the crosscoupling channels in the frequency domain is appropriate. The hardware utilized in the measurement campaign is described in Section 3.3.1, while the calibration process is discussed in Section Finally, the measurement setup is introduced in Section Measurement Hardware To study the cross-coupling channels between PLC and DSL networks within the home environment, an Agilent E5071B vector network analyzer (VNA) with an operating range of 300 khz to 8.5 GHz is utilized. One port of the VNA is connected to the twisted-pairs within the house, while the second port of the VNA is connected to the power lines. The scatter matrix between the two ports of the VNA is obtained, which is essentially the cross-coupling channels between the PLC and DSL networks. However, since a VNA has coaxial cable outputs (which are unbalanced with respect to the ground), and both DSL and PLC utilize DM signalling (which is composed of two complementary signals that are balanced with respect to the ground), the unbalanced coaxial ports of the VNA have to be matched to the balanced DSL and PLC networks [14]. To connect the balanced DSL network to the VNA, a balun is used. A balun is essentially a resistance transformer that converts a balanced input to an unbalanced output and vice versa. Fig. 3.4 shows the North Hills 0320BF Balun utilized in the measurement campaign. The 0320BF Balun operates over the range of 10 khz to 30 53

72 Figure 3.4: North Hills 0320BF Balun. MHz. This balun has two 50 Ω unbalanced coaxial ports (labelled J1 and J2 in Fig. 3.4) and two balanced ports with a combined 100 Ω resistance (labelled 1 and 2 in Fig. 3.4). To use the balun in differential mode, it is connected to the VNA via port J1; while for common mode readings, the balun is connected to the VNA via port J2. Note that the twisted-pairs are connected to the Balun via ports 1 and 2, shown in 3.4. Figure 3.5: Northern Microdesign PLC Coupler. Fig. 3.5 shows the Northern Microdesign PLC coupler that operates over the range from 1.8MHz to 39MHz, which is utilized to connect the VNA to the power lines 54

73 within the house. The PLC coupler has two sub-miniature version A (SMA) ports and a three-pin plug to connect to the the power lines via the mains outlets within the house. Both SMA ports are connected to the differential output of the coupler. Thus, to inject a DM signal onto the power lines, the PLC coupler is connected to the VNA via one of the SMA ports. Note that an SMA to BNC (baby N connector) adaptor is required to connect the coupler to the VNA coaxial ports Calibration Balun PLC Coupler 1.5 Gain in db Frequency in MHz Figure 3.6: Insertion loss of the balun and the PLC coupler before calibration. Fig. 3.6 shows the insertion loss due to the balun and the PLC coupler. Thus, before collecting the measurements, the VNA is calibrated to remove the effects of the balun and the PLC coupler. However, before calibration both the Balun and the PLC coupler have to be modified. A registered jack (RJ) is connected to ports 1 and 2 of the balun to facilitate the calibration procedure and promote a more secure connection to the twisted-pairs as shown in Fig.3.7. A female 3-pin socket is soldered to a BNC adaptor to enable the calibration of the PLC coupler as shown in Fig Calibration of the VNA, Balun, and PLC coupler is required to eliminate any 55

74 Figure 3.7: Modified Balun. Figure 3.8: Modified PLC Coupler. effect on the collected measurements. Each port of the VNA is calibrated for four scenarios: open, short, load, and through (between ports). A standard calibration kit provided by the manufacturer is usually utilized in the calibration process; however, since standard calibration kits are designed for the 50 Ω VNA ports, and since both the balun and PLC coupler require a 100 Ω calibration kit, a customized calibration kit is utilized in the calibration process. Three RJ 45 jacks are utilized for the open, short, and load calibration, as shown in Fig The through calibration, shown in Fig. 3.10, is performed via a pair of wires that are connected to a BNC adaptor at one end and to an RJ 45 jack at the other. 56

75 Figure 3.9: Right to Left: Open, Short, and Load (100 Ω). Figure 3.10: Through calibration Measurement Setup After calibration, the scatter matrix between port 1 and port 2 of the VNA is used to determine the PLC-to-DSL cross-coupling channels. One port of the VNA is connected to the power lines of the house through the PLC coupler, while the second port is connected to the telephone wires in the house through the balun, as shown in Fig The VNA is set to output 0 dbm sinusoidal signals and to collect 1500 samples, between frequencies 300 khz and 30 MHz. To measure the DM PLC-to-DSL channel frequency response matrix the balun is connected in the DM mode to the VNA, as shown in Figure 3.11(a). To measure the CM PLC-to-DSL cross-coupling channel, the balun is connected in the CM mode to the VNA, as shown in Figure 3.11(b). 57

76 VNA Port 1 Port 2 PLC adapter Balun Power Network Telephone Network Coupling (a) Setup for measuring DM coupling. VNA Port 1 Port 2 PLC adapter Balun Power Network Telephone Network Coupling (b) Setup for measuring CM coupling. Figure 3.11: Setup for PLC-to-DSL coupling. To study the stationarity of the PCL-to-DSL cross-coupling channels, measurements of the the channels have to be collected over time. An Agilent E5810A general purpose interface bus (GPIB) controller is used to control an Agilent E5071B vector network analyzer (VNA), via MATLAB as shown in Fig Note that the setup shown in Fig is required when consecutive measurements of the channel are needed. If the variation of the channel over time is not a concern, the GPIB is not necessary, and thus, the setup shown in Fig 3.11 is sufficient. In the literature, the background noise on the DM DSL line is usually assumed to be -140 dbm [4]. However, a similar assumption for the CM background noise Hz does not exist. Both the DM and the CM background noise levels are required to study the performance of the adaptive filters utilized by the interference mitigation solutions proposed in this thesis. The background noise present on the CM DSL line was measured with a spectrum analyzer which showed the background noise on the CM DSL channel is approximately -120 dbm Hz. 58

77 GPIB Controller VNA Port 1 Port 2 PLC adapter Balun Power Network Telephone Network Coupling MATLAB Balun Balun Differential Mode Setup Common Mode Setup Figure 3.12: GPIB controlled measurement setup. 3.4 Measurement Campaign The measurement campaign, introduced in this section, is composed of two case studies: Case Study A and Case Study B. The measurements were conducted in two residential test-sites (one site per case study). Since no field measurements have been conducted in a residential house before Case Study A, the case study was designed with two objectives in mind. The first objective is identifying the level of interference between a DSL modem and a colocated PLC network in the home environment. Identifying and determining the relationship between the DM and CM PLC-to-DSL cross-coupling channels within a residential setting is the second objective of Case study A. Note that only the Desk Modem Scenario was considered in Case Study A because the test-site s wiring did not permit measuring the cross-coupling channels for the Entry Point Scenario. From Case Study A, the existence of strong frequency selective cross-coupling channels between the DSL and PLC networks is identified. In addition, it was noticed that the cross-coupling channels were not dependent on the spatial separation between the PLC and DSL modems. However, Case Study A considered only the Desk Modem 59

78 Scenario. In addition, the variation of the cross-coupling channels over time was not studied. Case Study B was a more extensive case study, where a set of 480 DM and CM cross-coupling channel measurements between the DSL and PLC networks is collected, within a residential house. Also, both DSL modem installation scenarios discussed in Section 3.1 are considered in Case Study B. The rest of this section is organized as follows. In Section 3.4.1, the test-sites for both case studies are discussed, while the measurements results are presented in Section Test-Sites The test-sites for the two case studies of the measurement campaign are discussed in this section. In Section , the test-site of Case Study A is presented, while the test-site of Case Study B is discussed in Section Case Study A In this case study, only the Desk Modem Scenario discussed in Section 3.1, is considered. By not terminating the copper twisted-pair at the entrance of the house, the distance over which the twisted-pair DSL line coexists with the PLC power lines increases. This, in-turn, increases the coupling between the two systems. Both DM and CM cross-coupling channels were measured in four rooms, all on the same level, within a 1000 ft 2 residential house. The floor plan of the house is shown in Figure The circle indicates the location of the telephone outlet (i.e., the location of the DSL modem within the house), while the triangles indicates the various locations of the power outlets for which the cross-coupling channels were studied. Note that in Case Study A, the variation of the PLC-to-DSL cross-coupling channels over time is not considered, and thus, the setup shown in Fig 3.11 was utilized. 60

79 40 ft Room D Room B 25 ft Room C Case Study B Room A Telephone Outlet Power Outlet Figure 3.13: Case Study A: test-site floor plan. In Case Study B, a new measurement case study is conducted. The PLC-to-DSL cross-coupling channels within a residential house are measured. Note that the setup shown in Fig was used in this case study. A set of 480 measurements is collected to characterize the DM and CM PLC-to-DSL cross-coupling channels, for both DSL modem installation scenarios discussed in Section 3.1 (i.e., the Desk Modem Scenario and the Entry Point Scenario). The floor plan of the upper and lower floors of the residential single storey bungalow, which is used as a test-site, are shown in Fig. 3.14(a) and Fig. 3.14(b), respectively. The bungalow was built in 1930 and renovated to modern wiring standards in 1992, when the telephony network was professionally installed. Note that the distance between the two floors is approximately 9 ft. Within each room, there are multiple power outlets. Each power outlet where a PLC-to-DSL coupling measurement was collected is labelled. Thus, Room A Plug 1 refers to the PLC-to-DSL cross-coupling channels for Plug 1 in Room A. Note that the location of the DSL modem for the 61

80 Room F Room E Telephone Plug Power Outlet Cable T.V. 48 ft 1 1 Room G 1 Room Room I Room J 1 H DN 92 ft (a) Upper floor. 3 2 Room D 2 Telephone Plug Power Outlet Cable T.V. 42 ft 1 1 Room A Room C Room B UP 72 ft (b) Lower floor. Figure 3.14: Case Study B: test-site floor plan. Desk Modem Scenario and the Entry Point Scenario within the measurement house are in Room J in Fig. 3.14(a) and Room C in Fig. 3.14(b), respectively. For each of the measured power outlets, 10 scatter matrix measurements are collected 0.5 seconds apart. This is performed for both the Desk Modem Scenario and the Entry Point Scenario to study the changes in the CM and DM PLC-to-DSL crosscoupling channels over time. A single set of measurements, which consists of 1500 measurement points, is completed in 0.5 seconds. Both the smoothing (averaging over frequency) and averaging (averaging over time) functions of the VNA were turned off 62

81 during the measurements, and the VNA was not synchronized with the AC cycle. Given that the AC cycle has a duration of 1/60 seconds, over the 10 measurement sets, each measurement point has a different location in the AC cycle. Thus, any variation detected in the PLC-to-DSL cross-coupling channel measurements is attributed to the change in the loads of the power line network Results This section presents the results of the measurement campaign. The frequency responses of the measured DM and CM PLC-to-DSL cross-coupling channels are presented in Section The stationarity of the cross-coupling channels is discussed in Section , while the effect of the spatial separation on the coupling gain and coherent bandwidth of the cross-coupling channels is studied in Section Cross-Coupling Channel Frequency Responses To determine the frequency response matrix for a given channel, the scatter matrix between the ports of the VNA is used. In Section , the cross-coupling channels measured in Case Study A are presented, while the cross-coupling channels measured in Case Study B are discussed in Section Case Study A Fig shows the DM PLC-to-DSL cross-coupling channels, for Rooms A, B, C, and D. From Fig the following is noticed. First, the measurements indicate a very high degree of DM coupling between the PLC and DSL channels. In addition, it also indicates that the DM PLC-to-DSL cross-coupling channels are frequency selective. The figure also shows that the relative distance between the telephone outlet and the power outlet has no significant impact on the PLC-to-DSL coupling. Fig shows the CM PLC-to-DSL cross-coupling channels, respectively, for Rooms A, B, C, and D. Similar to the findings of the DM PLC-to-DSL cross-coupling 63

82 30 40 DM coupling Channel Gain in db Room A 90 Room B Room C Room D Frequency in MHz Figure 3.15: Coupling in Differential Mode. channel measurements, the CM PLC-to-DSL cross-coupling channel measurements show the expected high degree of coupling between the PLC and DSL channels in the common mode. Additionally, Fig indicates that the CM PLC-to-DSL crosscoupling channel is frequency selective and the relative distance between the telephone outlet and the power outlet has no significant impact on the PLC-to-DSL coupling. Finally, by comparing Figs and 3.16, it is noted that the DM coupling is lower than the CM coupling, which is expected since DM transmission is used to mitigate the effects of EMI. The CM to DM transfer function (C2DTF), which is defined as the ratio of the DM to CM cross-coupling channels, is calculated by G(i) = g d(i) g c (i) (3.1) where, G(i) is the C2DTF ratio for frequency bin i, and g d (i) and g c (i) are the DM and CM channel gain vectors for frequency bin i respectively. Fig shows that the C2DTF is quite different for each room. This is due to the tree topology of the power line network within the house. The various branches of the power line network cause either constructive or destructive interference to the DSL network. This tree 64

83 20 30 CM coupling Channel Gain in db Room A 80 Room B Room C Room D Frequency in MHz Figure 3.16: Coupling in Common Mode. also causes the relationship between the DM and CM cross-coupling channels to vary significantly from one power outlet to another. In essence, the topology of the network changes from room to room which causes the variation in the C2DTF Case Study B Fig shows the outlet-to-outlet variation in the DM PLC-to-DSL cross-coupling channels, for the Desk Modem Scenario. Only 5 cross-coupling channels are shown in Fig for the sake of clarity; however, variations among the DM and the CM PLCto-DSL cross-coupling channels for all labelled outlet in Fig have been studied. It was observed that both the DM and CM PLC-to-DSL cross-coupling channels are frequency selective and vary significantly from one power outlet to another, for both the Desk Modem Scenario and the Entry Point Scenario. Note that, as discussed in Section , the frequency selectivity of the cross-coupling channel, and its variation from one outlet to the other, is due to the tree structure of the power lines within the house. Similar to Room A Plug 1, the CM and DM PLC-to-DSL cross-coupling channels for the rest of the labelled power outlets in Fig were measured, for both 65

84 C2DTF Gain in db Room A 40 Room B Room C Room D Frequency in MHz Figure 3.17: Common mode to differential mode transfer function (C2DTF). Desk Modem Scenario and Entry Point Scenario. From these measurements, it was observed that the PLC-to-DSL cross-coupling channels are frequency selective, and vary significantly from one power outlet to another. In addition, the relationship between the DM and the CM cross-coupling channels is not the same for all power outlets. Note that these findings agree with the finding of Case Study A. To study the distribution of the coupling gains of the DM and CM PLC-to-DSL cross-coupling channels, the empirical probability density functions (PDFs) of the DM and CM PLC-to-DSL coupling gains, for both the Desk Modem Scenario and the Entry Point Scenario are generated. To generate the PDF for each PLC-to- DSL cross-coupling channel, a histogram of the gains at all frequency points, for the measured channels, is generated and normalized such that area under the curve tracing the heights of the histogram bins is equal to one. 66

85 30 40 Channel Gain in db Frequency in MHz Figure 3.18: Variation in DM cross-coupling channels from one power outlet to another. Emperical Probability Density Function DM Mean DM CM Mean CM Coupling Gain in db Figure 3.20: Entry Point Scenario coupling gain probability density function. For each scenario, the means of the DM and CM PLC-to-DSL cross-coupling channels, represented by vertical lines in Figs and 3.20, are calculated. Note that, on average, the DM coupling for the Entry Point Scenario is approximately 20 db lower than the DM coupling for the Desk Modem Scenario, which indicates the significant effect of the in-line LPF on the reduction of the DM interference. 67

86 Emperical Probability Density Function DM Mean DM CM Mean CM Coupling Gain in db Figure 3.19: Desk Modem Scenario coupling gain probability density function. However, as will be shown in Chapter 4, this reduction is not sufficient to prevent the PLC system from degrading the performance of the DSL system Coupling Gain in db Desk Modem: Mean DM 90 Desk Modem: Mean CM Entry Point: Mean DM Entry Point: Mean CM Frequency in MHz Figure 3.21: Mean DM and CM PLC-to-DSL coupling. The means of the DM and CM PLC-to-DSL cross-coupling channels for all rooms, for both the Desk Modem Scenario and the Entry Point Scenario, are shown in Fig The mean for each scenario represent the average of the gains for a particular frequency across all the measurement locations shown in Fig. 3.14, i.e., these plots 68

87 show the cross-coupling channels averaged across all outlets used to inject the PLC signal. From Fig. 3.21, it is noted that both the DM and CM PLC-to-DSL crosscoupling channels are frequency selective. Furthermore, it is noted that the in-line LPF present in the Entry Point Scenario reduces the amount of DM coupling relative to the Desk Modem Scenario. However, the CM coupling for both scenarios is not affected by the in-line LPF. In addition, it is evident that the CM coupling is higher than the DM coupling, for both scenarios. This is expected since the DM reception mode partially cancels EMI Stationarity of Cross-Coupling Channels Figure 3.22: Desk Modem Scenario.: DM PLC-to-DSL coupling for Room A Plug 1 over measurement interval. Figs and 3.23 show the changes in the DM PLC-to-DSL coupling for Room A Plug 1 over time for the Desk Modem Scenario and the Entry Point Scenario respectively. Recall, for each power outlet, the PLC-to-DSL cross-coupling channels are measured for 10 consecutive times, with a separation interval of 0.5 seconds, to study the changes in the channel over time. 69

88 Figure 3.23: Entry Point Scenario.: DM PLC-to-DSL coupling for Room A Plug 1 over measurement interval Channel gain in db DM Coupling CM Coupling Frequency in MHz Figure 3.24: Desk Modem Scenario: Room A Plug 1 DM vs. coupling. CM PLC-to-DSL 70

89 Channel gain in db DM Coupling CM Coupling Frequency in MHz Figure 3.25: Entry Point Scenario: Room A Plug 1 DM vs. CM PLC-to-DSL coupling. For each scenario, the average channel gains of the DM and CM PLC-to-DSL cross-coupling channels, for Room A Plug 1, are calculated, and the variations in the channel gains over time is determined by calculating the standard deviation of the channel gains at each frequency point. The average DM and CM PLC-to-DSL crosscoupling channels for Room A Plug 1, for the Desk Modem Scenario and the Entry Point Scenario, are shown in Figs and 3.25 respectively. Also, the variations in the channel gains over time are indicated via error bars in Figs and From Figs. 3.22, 3.23, 3.24, and 3.25 it is noted that the channel gain variations of the PLC-to-DSL cross-coupling channels over time are insignificant. However, over frequency, the variation of the PLC-to-DSL cross-coupling channel gains is quiet significant. Also, it is evident that the CM coupling is higher than the DM coupling, for both scenarios. In addition, it is noted that the DM and CM cross-coupling channels are frequency selective, for both scenarios. Additionally, for each room, the standard deviation for each frequency bin over the 10 measurements is determined; after which, the standard deviation values are 71

90 averaged across all the rooms and then across all the frequency bins. The average standard deviations for the DM and CM PLC-to-DSL cross-coupling channels for the Desk Modem Scenario are 1.06 and 0.98 db respectively, while for the Entry Point Scenario the average standard deviations for the DM and CM PLC-to-DSL coupling are 1.87 and 1.24 db respectively. These numbers indicate that the variation in the channel coupling gains over time is insignificant, and thus, the PLC-to-DSL crosscoupling channel is considered stationary. The measurements were performed in a furnished house, albeit, most of the appliances were not actively drawing power during the measurements. Only the fridge, deep freezer, and coffee maker were on during the measurement campaign. Thus, the test-site was not very electrically active. As will be discussed in Chapter 5, the proposed interference mitigation solutions presented in this thesis are insensitive to changes in the cross-coupling channels due to the non-stationarity of the PLC channel. For a given outlet, any variations in the PLC interference is reflected in both the DM and CM cross-coupling channels simultaneously. Additionally, the proposed interference mitigation solutions utilize adaptive filters that estimate the ratio of the DM to CM PLC-to-DSL cross-coupling channels. Since variations in the PLC interference simultaneously affect both the DM and CM PLC-to-DSL cross-coupling channels, the ratio of the DM to CM PLC-to-DSL cross-coupling channels for a given room remains constant over time Effect of Spatial Separation Fig shows the average DM and CM PLC-to-DSL cross-coupling channel gains versus the relative distance between the DSL and PLC modems for all the measurement locations. Each point represents the cross-coupling channel gain averaged across the measurement frequency range. Note that the Euclidean distance between the three dimensional positions of the DSL modem and the PLC modem is used in 72

91 30 40 Average Coupling Gain in db DM: Desk Modem DM: Entry Point CM: Desk Modem CM: Entry Point Euclidean Distance in ft Figure 3.26: Effect of relative distance between DSL and PLC modems on PLC to-dsl coupling. Fig Information on the actual run length of the cables was unavailable. From this figure, it is noted that the relative distance between the DSL modem and the PLC modem does not have a significant impact on the coupling, since the coupling levels do not decrease as the relative distance between the modems increase. This is likely because the unshielded power lines radiate interference approximately uniformly throughout the home due to the length of the power lines and the relatively small size of the home. Finally, the 90% coherence bandwidth (CB) of both the DM and CM PLC-to-DSL cross-coupling channels versus the Euclidean distance between the PLC and DSL modem is shown in Fig The 90% coherence bandwidth (CB) was determined by calculating the autocorrelation of each channel and determining the width of the bandwidth at which the magnitude of autocorrelation drops to 90% of its maximum value. Table 3.1: Average coherence bandwidth in khz. Scenario DM CM Desk Modem Entry Point

92 Coherence Bandwidth in khz DM: Desk Modem CM: Desk Modem DM: Entry Point CM: Entry Point Euclidean Distance in ft. Figure 3.27: Effect of relative distance between DSL and PLC modems on cross-coupling channels coherence bandwidth. It is noted that, on average, the CB for the DM and CM PLC-to-DSL crosscoupling channels for Desk Modem Scenario is lower than the CB for Entry Point Scenario, as shown in Table 3.1. This due to the fact that by allowing the DSL signal to travel within the house, more branches of the PLC tree contribute (constructively or destructively) to the cross-coupling channels. 74

93 Chapter 4 INTERFERENCE SYSTEM MODEL In Chapter 3, the cross-coupling channels between DSL and PLC networks within a residential setting have been measured. Two case studies have been performed, and among the findings was a strong DM cross-coupling channel between DSL and PLC networks. As a result of this coupling, the mutual interference between two co-located DSL and PLC networks, operating over the same frequency band, will inevitably affect the performance of both systems. Thus, an interference mitigation solution is required. In this chapter, we quantify the effect of the mutual DSL and PLC interference on the data rates of existing DSL and PLC systems. The current system model, which describes the interference environment between PLC and DSL systems as it exists today, is discussed in Section 4.1. The data rates achieved by a DSL modem in presence of an in-home PLC network and the data rates achieved by an in-home PLC network in presence of a DSL modem are studied in Section Current System Model Within a residential house, the DSL and PLC systems utilize the twisted-pairs and the power lines, respectively, as transmission mediums. The twisted-pairs are composed of two twisted copper wires, designed to minimize both EMI egress and ingress. On the other hand, power lines are made of two or three wires. The power lines are neither twisted nor shielded, which result in significant electromagnetic radiation. Fig. 4.1 shows the current system model for the PLC-to-DSL interference environment within a residential setting. In each house, there is a single DSL transceiver 75

94 DSL system Twisted-pair DSL Transceiver Balun DMT Transceiver Wide-band EMI PLC-D.C. Power Lines PLC Transceiver PLC Transceiver PLC Transceiver PLC Coupler DMT Transceiver PLC system PLC Transceiver Figure 4.1: Current system model. while there are multiple PLC transceivers. The measurement campaign presented in Chapter 3 indicates that the cross-coupling channels between the DSL and PLC systems vary significantly from one PLC transceiver to the other. The DSL transceiver is connected to the copper twisted-pairs, over which the DSL signal is carried to and from the house. The DSL transceiver is composed of a balun that outputs a DM signal, which is then passed to a DMT transceiver, as shown in Fig Similarly, the PLC transceivers form a network through the power lines which run through the house. Each PLC transceiver is composed of a PLC coupler that feeds its output to a DMT transceiver, as shown in Fig Note that the 76

95 structure of the DMT transceiver was discussed in Section Effect of Mutual DSL and PLC Interference on Bit Rates The measured cross-coupling channels presented in Chapter 3 are utilized to visualize the effect of the PLC interference on the received DSL signal in Section 4.2.1; while, the degradation of the PLC system performance due to the DSL interference is studied in Section Note that the measured cross-coupling channels are reciprocal; however, as will be shown in the following sections, the effect of the PLC and the DSL networks on one another is not Effect of PLC Interference on DSL Bit Rates To determine the effect of the PLC interference on the received DSL signal, the measured PLC-to-DSL cross-coupling channels are used to calculate the power spectral density (PSD) of the PLC interference. The PSD of the PLC interference is then compared with the PSD of a typical received DSL signal. Note that the direct DSL channel, used to determine the PSD of the received DSL signal, is obtained from the standard two-port model defined in [29]. Similarly, the average of the measured DM PLC-to-DSL cross-coupling channels, presented in Section 3.4 and shown in Fig. 3.21, is utilized to determine the PSD of the PLC interference. The PSDs of the received DSL signal for various DSL cable run lengths versus the PSD of the PLC interference, for both the Desk Modem Scenario and the Entry Point Scenario, are shown in Fig The maximum transmit power of both the DSL and PLC systems, as specified in their respective ITU-T standards, is multiplied by the DSL channel and the mean DM PLC-to-DSL cross-coupling channel to determine the PSD of the signals. For VDSL2, the maximum transmit power is -50 dbm [2]; while Hz for in-home PLC networks, the maximum allowed transmit power is -60 dbm Hz [9]. 77

96 PSD in db m Hz DSL: Recieved at 2kft 200 DSL: Recieved at 4kft PLC: Desk Modem Scen. PLC: Entry Point Scen Frequency in MHz Figure 4.2: Received DSL Signal versus PLC interference. Fig. 4.2 illustrates two points. First is the extremely high level of PLC interference on the DSL line for both the Desk Modem Scenario and the Entry Point Scenario. Although, in the Entry Point Scenario, the in-line LPF reduces the PLC interference levels, this reduction is not sufficient to prevent the DSL system from experiencing low signal to interference plus noise ratio (SINR) over the VDSL2 spectrum. The second observation is, while the PLC interference levels are unaffected by the distance between the central office (or distribution point) and the house, the received DSL signal power decreases as the distance increases. Thus, the PLC interference does not only impact VDSL2 services, but also it poses a significant risk to DSL services that utilize narrower spectrum, such as ADSL2+ (up to 2.2 MHz [39]) and VDSL (up to 12 MHz [40]), as the DSL cable run increases. Bit loading, where the number of bits allocated to each sub-channel is dependent on the sub-channel s SINR, is used in both DSL and PLC [22]. The number of bits b(i) that can be loaded on frequency bin i is calculated by ( ( b(i) = min b max, log γ(i) ) ), (4.1) Γ where γ(i) is the SINR of frequency bin i, Γ is the SNR gap, and b max is the maximum 78

97 number of bits that can be allocated to a frequency bin. After determining the number of bits that can be loaded on each frequency bin, the total bit rate R is then calculated by R = λ f i b(i), (4.2) where λ is the normalizing factor and f is the sub-carrier spacing [27]. Note that the value of b max, λ, and f varies from one system to the other. In presence of AWGN with PSD of N o = -140 dbm, the signal to the interference plus noise power is used Hz to determine the available bits that can be loaded to each of the DSL sub-channels via (4.1). After which, from (4.2), the total number of bits is determined. Note that for VDSL2, b max =15 bits, Γ=9.45 db, λ=0.79, and f DSL =8.6 khz [27]. Recall, DSL utilizes FDD, where the sub-channels are grouped and the groups are allocated to US and DS transmission in an alternating fashion. The number of sub-channels per transmission direction varies from one band plan to the other. Note that band plan 998E30 [2], shown in Fig. 4.3, is utilized in this thesis. Downstream Upstream Figure 4.3: US and DS frequencies for band plan 998E30 [2]. Fig. 4.4 shows the available DSL bit rates vs the length of the DSL cable run lengths. Six plots are shown in Fig. 4.4, where Upstream and Downstream indicate 79

TELECOMMUNICATIONS STANDARDS ADVISORY COMMITTEE WORKING GROUP ON COMMON CONNECTION STANDARDS (CCS)

TELECOMMUNICATIONS STANDARDS ADVISORY COMMITTEE WORKING GROUP ON COMMON CONNECTION STANDARDS (CCS) TELECOMMUNICATIONS STANDARDS ADVISORY COMMITTEE WORKING GROUP ON COMMON CONNECTION STANDARDS (CCS) Overview of Very High Speed Digital Subscriber Line 2 (VDSL2) 甚 高 速 數 碼 用 戶 線 路 2 的 概 覽 Introduction ADSL

More information

Broadband 101: Installation and Testing

Broadband 101: Installation and Testing Broadband 101: Installation and Testing Fanny Mlinarsky Introduction Today the Internet is an information superhighway with bottlenecks at every exit. These congested exits call for the deployment of broadband

More information

Evolution from Voiceband to Broadband Internet Access

Evolution from Voiceband to Broadband Internet Access Evolution from Voiceband to Broadband Internet Access Murtaza Ali DSPS R&D Center Texas Instruments Abstract With the growth of Internet, demand for high bit rate Internet access is growing. Even though

More information

www.zte.com.cn VDSL2 A feasible Solution for Last Mile

www.zte.com.cn VDSL2 A feasible Solution for Last Mile www.zte.com.cn VDSL2 A feasible Solution for Last Mile Version Date Author Approved By Remarks V1.00 009-08-8 MichaelSong Not open to the Third Party 009 ZTE Corporation. All rights reserved. ZTE CONFIDENTIAL:

More information

Current access technologies overview

Current access technologies overview White Paper Current access technologies overview In this paper, we explore six basic technology choices for deploying broadband services to the end customer xdsl, DOCSIS, G.fast, satellite, wireless and

More information

(Refer Slide Time: 2:10)

(Refer Slide Time: 2:10) Data Communications Prof. A. Pal Department of Computer Science & Engineering Indian Institute of Technology, Kharagpur Lecture-12 Multiplexer Applications-1 Hello and welcome to today s lecture on multiplexer

More information

Public Switched Telephone System

Public Switched Telephone System Public Switched Telephone System Structure of the Telephone System The Local Loop: Modems, ADSL Structure of the Telephone System (a) Fully-interconnected network. (b) Centralized switch. (c) Two-level

More information

RADIO FREQUENCY INTERFERENCE AND CAPACITY REDUCTION IN DSL

RADIO FREQUENCY INTERFERENCE AND CAPACITY REDUCTION IN DSL RADIO FREQUENCY INTERFERENCE AND CAPACITY REDUCTION IN DSL Padmabala Venugopal, Michael J. Carter*, Scott A. Valcourt, InterOperability Laboratory, Technology Drive Suite, University of New Hampshire,

More information

Residential Broadband: Technologies for High-Speed Access To Homes

Residential Broadband: Technologies for High-Speed Access To Homes Residential Broadband: Technologies for High-Speed Access To Homes The Ohio State University Columbus, OH 43210-1277 1277 http://www.cse.ohio-state.edu/~jain/ 1 Overview 56 kbps Modems, ISDN ADSL, VDSL

More information

Residential Broadband: Technologies for High-Speed Access To Homes

Residential Broadband: Technologies for High-Speed Access To Homes Residential Broadband: Technologies for High-Speed Access To Homes Columbus, OH 43210 Jain@CIS.Ohio-State.Edu http://www.cis.ohio-state.edu/~jain/cis788-97/ Email questions to mbone@netlab.ohio-state.edu

More information

ECE 510 -- Chapter 1

ECE 510 -- Chapter 1 ECE 510 -- Chapter 1 Definition: Digital Subscriber Line (DSL) Public network technology that delivers high bandwidth over conventional copper wiring at limited distances. There are four major types of

More information

Digital Subscriber Line (DSL) Transmission Methods

Digital Subscriber Line (DSL) Transmission Methods Digital Subscriber Line (DSL) Transmission Methods 1. Overview... 1 2. SHDSL Transmission Methods... 1 SHDSL Transmission System Versions... 1 SHDSL Transmission Subsystem Structure... 1 SHDSL Modulation

More information

Data Transmission. Data Communications Model. CSE 3461 / 5461: Computer Networking & Internet Technologies. Presentation B

Data Transmission. Data Communications Model. CSE 3461 / 5461: Computer Networking & Internet Technologies. Presentation B CSE 3461 / 5461: Computer Networking & Internet Technologies Data Transmission Presentation B Kannan Srinivasan 08/30/2012 Data Communications Model Figure 1.2 Studying Assignment: 3.1-3.4, 4.1 Presentation

More information

How Enhanced DSL Technologies Optimize the Last Copper Mile By John Williams

How Enhanced DSL Technologies Optimize the Last Copper Mile By John Williams White Paper How Enhanced DSL Technologies Optimize the Last Copper Mile By John Williams As broadband speeds increase to support the connected home, a debate continues over which type of network architecture

More information

Broadband access. Nils Holte, NTNU. NTNU Department of Telecommunications Kursdagene ved NTNU, Digitale telenett, 9. januar 2002 1.

Broadband access. Nils Holte, NTNU. NTNU Department of Telecommunications Kursdagene ved NTNU, Digitale telenett, 9. januar 2002 1. Broadband access - properties of the copper network Nils Holte, Kursdagene ved, Digitale telenett, 9. januar 2002 1 Definition of broadband Overview Alternatives for fixed broadband access pair cable,

More information

Chapter 9 Using Telephone and Cable Networks for Data Transmission

Chapter 9 Using Telephone and Cable Networks for Data Transmission 9-11 TELEPHONE NETWORK Chapter 9 Using Telephone and Cable Networks for Data Transmission 1 McGraw-Hill Copyright The McGraw-Hill Companies, Inc. Permission required The for reproduction McGraw-Hill or

More information

Broadband Technology Clinic. Burlington Telecom Advisory Board

Broadband Technology Clinic. Burlington Telecom Advisory Board Broadband Technology Clinic Burlington Telecom Advisory Board 1 What are the Defining Characteristics of a Broadband Service? Speed - Throughput capability both down and upstream Performance - Latency

More information

Appendix A: Basic network architecture

Appendix A: Basic network architecture Appendix A: Basic network architecture TELECOMMUNICATIONS LOCAL ACCESS NETWORKS Traditionally, telecommunications networks are classified as either fixed or mobile, based on the degree of mobility afforded

More information

XDSL and DSLAM Access Technologies

XDSL and DSLAM Access Technologies XDSL and DSLAM Access Technologies Abstract What are the differences between the different forms of xdsl technology, such as ADSL and HDSL? How are they implemented. What are the limitations? What are

More information

A COMPARATIVE ON PERFORMANCE OF VOIP USING POWER LINE AND WIRED (UTP CAT5)

A COMPARATIVE ON PERFORMANCE OF VOIP USING POWER LINE AND WIRED (UTP CAT5) A COMPARATIVE ON PERFORMANCE OF VOIP USING POWER LINE AND WIRED (UTP CAT5) BY KHAINORIZAN BINTI KHALID BACHELOR OF SCIENCE (Hons) DATA COMMUNICATION AND NETWORKING Thesis proposal submitted in fulfillment

More information

Broadcast digital subscriber lines using discrete multitone for broadband access

Broadcast digital subscriber lines using discrete multitone for broadband access Microprocessors and Microsystems 22 (1999) 605 610 Broadcast digital subscriber lines using discrete multitone for broadband access K.-P. Ho* Department of Informational Engineering, The Chinese University

More information

Next Generation of High Speed. Modems8

Next Generation of High Speed. Modems8 Next Generation of High Speed Modems High Speed Modems. 1 Traditional Modems Assume both ends have Analog connection Analog signals are converted to Digital and back again. Limits transmission speed to

More information

Intel System Engineers Documents. DSL General Overview

Intel System Engineers Documents. DSL General Overview Intel System Engineers Documents DSL General Overview Alex Lattanzi SC LAR Whatt IIs Brroadband? Broadband describes a number of different technologies that deliver digital data to homes and businesses

More information

Detecting Bridged Tap and Noise Interference in VDSL2 Access Networks using the JDSU SmartClass TPS

Detecting Bridged Tap and Noise Interference in VDSL2 Access Networks using the JDSU SmartClass TPS Application Note Detecting Bridged Tap and Noise Interference in VDSL2 Access Networks using the JDSU SmartClass TPS The JDSU SmartClass TPS tester is the ideal tool for technicians who install, troubleshoot,

More information

Black Box Explains: DSL

Black Box Explains: DSL Black Box Explains: DSL History It was realized as early as the late eighties, early nineties, that conventional data transmission systems did not meet the requirements of the growing internet community

More information

Analog vs. Digital Transmission

Analog vs. Digital Transmission Analog vs. Digital Transmission Compare at two levels: 1. Data continuous (audio) vs. discrete (text) 2. Signaling continuously varying electromagnetic wave vs. sequence of voltage pulses. Also Transmission

More information

Module: Digital Communications. Experiment 784. DSL Transmission. Institut für Nachrichtentechnik E-8 Technische Universität Hamburg-Harburg

Module: Digital Communications. Experiment 784. DSL Transmission. Institut für Nachrichtentechnik E-8 Technische Universität Hamburg-Harburg Module: Digital Communications Experiment 784 DSL Transmission Institut für Nachrichtentechnik E-8 Technische Universität Hamburg-Harburg ii Table of Contents Introduction... 1 1 The DSL System... 2 1.1

More information

How DSL Works. by Curt Franklin

How DSL Works. by Curt Franklin by Curt Franklin How DSL Works When you connect to the Internet, you might connect through a regular modem, through a localarea network connection in your office, through a cable modem or through a digital

More information

Introduction to Digital Subscriber s Line (DSL)

Introduction to Digital Subscriber s Line (DSL) Introduction to Digital Subscriber s Line (DSL) Professor Fu Li, Ph.D., P.E. Chapter 3 DSL Fundementals BASIC CONCEPTS maximizes the transmission distance by use of modulation techniques but generally

More information

EECC694 - Shaaban. Transmission Channel

EECC694 - Shaaban. Transmission Channel The Physical Layer: Data Transmission Basics Encode data as energy at the data (information) source and transmit the encoded energy using transmitter hardware: Possible Energy Forms: Electrical, light,

More information

Broadband Primer. A Guide to High Speed Internet Technologies. Indiana Office of Utility Consumer Counselor

Broadband Primer. A Guide to High Speed Internet Technologies. Indiana Office of Utility Consumer Counselor Broadband Primer A Guide to High Speed Internet Technologies Indiana Office of Utility Consumer Counselor 100 N. Senate Av., Room N501 Indianapolis, IN 46204-2215 www.openlines.in.gov toll-free: 1-888-441-2494

More information

VDSL (VERY HIGH DATA BIT RATE DIGITAL SUBSCRIBER LINE)

VDSL (VERY HIGH DATA BIT RATE DIGITAL SUBSCRIBER LINE) 1 VDSL (VERY HIGH DATA BIT RATE DIGITAL SUBSCRIBER LINE) INTRODUCTION 1. Recent events in the telecommunications environment are giving rise to a new class of service providers, setting the stage for how

More information

INTRODUCTION TO COMMUNICATION SYSTEMS AND TRANSMISSION MEDIA

INTRODUCTION TO COMMUNICATION SYSTEMS AND TRANSMISSION MEDIA COMM.ENG INTRODUCTION TO COMMUNICATION SYSTEMS AND TRANSMISSION MEDIA 9/6/2014 LECTURES 1 Objectives To give a background on Communication system components and channels (media) A distinction between analogue

More information

Digital Subscriber Line (DSL)

Digital Subscriber Line (DSL) CHAPTER 9 This chapter describes the level of support that Cisco ANA provides for DSL, as follows: Technology Description, page 9-1 Inventory and Information Model Objects (IMOs), page 9-2 Vendor-Specific

More information

Introduction to ADSL. NEXTEP Broadband White Paper. Broadband Networks Group. A primer on Asymmetric Digital Subscriber Line transmission technology.

Introduction to ADSL. NEXTEP Broadband White Paper. Broadband Networks Group. A primer on Asymmetric Digital Subscriber Line transmission technology. NEXTEP Broadband White Paper Introduction to ADSL A primer on Asymmetric Digital Subscriber Line transmission technology. A NEXTEP Broadband White Paper May 2001 Broadband Networks Group Introduction to

More information

TELECOMMUNICATIONS STANDARDS ADVISORY COMMITTEE TSAC WORKING GROUP ON NEW STANDARDS AND POLICY (NSP)

TELECOMMUNICATIONS STANDARDS ADVISORY COMMITTEE TSAC WORKING GROUP ON NEW STANDARDS AND POLICY (NSP) TELECOMMUNICATIONS STANDARDS ADVISORY COMMITTEE TSAC WORKING GROUP ON NEW STANDARDS AND POLICY (NSP) Introduction Development of Digital Subscriber Line (DSL) Technology This paper introduces the development

More information

CS263: Wireless Communications and Sensor Networks

CS263: Wireless Communications and Sensor Networks CS263: Wireless Communications and Sensor Networks Matt Welsh Lecture 4: Medium Access Control October 5, 2004 2004 Matt Welsh Harvard University 1 Today's Lecture Medium Access Control Schemes: FDMA TDMA

More information

:-------------------------------------------------------Instructor---------------------

:-------------------------------------------------------Instructor--------------------- Yarmouk University Hijjawi Faculty for Engineering Technology Computer Engineering Department CPE-462 Digital Data Communications Final Exam: A Date: 20/05/09 Student Name :-------------------------------------------------------Instructor---------------------

More information

Network+ Guide to Networks 6 th Edition. Chapter 7 Wide Area Networks

Network+ Guide to Networks 6 th Edition. Chapter 7 Wide Area Networks Network+ Guide to Networks 6 th Edition Chapter 7 Wide Area Networks Objectives Identify a variety of uses for WANs Explain different WAN topologies, including their advantages and disadvantages Compare

More information

ADSL part 2, Cable Internet, Cellular

ADSL part 2, Cable Internet, Cellular ADSL part 2, Cable Internet, Cellular 20 June 2016 Lecture 12 20 June 2016 SE 428: Advanced Computer Networks 1 Topics for Today ADSL Cable Internet Cellular Radio Networks 20 June 2016 SE 428: Advanced

More information

Chapter 2 from Tanenbaum - modified. The Physical Layer. Ref: A.S. Tanenbaum, Computer Networks, 4 th Ed., Prentice-Hall, 2003, ISBN: 0-13-038488-7.

Chapter 2 from Tanenbaum - modified. The Physical Layer. Ref: A.S. Tanenbaum, Computer Networks, 4 th Ed., Prentice-Hall, 2003, ISBN: 0-13-038488-7. Chapter 2 from Tanenbaum - modified The Physical Layer Ref: A.S. Tanenbaum, Computer Networks, 4 th Ed., Prentice-Hall, 2003, ISBN: 0-13-038488-7. Data Communications over Wireless and Digital Wired Systems

More information

A Performance Study of Wireless Broadband Access (WiMAX)

A Performance Study of Wireless Broadband Access (WiMAX) A Performance Study of Wireless Broadband Access (WiMAX) Maan A. S. Al-Adwany Department of Computer & Information Engineering, College of Electronics Engineering University of Mosul Mosul, Iraq maanaladwany@yahoo.com

More information

Public Network. 1. Relatively long physical distance 2. Requiring a service provider (carrier) Branch Office. Home. Private Network.

Public Network. 1. Relatively long physical distance 2. Requiring a service provider (carrier) Branch Office. Home. Private Network. Introduction to LAN TDC 363 Week 4 Connecting LAN to WAN Book: Chapter 7 1 Outline Wide Area Network (WAN): definition WAN Topologies Choices of WAN technologies Dial-up ISDN T1 Frame Relay DSL Remote

More information

Narrowband and Broadband Access Technologies

Narrowband and Broadband Access Technologies Computer Networks and Internets, 5e Chapters 12 and 16 Access and Interconnection Technologies (slidesets abridged/combined) By Douglas Comer Modified from the lecture slides of Lami Kaya (LKaya@ieee.org)

More information

EFM Copper (EFMC) Tutorial. June 2004

EFM Copper (EFMC) Tutorial. June 2004 EFM Copper (EFMC) Tutorial June 2004 2004 Ethernet in the First Mile Alliance. All Rights Reserved. 1 EFM Copper (EFMC) Tutorial INTRODUCTION... 2 EFMC: BASED ON DSL... 3 BACKGROUND ON DSL... 3 DSL ENHANCEMENTS

More information

SECTION 2 TECHNICAL DESCRIPTION OF BPL SYSTEMS

SECTION 2 TECHNICAL DESCRIPTION OF BPL SYSTEMS SECTION 2 TECHNICAL DESCRIPTION OF SYSTEMS 2.1 INTRODUCTION Access equipment consists of injectors (also known as concentrators), repeaters, and extractors. injectors are tied to the backbone via fiber

More information

Nexus Technology Review -- Exhibit A

Nexus Technology Review -- Exhibit A Nexus Technology Review -- Exhibit A Background A. Types of DSL Lines DSL comes in many flavors: ADSL, ADSL2, ADSL2+, VDSL and VDSL2. Each DSL variant respectively operates up a higher frequency level.

More information

xdsl Technology and Applications:

xdsl Technology and Applications: xdsl Technology and Applications: Removing the Telephone Line Bottleneck Krista S. Jacobsen Texas Instruments Broadband Access Group jacobsen@ti.com Texas Instruments 1999 1 Overview/Goals Introduction

More information

INTERNET CONNECTIVITY

INTERNET CONNECTIVITY INTERNET CONNECTIVITY http://www.tutorialspoint.com/internet_technologies/internet_connectivity.htm Copyright tutorialspoint.com Here in this tutorial, we will discuss how to connect to internet i.e. internet

More information

Broadband Access Technologies

Broadband Access Technologies Broadband Access Technologies Chris Wong Communications Engineering Sector Analysis & Reporting Branch International Training Program 23 October 2007 Presentation Outline What is broadband? What are the

More information

HIGH CAPACITY DSL-SYSTEMS

HIGH CAPACITY DSL-SYSTEMS Electronic Communications Committee (ECC) within the European Conference of Postal and Telecommunications Administrations (CEPT) HIGH CAPACITY DSL-SYSTEMS Oulu, March 2006 Page 2 EXECUTIVE SUMMARY In this

More information

Technical Paper. Digital Subscriber Line (DSL): Using Next Generation Technologies to Expand Traditional Infrastructures

Technical Paper. Digital Subscriber Line (DSL): Using Next Generation Technologies to Expand Traditional Infrastructures Technical Paper Digital Subscriber Line (DSL): Using Next Generation Technologies to Expand Traditional Infrastructures USB ADSL Modem Contents Digital Subscriber Line (DSL): Using Next Generation Technologies

More information

Characterization of a new copper cable for next generation DSL systems

Characterization of a new copper cable for next generation DSL systems la The Tour Marine Tanguy, Drive,, Brest Mumbai. (FR). Characterization of a new copper cable for next generation DSL systems Project Co-ordinates ordinates Kalpesh PATEL TELECOM ParisTech Paris, France.

More information

White Paper: Broadband Access Technologies A White Paper by the Deployment & Operations Committee

White Paper: Broadband Access Technologies A White Paper by the Deployment & Operations Committee White Paper: Broadband Access Technologies A White Paper by the Deployment & Operations Committee Contributors: Rong Zhao (Detecon International) Wolfgang Fischer (Cisco) Edgar Aker (Prysmian Group) Pauline

More information

NEW WORLD TELECOMMUNICATIONS LIMITED. 2 nd Trial Test Report on 3.5GHz Broadband Wireless Access Technology

NEW WORLD TELECOMMUNICATIONS LIMITED. 2 nd Trial Test Report on 3.5GHz Broadband Wireless Access Technology NEW WORLD TELECOMMUNICATIONS LIMITED 2 nd Trial Test Report on 3.5GHz Broadband Wireless Access Technology Issue Number: 01 Issue Date: 20 April 2006 New World Telecommunications Ltd Page 1 of 9 Issue

More information

Long Distance Connection and WAN

Long Distance Connection and WAN Lecture 6 Long Distance Connection and WAN Digital Telephone, PCM and Nyquist Sampling Theorem DSU/CSU, T Line Series and OC line Series Local Loop DSL Technologies - ADSL, HDSL, SDSL, VDSL Cable Modem

More information

Guideline for the Implementation of Coexistence for Broadband Power Line Communication Standards

Guideline for the Implementation of Coexistence for Broadband Power Line Communication Standards NISTIR 7862 Guideline for the Implementation of Coexistence for Broadband Power Line Communication Standards Dr. David Su Dr. Stefano Galli http://dx.doi.org/10.6028/nist.ir.7862 1 NISTIR 7862 Guideline

More information

Chapter 9A. Network Definition. The Uses of a Network. Network Basics

Chapter 9A. Network Definition. The Uses of a Network. Network Basics Chapter 9A Network Basics 1 Network Definition Set of technologies that connects computers Allows communication and collaboration between users 2 The Uses of a Network Simultaneous access to data Data

More information

XDSL TECHNIQUES FOR POWER LINE COMMUNICATIONS

XDSL TECHNIQUES FOR POWER LINE COMMUNICATIONS XDSL TECHNIQUES FOR POWER LINE COMMUNICATIONS Costas Assimakopoulos*, P.L. Katsis*, F.-N. Pavlidou*, Danilo Obradovic**, Milorad Obradovic** Abstract *Aristotle University of Thessaloniki Dept. of Elec.

More information

Future Proof Telecommunications Networks with VDSL2

Future Proof Telecommunications Networks with VDSL2 Future Proof Telecommunications Networks with VDSL2 By: Stephan Wimoesterer Product Marketing Manager, VDSL2 Stephan. Wimoesterer@infineon.com V1.1 July, 2005 Future Proof Telecommunications Networks With

More information

How To Understand The Theory Of Time Division Duplexing

How To Understand The Theory Of Time Division Duplexing Multiple Access Techniques Dr. Francis LAU Dr. Francis CM Lau, Associate Professor, EIE, PolyU Content Introduction Frequency Division Multiple Access Time Division Multiple Access Code Division Multiple

More information

Objectives. Lecture 4. How do computers communicate? How do computers communicate? Local asynchronous communication. How do computers communicate?

Objectives. Lecture 4. How do computers communicate? How do computers communicate? Local asynchronous communication. How do computers communicate? Lecture 4 Continuation of transmission basics Chapter 3, pages 75-96 Dave Novak School of Business University of Vermont Objectives Line coding Modulation AM, FM, Phase Shift Multiplexing FDM, TDM, WDM

More information

TABLE OF CONTENTS. Dedication. Table of Contents. Preface. Overview of Wireless Networks. vii 1.1 1.2 1.3 1.4 1.5 1.6 1.7. xvii

TABLE OF CONTENTS. Dedication. Table of Contents. Preface. Overview of Wireless Networks. vii 1.1 1.2 1.3 1.4 1.5 1.6 1.7. xvii TABLE OF CONTENTS Dedication Table of Contents Preface v vii xvii Chapter 1 Overview of Wireless Networks 1.1 1.2 1.3 1.4 1.5 1.6 1.7 Signal Coverage Propagation Mechanisms 1.2.1 Multipath 1.2.2 Delay

More information

ADSL TUTORIAL. Figure 1: Typical DSL system.

ADSL TUTORIAL. Figure 1: Typical DSL system. ADSL TUTORIAL Matthew J. Langlois, University of New Hampshire InterOperability Laboratory 121 Technology Drive, Suite 2, Durham, NH 03824 USA. Extracted from the Introduction and Chapter 1 of A G.hs Handshaking

More information

Simulation Study on Internet Applications over DSL Access Network: KFUPM Campus as an Example

Simulation Study on Internet Applications over DSL Access Network: KFUPM Campus as an Example Simulation Study on Internet Applications over DSL Access Network: KFUPM Campus as an Example Uthman Baroudi, Saeed Bawazir and Wael Hamri Computer Engineering Department KFUPM, Dhahran Saudi Arabia Email:

More information

ADSL over ISDN, DAML, and Long Loops

ADSL over ISDN, DAML, and Long Loops Avi Vaidya Vice President and Chief Technology Officer Charles Industries, Ltd. over ISDN, DAML, and Long Loops As new technologies evolve, new challenges arise for telephone companies. Some of these challenges

More information

Network Requirements for DSL systems, (ADSL through G.Fast) (A summarized view)

Network Requirements for DSL systems, (ADSL through G.Fast) (A summarized view) Network Requirements for DSL systems, (ADSL through G.Fast) (A summarized view) Gilberto GG Guitarte, BB Connectivity Director TE Connectivity FTTH LATAM Chapter CHAIRMAN 2/24/2014 G.A.Guitarte 1 Executive

More information

Chapter 1: roadmap. Access networks and physical media

Chapter 1: roadmap. Access networks and physical media Chapter 1: roadmap 1.1 What is the nternet? 1.2 Network edge 1.3 Network core 1.4 Network access and physical media 1.5 nternet structure and SPs 1.6 elay & loss in packet-switched networks 1.7 Protocol

More information

Optimal Transmit Spectra for Communication on Digital Subscriber Lines

Optimal Transmit Spectra for Communication on Digital Subscriber Lines Optimal Transmit Spectra for Communication on Digital Subscriber Lines Rohit V. Gaikwad and Richard G. Baraniuk Department of Electrical and Computer Engineering Rice University Houston, Texas, 77005 USA

More information

Cable 101. A Broadband Telecommunications Primer for Non-technical Personnel

Cable 101. A Broadband Telecommunications Primer for Non-technical Personnel Cable 101 KnowledgeLink, Inc. A Broadband Telecommunications Primer for Non-technical Personnel Presented by: Justin J. Junkus President, KnowledgeLink, Inc. November 20, 2013 Agenda Broadband Cable Systems

More information

Cable Modems. Definition. Overview. Topics. 1. How Cable Modems Work

Cable Modems. Definition. Overview. Topics. 1. How Cable Modems Work Cable Modems Definition Cable modems are devices that allow high-speed access to the Internet via a cable television network. While similar in some respects to a traditional analog modem, a cable modem

More information

Multiplexing on Wireline Telephone Systems

Multiplexing on Wireline Telephone Systems Multiplexing on Wireline Telephone Systems Isha Batra, Divya Raheja Information Technology, Dronacharya College of Engineering Farrukh Nagar, Gurgaon, India ABSTRACT- This Paper Outlines a research multiplexing

More information

ADSL2 AND ADSL2plus THE NEW ADSL STANDARDS

ADSL2 AND ADSL2plus THE NEW ADSL STANDARDS ADSL2 AND ADSL2plus THE NEW ADSL STANDARDS March 25, 2003 ADSL2 AND ADSL2plus - THE NEW ADSL STANDARDS In July 2002, the ITU completed G.992.3 and G.992.4 1, two new standards for ADSL technology collectively

More information

Data Transmission via Modem. The Last Mile Problem. Modulation of Digital Signals. Modem Standards (CCITT)

Data Transmission via Modem. The Last Mile Problem. Modulation of Digital Signals. Modem Standards (CCITT) The Last Mile Problem LN, MN, WN how to connect private users at home to such networks? Problem of the last mile: somehow connect private homes to the public Internet without laying many new cables By

More information

V-DSL technology in the NBN

V-DSL technology in the NBN CommsDay Melbourne Congress 2013 V-DSL technology in the NBN Wednesday 9 th October 2013 Cliff Gibson Outline VDSL2 technologies which extend the capability of existing copper are expected to be included

More information

CSCI 491-01 Topics: Internet Programming Fall 2008

CSCI 491-01 Topics: Internet Programming Fall 2008 CSCI 491-01 Topics: Internet Programming Fall 2008 Introduction Derek Leonard Hendrix College September 3, 2008 Original slides copyright 1996-2007 J.F Kurose and K.W. Ross 1 Chapter 1: Introduction Our

More information

Channel Bandwidth, MHz. Symbol Rate, Msym/sec

Channel Bandwidth, MHz. Symbol Rate, Msym/sec APPENDIXB Information in the following tables is from the DOCSIS and EuroDOCSIS Radio Frequency Interface Specification, and should be considered minimum recommended performance criteria for reliable data

More information

ZHONE VDSL2 TECHNOLOGY. Access Technology for the Future. November 2009 CONTENTS

ZHONE VDSL2 TECHNOLOGY. Access Technology for the Future. November 2009 CONTENTS ZHONE VDSL2 TECHNOLOGY Access Technology for the Future November 2009 CONTENTS Introduction 2 VDSL2 technology 2 VDSL2 Deployments: Zhone findings 4 Zhone s VDSL2 solutions 6 Conclusion 8 Access for a

More information

What Does Communication (or Telecommunication) Mean?

What Does Communication (or Telecommunication) Mean? What Does Communication (or Telecommunication) Mean? The term communication (or telecommunication) means the transfer of some form of information from one place (known as the source of information) to

More information

Copyright. Transport networks. Physical layer Transport and access networks. Pag. 1

Copyright. Transport networks. Physical layer Transport and access networks. Pag. 1 Physical layer Transport and access networks Gruppo Reti TLC nome.cognome@polito.it http://www.telematica.polito.it/ COMPUTER NETWORK DESIGN Physical layer review - 1 Copyright Quest opera è protetta dalla

More information

The Effect of Network Cabling on Bit Error Rate Performance. By Paul Kish NORDX/CDT

The Effect of Network Cabling on Bit Error Rate Performance. By Paul Kish NORDX/CDT The Effect of Network Cabling on Bit Error Rate Performance By Paul Kish NORDX/CDT Table of Contents Introduction... 2 Probability of Causing Errors... 3 Noise Sources Contributing to Errors... 4 Bit Error

More information

COMMITTEE T1 TELECOMMUNICATIONS Working Group T1E1.4 (DSL Access) Ottawa, Canada; June 7, 1999

COMMITTEE T1 TELECOMMUNICATIONS Working Group T1E1.4 (DSL Access) Ottawa, Canada; June 7, 1999 COMMITTEE T1 TELECOMMUNICATIONS Working Group T1E1.4 (DSL Access) Ottawa, Canada; June 7, 1999 T1E1.4/99-261 CONTRIBUTION TITLE: SOURCE: PROJECT: Regarding the Spectral Compatibility of 2B1Q SDSL Telcordia

More information

Mixed High-Speed Ethernet Operations over Different Categories of Bundled UTP Cable

Mixed High-Speed Ethernet Operations over Different Categories of Bundled UTP Cable Mixed High-Speed Ethernet Operations over Different Categories of Bundled UTP Cable June 10, 2010 Contributors: Yinglin (Frank) Yang CommScope, Inc. Charles Seifert Ixia Ethernet Alliance 3855 SW 153 rd

More information

DSL: An Overview. By M. V. Ramana Murthy. All Rights Reserved

DSL: An Overview. By M. V. Ramana Murthy. All Rights Reserved By M. V. Ramana Murthy All Rights Reserved COPYRIGHT: This document is a property of MultiTech Software Systems India Pvt. Ltd. No part of this document may be copied or reproduced in any form or by any

More information

ADSL or Asymmetric Digital Subscriber Line. Backbone. Bandwidth. Bit. Bits Per Second or bps

ADSL or Asymmetric Digital Subscriber Line. Backbone. Bandwidth. Bit. Bits Per Second or bps ADSL or Asymmetric Digital Subscriber Line Backbone Bandwidth Bit Commonly called DSL. Technology and equipment that allow high-speed communication across standard copper telephone wires. This can include

More information

T = 1 f. Phase. Measure of relative position in time within a single period of a signal For a periodic signal f(t), phase is fractional part t p

T = 1 f. Phase. Measure of relative position in time within a single period of a signal For a periodic signal f(t), phase is fractional part t p Data Transmission Concepts and terminology Transmission terminology Transmission from transmitter to receiver goes over some transmission medium using electromagnetic waves Guided media. Waves are guided

More information

Orion2+ SHDSL.bis Solution with 11Mbit/s and 15Mbit/s per Copper Pair

Orion2+ SHDSL.bis Solution with 11Mbit/s and 15Mbit/s per Copper Pair Orion2+ SHDSL.bis Solution with 11Mbit/s and 15Mbit/s per Copper Pair Good to Know about Orion s Orion1 and Orion2 Modems from FlexDSL have always some more features than corresponding products from competitors.

More information

Managing Interference in a Vectored VDSL2 environment Communications Alliance WC58

Managing Interference in a Vectored VDSL2 environment Communications Alliance WC58 Managing Interference in a Vectored VDSL2 environment Communications Alliance WC58 18 November 2014 Stefan Keller-Tuberg, Acting Editor Communications Alliance Working Committee 58 1 CA Working Committee

More information

An Analysis of Speed Drop in ADSL Lines in Sri Lanka

An Analysis of Speed Drop in ADSL Lines in Sri Lanka An Analysis of Speed Drop in ADSL Lines in Sri Lanka CPSY Dalpathadu # and PN Karunanayaka General Sir John Kotelawala Defence University, Ratmalana, Sri Lanka # sach.yamani@gmail.com Abstract: This paper

More information

INTERNET DELIVERY ON CATV - A BASIC OVERVIEW

INTERNET DELIVERY ON CATV - A BASIC OVERVIEW INTERNET DELIVERY ON CATV - A BASIC OVERVIEW THE BASICS A basic question that is often not answered is " What really IS the Internet? " The Internet is simply a network of computers that are linked together.

More information

Cabling LANs and WANs

Cabling LANs and WANs Cabling LANs and WANs CCNA 1 v3 Module 5 10/11/2005 NESCOT CATC 1 Cabling the LAN Each media has advantages and disadvantages: Cable length Cost Ease of installation Susceptibility to interference The

More information

Innovative Synergies

Innovative Synergies 20030323 Digital CAN Innovative Synergies (2003, 2005, 2007, August 2012) Malcolm Moore 32-Mar-2003 Positioning ADSL (Asymmetric Digital Subscribers Line) In the picture below of the stylised network infrastructures

More information

DSL Variations. NEXTEP Broadband White Paper. Broadband Networks Group. Definitions and differences of Digital Subscriber Line variations.

DSL Variations. NEXTEP Broadband White Paper. Broadband Networks Group. Definitions and differences of Digital Subscriber Line variations. NEXTEP Broadband White Paper DSL Variations Definitions and differences of Digital Subscriber Line variations. A NEXTEP Broadband White Paper May 2001 Broadband Networks Group DSL Variations EXECUTIVE

More information

Chapter 4 Connecting to the Internet through an ISP

Chapter 4 Connecting to the Internet through an ISP Chapter 4 Connecting to the Internet through an ISP 1. According to Cisco what two things are essential to gaining access to the internet? a. ISPs are essential to gaining access to the Internet. b. No

More information

Lecture 1. Introduction to Wireless Communications 1

Lecture 1. Introduction to Wireless Communications 1 896960 Introduction to Algorithmic Wireless Communications Lecture 1. Introduction to Wireless Communications 1 David Amzallag 2 May 25, 2008 Introduction to cellular telephone systems. How a cellular

More information

Wireless SDSL for the Business Sector

Wireless SDSL for the Business Sector Wireless SDSL for the Business Sector Broadband Services over BreezeACCESS VL June 2005 Alvarion Ltd. All rights reserved. The material contained herein is proprietary. No part of this publication may

More information

Line Simulator (LiSi) for Asymmetric and Very High-Speed Digital Subscriber Line

Line Simulator (LiSi) for Asymmetric and Very High-Speed Digital Subscriber Line Line Simulator (LiSi) for Asymmetric and Very High-Speed Digital Subscriber Line Nikolaos Choulakis, Research Masters in Telecommunications, University College London Abstract: A brief introduction describing

More information

Proposal: Option for in-band POTS and ISDN. Mikael Isaksson, Tomas Stefansson, Per Ödling, Frank Sjöberg, Kate Wilson

Proposal: Option for in-band POTS and ISDN. Mikael Isaksson, Tomas Stefansson, Per Ödling, Frank Sjöberg, Kate Wilson 1(5) Standards project: T1E1.4: VDSL Title : Proposal: Option for in-band POTS and ISDN Source : Contact: Telia Research AB Mikael Isaksson, Tomas Stefansson, Per Ödling, Frank Sjöberg, Kate Wilson Mikael

More information

Network Design. Yiannos Mylonas

Network Design. Yiannos Mylonas Network Design Yiannos Mylonas Physical Topologies There are two parts to the topology definition: the physical topology, which is the actual layout of the wire (media), and the logical topology, which

More information

VDSL: The Next Step in the DSL Progression

VDSL: The Next Step in the DSL Progression VDSL: The Next Step in the DSL Progression Krista S. Jacobsen (jacobsen@ti.com) Broadband Access Group San Jose, CA DSPS Fest August 1999 Houston, TX 1 July 16, 1999 2 Table of Contents 1 General overview...5

More information