End-to-End Network Architecture for Cognitive Reconfigurable Mobile Systems

Transcription

1 E 2 R II White Paper End-to-End Network Architecture for Cognitive Reconfigurable Mobile Systems ABSTRACT Reconfigurability is set to be an important facet in the evolving world of mobile and wireless communications, through which technologies such as cognitive radio are greatly facilitated. Given advances in the scope and applicability of software functionalities, reconfigurability is increasingly becoming possible at many layers of the protocol stack which would until recently have been maintained only by intransigent fixed hardware components. Moreover, given the complexity and range of emerging reconfiguration possibilities, it would be useful for devices and networks to be able to efficiently partake in reconfiguration procedures with reduced - or even zero - human input. This would maximise the realisation of potential performance improvements through reconfiguration-associated technologies. This white paper discusses the development of an end-to-end network architecture that incorporates cognition and autonomics aspects into reconfiguration procedures. The architecture, its overall context, and its mapping to the emerging 3GPP System Architecture Evolution (SAE) as well as to an All-IP network configuration, are discussed. 1 INTRODUCTION End-to-End Reconfigurability (E²R II) project [1] aims to devise, design, and validate architectural solutions and support algorithms, protocols, and mechanisms in cognitive networks, wherein the notion of cognitive networks accommodates software and cognitive radio facilities enriched by self-management capabilities. Cognitive radio [2-4] has recently been the subject of intensive research, targeted at providing a breakthrough spectrum-hopping capability to utilise the white spaces observed in spectrum occupancy measurements. Software-defined radio [5-6] is one enabler of cognitive radio, offering software facilities for tuning and reconfiguring RF and baseband processing functions and parameters such as operating frequency, power, and modulation type. Cognitive networks bear the same planning, decision-making, and enforcement stages to optimise the overall performance taking into account end-to-end goals [7]. A cognition engine, which senses the environment, learns from past observations, and plans the future actions, is the central part of such a framework. A similar feedback loop is applied in autonomic computing targeting self-management capabilities [8]. Whereas [7] and [8] focus on networking and computing disciplines, E 2 R II project has proposed a three-tier control and management architecture for end-to-end reconfigurable End-to-End Reconfigurability II (E 2 R II) White Paper December 07 1/12

2 systems [9], within the scope of next-generation All-IP mobile networks. This proposed model accommodates solutions for software and cognitive radios, focusing on equipment upgrades through over-the-air software download, and dynamic spectrum allocation and access. Such architecture aims at minimum impact on legacy or future infrastructures, through standalone servers for implementing the functional capabilities. Besides, E 2 R II has introduced a new logical model for applying the autonomic computing paradigm in reconfigurable wireless networks [10], in the form of UML modelling and functional specification. This white paper presents extensions to previous work [9-10] within the E 2 R II project in the following ways [11]. Firstly, an overall functional model for autonomously reconfigurable user equipment and network elements is proposed, integrating functional entities for software upgrades and spectrum/radio resource efficiency, also highlighting the identified interfaces. More importantly, the paper enhances the ongoing 3GPP Figure 1: E 2 R II vision of heterogeneous end-to-endreconfigurable systems Long-Term Evolution / System Architecture Evolution (LTE/SAE) work [12] via new functionalities in the software/cognitive radio and autonomic communication areas. In addition, an All-IP network configuration is presented. The paper is organised as follows: the E 2 R II functional model for cognitive reconfigurable wireless networks is briefly described in the following section, with section 3 giving an overview of the 3GPP LTE/SAE architecture. The enhancement of 3GPP SAE elements and architectural extensions are presented in Section 4, and section 5 introduces the E 2 R II All- IP architecture. Conclusions and next steps are highlighted in Section 6. 2 OVERVIEW OF E 2 R II SYSTEM ARCHITECTURE Fig. 2 shows the E 2 R II system architecture for cognitive reconfigurable wireless networks [11]. The architecture is based on the concept of a Reconfiguration Plane (RMP) [10], which comprises a modular, intermediary, multi-plane- and layer-based solution allowing independent evolution paths and refinements of the legacy control and management planes. The depicted functional entities are grouped under the following categories: Control and management functions for discovery of the reconfiguration capabilities/services offered by an E 2 R-II-enabled infrastructure (RSD function), context and resource monitoring (RsM and functions), as well as for processing and transforming such contextual/profile information ( function) and generating performance reports ( function); An overarching entity for negotiation and session control that produces definitive reconfiguration decisions (ADM&RM function); End-to-End Reconfigurability II (E 2 R II) White Paper December 07 2/12

3 Control functions that cater for software download (SDM function), apply content control and service adaptation due to reconfiguration events ( function), and execute handover decisions (MM function); Control and management functions that offer autonomic features, specifically, selfconfiguration of protocols (SeCo function), and self-healing and self-governance capabilities (CoSH and PMSG functions, respectively); Control and management functions for spectrum and radio resource efficiency; specifically, the JRRM function for jointly optimising the radio resources of a set of Radio Access Technologies (RATs), the DNPM function for network re-planning, the SAM, STM, and SEM functions for spectrum allocation, trading, and related economic functionalities (e.g., spectrum auctioning, pricing, and billing), and the TE function for spectrum demand prediction. Details on these functions can be found in [13]; Other auxiliary functions for charging, access and security control purposes (RCC and PASeC functions, respectively). A detailed specification of the functional entities can be found in [10] as well as in a dedicated E 2 R II system architecture white paper [14]. Control functions RSD PMSG SDM SeCo ADM & RM functions CoSH PASeC RCC RsM TE MM JRRM DNPM SAM SEM STM Legend ADM&RM Autonomic Decision-Making & Reconfiguration CoSH Cognitive Self-Healing Cognitive Pilot Channel Controller Content and Service Adaptation DNPM Dynamic Network Planning and JRRM Joint Radio Resource Manager MM Mobility PASeC Pervasive Access and Security Control Performance PMSG Policy and Self-Governance Profile RCC Reconfiguration Charging Control RSD Reconfiguration Services Discovery RsM Resource SDM Software Download SAM Spectrum Allocation Manager SeCo Self-Configuration SEM Spectrum Economic Manager STM Spectrum Trading Manager TE Traffic Estimator Figure 2: E 2 R II system architecture for cognitive reconfigurable wireless networks End-to-End Reconfigurability II (E 2 R II) White Paper December 07 3/12

4 3 OVERVIEW OF 3GPP SYSTEM ARCHITECTURE EVOLUTION Fig. 3 provides an overview of the evolved 3GPP system along with baseline Release 7 domains [12]. The evolved 3GPP system consists of the User Equipment (UE) connected to one or more Evolved Radio Access Networks (E-RANs), and E-RANs connected to one ore more Evolved Packet Core (EPC) networks, with EPCs providing connectivity to external Packet Data Networks (PDNs). The evolved packet core interworks with a) legacy 3GPP access systems (2G/3G) through the legacy GPRS core network, b) trusted non-3gpp IP access networks, and c) Interworking WLAN (I-WLAN) access networks through the evolved Packet Data Gateway (epdg). The evolved 3GPP system also consists of the Home Subscriber Server (HSS) and 3GPP service-stratum systems/functions such as the Policy and Charging Rule Function (PCRF), the Application Function (AF), the Broadcast/Multicast Service Centre (BM-SC), and the IP Multimedia Subsystem (IMS). The evolved RAN consists of a number of evolved Node-Bs (enb). The evolved Node B manages radio resources (e.g., packet scheduling in the downlink) and performs policing of uplink user-plane traffic. The EPC consists of the Mobility Entity (MME), the User Plane Entity (UPE), and the Inter-Access System Anchor (IASA). The MME and UPE are collectively referred to as Access Gateway (AGW). The IASA is in charge of mobility between different access systems. It is composed of a 3GPP anchor that executes mobility between 3GPP access systems and of a SAE anchor, which handles mobility between 3GPP and non-3gpp access systems. The SAE anchor does not make any decisions regarding mobility; it just executes such decisions made by the UE. It should be noted that the latest 3GPP technical reports use the terms serving and PDN Gateways (S-GW and P-GW, respectively) to refer to core network elements [15-16]. These roughly correspond to the UPE and IASA elements depicted in fig. 3. SAE targets the long-term evolution of the 3GPP access technology for ensuring competitiveness in the next ten years. The focus will be on the support of IP-based 3GPP services through various access technologies offering seamless mobility between heterogeneous access networks. Nevertheless, SAE does not capture software and cognitive radio aspects, and does not accommodate facilities for self-configuration and selfmanagement. The E 2 R-II-enhanced 3GPP SAE network architecture attempts to cover this gap. 4 MAPPING TO 3GPP LTE/SAE FOR COGNITIVE RECONFIGURABLE USER EQUIPMENT This section provides the mapping of the E 2 R II system architecture to the 3GPP LTE/SAE network for cognitive reconfigurable UEs. The objective is to yield a network architecture that facilitates optimal split of reconfiguration intelligence and functionalities between cognitive network elements and reconfigurable end-user equipment. End-to-End Reconfigurability II (E 2 R II) White Paper December 07 4/12

5 GERAN Gb UE UTRAN E-RAN Iu S1 S1 GPRS Core S6 S3 S4 MME S5 UPE IASA EPC S2b S2a epdg HSS S7 SGi Rx+(Rx/Gq) PCRF AF SGi BM-SC IMS SGi Trusted non-3gpp IP Access Network WLAN Access Network PDN Legend AF BM-SC epdg EPC E-RAN GERAN GPRS HSS IASA Application Function Broadcast/Multicast Service Centre Evolved Packet Data Gateway Evolved Packet Core Evolved Radio Access Network GSM Enhanced Radio Access Network General Packet Radio Service Home Subscriber Server Inter-Access System Anchor IMS MME PCRF PDN UE UPE UTRAN WLAN IP Multimedia Subsystem Mobility Entity Policy and Charging Rule Function Packet Data Network User Equipment User Plane Entity UMTS Terrestrial Radio Access Network Wireless Local Area Network Figure 3: The Evolved 3GPP System and interworking with the baseline Release 7 architecture The goal of this mapping assessment is to identify the potential evolution and impact on the 3GPP LTE/SAE, augmenting legacy functionalities with reconfiguration-induced procedures. The main criteria for the mapping are: a) what are the elements that are responsible for managing the additional functionality. b) What is the entity that should hold a primary or secondary copy of the introduced information elements, with the former having access rights to modify this specific set of parameters, whereas the latter having read-only access rights. In this context, three different roles are used in the mapping to each network element: Main storage: This entity initiates the storage process as soon as there is a change in the entity. Copy: This is a copy of the main storage. The entity does not change the stored information; instead, it only gets a copy from the main storage when there is a change. The entity executes the store process initiated by the main storage. : This entity has the right to manage the concerned function. The term management here denotes both control plane (online/real-time tasks related to traffic exchange of specific user sessions/calls) and management plane functionality (offline tasks). The mapping analysis below focuses on key functional entities that address aspects related to terminal reconfiguration and autonomics. Table 1 shows the distribution of ADM&RM, SeCo/CoSH, and functionalities in the LTE/SAE elements. AN/HN refer to the access/home network, whereas the network backend includes other network elements End-to-End Reconfigurability II (E 2 R II) White Paper December 07 5/12

6 besides the elements already listed in the operator network (e.g., the Operations and Maintenance (O&M) server, the device management server, etc.) Finally, the external server comprises an application/service server. 4.1 Autonomic Decision-Making & Reconfiguration Function The ADM&RM function produces up-to-date decisions for reconfiguration and optimization, based on contextual information and evaluation of policy rules. In LTE/SAE, reconfiguration has a smaller scope than in E 2 R II: it concerns the application characteristics and the bearers allocated. It is also possible to configure via O&M some network elements. Thus, reconfiguration does not concern all network elements and is not based on autonomous and real-time decisions. In LTE/SAE architecture, decisions concern aspects related to inter-access system mobility and are performed in the IASA. In LTE/SAE, network elements can be reconfigured partly (e.g., only parameters setting) and it is possible that this require the re-setting of the machine. Decisions regarding advanced reconfiguration services proposed in this paper (e.g., software upgrades, dynamic spectrum access) are implemented in the ADM function as shown in Table 1a. Fig. 4 maps all system architecture functions to LTE/SAE elements. It is worth mentioning that the ADM function is depicted only in the UE (not in any network elements), as this figure addresses the case when only UEs can be autonomously reconfigurable. UE AN AGW IASA HN I-WLAN IMS Network Backend External Server S M S M S M S M S M S M S M S M S M (a) Autonomic decision-making and reconfiguration management UE AN AGW IASA HN I-WLAN IMS Network Backend External Server S M S M S M S M S M S M S M S M S M (b) Self-configuration and self-healing UE AN AGW IASA HN I-WLAN IMS Network Backend External Server S M S M S M S M S M S M S M S M S M (c) Storage and management of network profile Table 1: Distribution of E 2 R II functionalities in 3GPP LTE/SAE elements: (a) ADM&RM function, (b) SeCo and CoSH functions, (c) function (network profile) End-to-End Reconfigurability II (E 2 R II) White Paper December 07 6/12

7 Reconfigurable UE ADM&RM Evolved Packet Core MME PCRF PMSG RCC RSD PMSG CoSH SDM PASeC SeCo RCC RsM L-MM LSAM Evolved RAN enb Controller LSEM LSAM JRRM L-MM JRRM UPE IASA G-MM ARM RM SDM RSD PASeC DNPM TE STM DNPM IOSEM JRRM L-MM G-MM GSAM RsM RCC BM-SC IMS Legend ADM&RM CoSH DNPM G-MM IOSEM JRRM L-MM LSAM LSEM Autonomic Decision-Making & Reconfiguration Cognitive Self-Healing Cognitive Pilot Channel Controller Content and Service Adaptation Dynamic Network Planning and Global Mobility Inter-Operator Spectrum Economic Manager Joint Radio Resource Manager Local Mobility Local Spectrum Allocation Manager Local Spectrum Economic Manager PASeC Pervasive Access and Security Control Performance PMSG Policy and Self-Governance Profile RCC Reconfiguration Charging Control RSD Reconfiguration Services Discovery RsM Resource SDM Software Download SeCo Self-Configuration STM Spectrum Trading Manager TE Traffic Estimator RM PMSG RSD SDM Figure 4: E 2 R-II-enhanced 3GPP SAE network architecture for cognitive reconfigurable wireless networks 4.2 Self-Configuration and Cognitive Self-Healing Since in LTE/SAE reconfiguration concerns the application and the network bearers, these functions should be managed and stored not only by the device but also by the network. Although the second case may violate the self-ware principle, SeCo and CoSH UE-peer functionalities may be needed in the network. Other service and application providers might need diagnosis for the service/application, so the external server should manage the processes as well. As this function optimizes network resources, the network backend is also involved. 4.3 Storage and of Network Profile Network profile contains information about the currently used RAT and a given RAT among those that could be used by the terminal. Therefore, it should be managed by the access network. The network operator may want to change certain network parameters, so it should store and be able to manage this profile. Since new access networks may be added, network backend should control this profile. However, some network connections may be created by the user (e.g., the WLAN UE), so the device may have the control in some scenarios as well. Since IASA manages the inter-access mobility and makes related decisions, it should contain information about the possible RATs for a given user to make the handover decision. 4.4 Cognitive Pilot Channel Controller The evolved Node B is responsible for retrieving and distributing the Cognitive Pilot Channel (CPC) information from/to other functional entities. The CPC [17] is envisaged as a radio enabler for reconfiguration by conveying a directory of the <operators, RATs, frequencies> offered in the concerning areas through a specific RAT and mechanism (socalled outband CPC ). The CPC may convey additional context/resource/signalling End-to-End Reconfigurability II (E 2 R II) White Paper December 07 7/12

8 information in both uplink and downlink using existing RATs (referred to as inband CPC ). Section 4.6 provides a CPC use case. After a base station reconfiguration, e.g., when new RAT is added or spectrum is allocated, all associated terminals in the coverage have to be informed about the new configuration. Otherwise, it will be difficult for inactive terminal to get this information without a tedious scanning process. With CPC as the cognitive radio enabler, the changed configuration information can be easily distributed to terminals. Among different CPC concepts, the architectural issues for hierarchical CPC are discussed below. In the hierarchical CPC concept [18], there are three levels of pilot signal. Level 1 CPC (country-level CPC) indicates the operators that are in the coverage area and the frequencies and RATs that are available for level 2 CPC (operator-level CPC). This information will not be frequently changed. Level 2 CPC informs which RAT at specific frequency is operated by the operator; this information can be frequently adapted according to base station reconfiguration. Level 3 CPC indicates the specific RAT information. To implement this hierarchical structure, it is helpful to apply hierarchical network architecture with a form of meta-operator on the top of the hierarchy. As shown in fig. 4, the country-level CPC is implemented by the functional entity at the anchor reconfiguration manager (ARM). This assumes that country-level CPC is maintained by a meta-operator that implements regulator s rules and policies and is independent of individual operators. 4.5 Broadcast/Multicast Service Centre The BM-SC SDM function orchestrates the transfer via multicast/broadcast of the reconfiguration download, where the BM-SC might provide the end-point for the reconfiguration content as beforehand provided by an external server. This implies selection of a transport means among other characteristics for the download. An extended functionality for the SDM, in a heterogeneous networking scenario where the download is conveyed directly from an external source via MBMS, might be participation in procedures for the selection and dynamic adaptation of the network layer mode (e.g., manyunicast/multicast) for the download, and coordination of resulting implications for the higher transport layer design. The SDM might also partake in transport layer characteristics selection for reliable multicast per se. In the case where a download is provided by another entity for relaying by the BM-SC, the SDM might also validate the originator of the download and the integrity of content (both as regards the possibilities of corruption and malicious interception). The BM-SC function executes adaptation of other multimedia services due to reconfiguration events. Finally, the RSD function broadcasts or multicasts information on the reconfiguration services provided by the PLMN. This necessitates subscription of the UE to the MBMS service. This solution should be juxtaposed with the option of placing the RSD function in the ARM. End-to-End Reconfigurability II (E 2 R II) White Paper December 07 8/12

9 4.6 Spectrum Resources Efficiency and CPC Use Fig. 4 depicts the distribution of functional entities for spectrum and radio resource efficiency to the evolved Node B and ARM. Whereas placing the JRRM, LSAM, LSEM,, and DNPM in the enb could be considered rather straightforward, there are options for the mapping of the GSAM, either in the ARM (as in fig. 4) or in evolved packet core entities. Besides, whereas the Inter-Operator SEM (IOSEM) and the inter-operator STM fit perfectly into the ARM element, the TE function could reside in either the ARM (as in fig. 4), or the IASA, or even the UPE, depending on operator implementation options for traffic monitoring. Besides, resides in the ARM to implement the functional realisation of the country-level and operator-level CPC. JRRM is also depicted in the ARM in order to coordinate the decision-making for a resource pool of multiple access systems, thus playing a role similar to the common RRM server in [19]. Finally, it is worth noting the distinction between intra-operator and inter-operator JRRM. Whereas intra-operator JRRM resides in enb (as depicted in fig. 4), the interoperator JRRM function may reside in network elements such as the ARM. Furthermore, JRRM could also reside in the MME (as depicted in fig. 4) if S101 interface is used for optimised handovers with Mobile WiMAX or CDMA2000 HRPD [16][20]. Fig. 5 presents an example message flow showing the role of the cognitive pilot channel. Resulting from current load situation at the UPE RsM (reported from the UPE ) and predicted traffic at the ARM TE, the ARM RM function asks the enb LSAM/LSEM for more spectral resources for a specific time period with affordable price. The enb LSAM/LSEM may communicate with the ARM STM/IOSEM to cater for the corresponding inter-operator spectrum trading. After receiving the corresponding trading result, the enb LSAM performs spectrum rearrangement taking into account all available information including network context and policy information. UE ADM&RM UE enb enb LSAM/LSEM UPE ARM TE ARM STM/IOSEM ARM RM Load report Spectrum Resource Query Traffic prediction Spectrum Resource Query Forward Monitor CPC Spectrum Resource Query Forward Response Spectrum reallocation Spectrum Resource Query Response Update directory Compile new directory Transmit new directory Inform new spectrum allocation Decide on spectrum hopping or HO Figure 5: CPC use case End-to-End Reconfigurability II (E 2 R II) White Paper December 07 9/12

10 Next, the ARM RM function informs the enb about the reallocated spectrum. The latter combines/filters the updated information and delivers it to the CPC channel timely. The UE monitors the CPC and retrieves the updated information; based on local policy rules and UE mobility context, the UE ADM&RM makes an autonomous decision so as to adapt to the changed network conditions (e.g., performs a spectrum hopping or handover to maintain the desired QoS for an ongoing session). 5 MAPPING TO ALL-IP NETWORKS Fig. 6 shows the mapping of all E2R II system architecture functional entities to a future All-IP architecture, satisfying the functional requirements for future All-IP networks identified by ITU [21] and 3GPP [22]. The figure depicts the case of reconfiguration of both the UE and the base station. The composite RAN manager includes functionality similar to the MME/UPE in the 3GPP SAE case, whereas the anchor reconfiguration manager accommodates the functional entities similar to the 3GPP-enhanced ARM, as well as the PCRF and IMS entities. Due to the flat core network and the existence of the JRRM function in the composite RAN manager, there is no need to place the JRRM function in the ARM as well. Tier 1 Reconfiguration Control & Tier 2 Reconfiguration Control & Tier 3 Reconfiguration Control & Multi-Standard Base Station Reconfigurable UE Control RSD PMSG SDM SeCo RsM ADM&RM CoSH PASeC RCC LSAM Control PMSG SDM SeCo RsM ADM&RM CoSH PASeC RCC LSAM JRRM Composite RAN Manager Control CoSH LSEM DNPM Anchor Reconfiguration Manager Control RM RSD PMSG SDM PASeC DNPM IOSEM GSAM RCC TE STM MM JRRM IMS AP DVB APC MSBS Post 3G PDG/WAG AP WiMAX APC HSS CSCF MRFC MRFP Composite RAN AP Wi-Fi Node B 3G RNC Post-3G CN (Native IP) Figure 6: Mapping of E 2 R II system architecture to All-IP network configuration 6 CONCLUSION This white paper has presented the end-to-end network architecture for cognitive reconfigurable mobile systems developed in the context of the EU-funded E 2 R II research project. Firstly, the paper provided an overview of the overall E 2 R II functional model and highlighted the key architectural elements of the evolved 3GPP system, which is currently specified within the 3GPP system-architecture evolution work. Next, the paper presented how the E 2 R II system architecture was mapped to the 3GPP SAE configuration: the mapping criteria were analysed and key aspects related to the placement of autonomic End-to-End Reconfigurability II (E 2 R II) White Paper December 07 10/12

11 decision-making and reconfiguration management, self-configuration and cognitive selfhealing, network profiling, cognitive pilot channel, and broadcast/multicast functionalities to SAE-enhanced or new network elements were discussed. An exemplary signalling flow for spectrum resource efficiency and CPC use was also presented. Finally, the paper proposed a mapping to an All-IP network configuration based on the SAE case. The E 2 R II system and network architecture embody innovative software and cognitive radio features whilst introducing key autonomic capabilities. The mapping of the architecture to LTE/SAE has been presented recently in 3GPP [23]. E 2 R II Consortium carefully considered the potential next steps and industrial standardization interests, and assessed the path towards 3GPP as well as related bodies such as ETSI and IEEE. AUTHORS Zachos Boufidis, University of Athens, Greece, boufidis@di.uoa.gr Nancy Alonistioti, University of Athens, Greece, nancy@di.uoa.gr Rémi Feuillette, LG Electronics, France, rfeuillette@lge.com Oliver Holland, King s College London, UK, oliver.holland@kcl.ac.uk Jianming Pan, Siemens AG Oesterreich, Austria, jianming.pan@siemens.com Klaus Moessner, University of Surrey, UK, k.moessner@surrey.ac.uk REFERENCES [1] IST End-to-End Reconfigurability (E 2 R II) Project, [2] J. Mitola, Cognitive Radio: An Integrated Agent Architecture for Software Defined Radio, Ph.D. thesis, Royal Inst. Technology, 2000 [3] J. Mitola and G. Maguire, Cognitive Radio: Making Software Radios more Personal, IEEE Pers. Commun., vol. 6, no. 4, pp , Aug [4] S. Haykin, Cognitive radio: Brain-empowered wireless communications, IEEE Journal on Selected Areas in Communications, vol. 23, no. 2, Feb [5] The Software Defined Radio (SDR) Forum, [6] M. Dillinger, K. Madami, and N. Alonistioti, Eds., Software Defined Radio: Architectures, Systems and, Wiley, 2003 [7] R. Thomas, D. Friend, L. DaSilva, and A. MacKenzie, Cognitive networks: Adaptation and learning to achieve end-to-end performance objectives, IEEE Comm. Mag., vol. 44, no. 12, pp , Dec [8] J. Kephart and D. Chess, The vision of autonomic computing, IEEE Computer, vol. 36, no.1, pp , Jan [9] Z. Boufidis et al., End-to-End Architecture for Adaptive Communication Systems, Proc. 64 th IEEE Vehicular Technology Conference (VTC), Montreal, Canada, Sept [10] Z. Boufidis et al., Evolution of the Reconfiguration Plane for Autonomic Communications, Proc. 15 th IST Mobile and Wireless Communications Summit, Mykonos, Greece, June 2006 [11] Z. Boufidis et al., End-to-End Architecture for Cognitive Reconfigurable Wireless Networks, Proc. 16 th IST Mobile and Wireless Communications Summit, Budapest, Hungary, July 2007 [12] 3GPP TR : 3GPP system architecture evolution (SAE): Report on technical options and conclusions, Oct End-to-End Reconfigurability II (E 2 R II) White Paper December 07 11/12

12 [13] C. Kloeck et al., Functional Architecture of Reconfigurable Systems, Proc. WWRF#14, San Diego, California, July 2005 [14] Z. Boufidis, E. Patouni, and N. Alonisitioti, End-to-End Reconfiguration and Control System Architecture, E 2 R II White Paper, June 2007, available at [15] 3GPP TS : GPRS enhancements for E-UTRAN access (Release 8) [16] 3GPP TS : "Architecture Enhancements for non-3gpp accesses (Release 8)" [17] P. Codier et al., Cognitive Pilot Channel, Proc. WWRF#15, Paris, France, Dec [18] E. Mohyeldin, J. Luo, J. Pan, and P. Slanina, Common Pilot Method Enabling Network Assisted Fast Scanning for Reconfigurable Terminals, Proc. 4 th Karlsruhe Workshop on Software Radios, Karlsruhe, Germany, Mar [19] 3GPP TR : 3rd Generation Partnership Project; Technical Specification Group Radio Access Network; Improvement of RRM across RNS and RNS/BSS (Release 5) [20] 3GPP TR : "Improved Network Controlled Mobility between E-UTRAN and 3GPP2/Mobile WiMAX Radio Technologies" [21] ITU-R Rec. M.1645, Framework for systems beyond IMT-2000, [22] 3GPP TS : 3rd Generation Partnership Project; Technical Specification Group Services and System Aspects; Service Requirements for the All-IP Network (AIPN); Stage 1, [23] Z. Boufidis and N. Alonistioti, E 2 R II Reconfiguration and Control Architecture, 3GPP doc. RP , Dec. 2006, available at End-to-End Reconfigurability II (E 2 R II) White Paper December 07 12/12