REDUCING HANDOFF LATENCY IN MOBILE IPv6 BASED WLAN

Transcription

1 REDUCING HANDOFF LATENCY IN MOBILE IPv6 BASED WLAN Muhammad Arif Amin PhD Student Department of Computer Science and Information Systems Universiti Teknologi Malaysia UTM Skudai Johor D. T. Malaysia Dr. Kamalrulnizam Abu Bakar Department of Computer System and Communication Faculty of Computer Science & Information Systems Universiti Teknologi Malaysia UTM Skudai Johor D. T. Malaysia Tel: Prof. Abdul Hanan Abdullah Department of Computer System and Communication Dean Computer Science & Information Systems Universiti Teknologi Malaysia UTM Skudai Johor D. T. Malaysia ABSTRACT Internet technology has widespread throughout the world over the past few years; this technology has offered variety of services and resources to the users. In contrast to traditional fixed networks the world has become mobile. Mobile computing offers users access to resource anytime anywhere. People have started communicating over wireless area networks (WLAN) using laptops and handheld devices such as new generation mobile phones, and PDAs. Internet Engineering Task Force has designed protocols to provide solution to mobile computing. Mobile IP, Mobile IPv4 and Mobile IPv6 provide mobility to users and enable them to communicate with each other while moving. However there are still technical issues that must be overcome before mobility can be widely deployed. One of the most important one is the handoff latency which occurs during change in point of attachment. MIPv6 is designed to provide user mobility by providing different network addresses when moved to a new network; however it still posse s challenges during handoff in which realtime application such as VoIP, audio and video streaming stops working. This research provides a solution that can improve MIPv6 communication performance during handoff in WLANs. An experimental test bed will be setup in a lab to investigate the real-time applications performance during handoff and latencies will be measured. The experimental results will be used in a simulation scenario to develop methods to speed up the handoff process and reduce the handoff delay. Categories and Subject Descriptors C.2.1 [Network Architecture and Design]: Wireless Communication General Terms Performance, design, experimentation and verification Keywords Mobility, Mobile IP, Mobile IPv4, MIPv6 and Handoff 1 INTRODUCTION & BACKGROUND Traditionally people used to avail network services using cables plugged in to the wall jack and were considered to be stationary [1]. This has led to the researcher to invent wireless technology to allow users to move from one place to other and hence called as mobile user. Wireless technology provides users to move within set boundaries of the network however with the advancement in technology and user requirements, organization started to deploy wireless across campus and enterprise network. However this facility did not provide uninterrupted real-time communication such as Voice over IP (VoIP), audio streaming and video streaming to users when they move to different subnets, between logical networks called virtual Local Area Network (VLAN) or Internet Service Providers (ISP). Mobile computing provides mobile users bidirectional communication anytime anywhere, however it includes the requirement that ongoing session should not break when the user is moving from one point of attachment to other. Internet Protocol (IP) was originally designed for the fixed networks (RFC 791) [2]; IP addresses were associated with fixed network computers and were required to be unchanged for the current session, however if the user moves to a different network, the computer was rebooted to gain network connectivity by obtaining a new IP address. Therefore to satisfy the requirements of the mobile users, Mobile IP (RFC 2002) [3] was proposed by the Internet Engineering Task Force (IETF). Mobile IPv4 (RFC 3344) [4] was developed to provide mobility to IPv4 users which were mainly fixed nodes users, however due to the address space restriction in IPv4 Internet Engineering Task Force (IETF) began to work on new protocol version of IP called IP version 6 (IPv6) (RFC 2460) [5]. IPv6 follows the Internet addressing architecture which allows nodes to communicated freely with another node regardless of their physical location (RFC 4291) [6]. Since MIPv4 did not get deployed commercially due to the limitations in the IPv4 addressing architecture, mobility in IPv6 was considered from the beginning of IPv6 design hence MIPv6 (RFC 3775) [7] was developed as an extension in the header of IPv6 protocol. Commercial deployment of native IPv6 has only begun in few countries, with ASIA as the faster and North America as lowest. 668

2 However ISPs has also started to deploy IPv6 websites and fewer than 78 ISPs are listed on the website of IPv6tf until to date [8]. The focus of this research is to better understand the architecture and handoff operation of MIPv6 protocol by implementing in a test lab. In addition to this evaluate the consequences of key Mobile IPv6 parameters responsible for the handoff delays which cause real-time applications to disconnect. Once these parameters are known a simulation model will be developed to reduce the handoff latency of Mobile IPv6 in Wireless Local Area Network (WLAN). 2 OBJECTIVES OF THE STUDY 2.1 Technical Issues with Mobile IPv6 Mobile IP [4] was developed for host configured with IPv4 [2] addresses, however the protocol is known since more than 10 years, the deployment of Mobile IP did not commercialized. One of the main reasons for that is the shortage of IPv4 addresses which are almost occupied and will not be available more in the near future. Since the development of IPv4 did involve mobility part, therefore due to the growth in the wireless world, Mobile IP was developed as an additional component to IPv4 protocol. An extensive research during 1990s and 2000s on Mobile IPv4 has led to many drawbacks such as triangular routing, packet flow, poor detection of network changes, authentication and authorization [9]. Since more and more electronic devices are been invented and the usage of Internet has increased, people need to access services while in motion, at hotspots and even during flights. At the same time to leverage these services deployment such as WiMax (broadband wireless) [10], General Packet Radio Service (GPRS) [11] and Wi-Fi has been introduced commercially. This has led to the demand of more and more IP addresses; therefore IPv6 came as a savior of network world to overcome the problem of shortage of IPv4 addresses. At the same time based on the experience from IPv4, the mobility part was included from the beginning of the development of IPv6. Due to the high demand and increase popularity of services Mobile IPv6 has gained strong research interest in the research world as platform for supporting mobility in real-time applications. One of the most important areas which need attention is the handoff latency (handoff: moving from one network to another) during which packets loss occurs due to the change at the point of attachment (link layer: L2) and IP address (network layer: L3), this leads to the disconnection of the realtime application. 3 RESEARCH GOALS The goals of this research are described below. Analyze the characteristics of standard Mobile IPv6 technology. Describe and analyze the handoff Latencies of Mobile IPv6. Describe and analyze handoff latencies of MIPv6 in WLAN environments. Evaluate the performance impact of MIPv6 handoff in WLAN using real-time applications (Voice and Video) by performing experiments in a test bed. This research will run through analysis of the Mobile IPv6 handoff latency in a WLAN environment. An experimental test bed will be setup in a lab environment which will include wireless access points, routers, fixed hosts and mobile nodes. Experimentation will include running of real-time applications for example Voice over IP and video streaming. Latency will be measured and recorded during the motion of the Mobile node using networking management tools [12] and sniffers [13].These results will then be used in setting up a simulation module for OMNET++ [14] to develop a method to reduce the latency in WLAN. 4 LITERATURE REVIEW 4.1 Traditional Computer Networks Traditional computer networks consisted of computers connected to network using cables, therefore also referred as fixed computers or nodes because they cannot be moved from one location to other. Internet protocol version 4 (IPv4) [2] was used to provide communication between them and share information. IPv4 provides unique 32 bit address to the nodes in the network and follows hierarchical structure to allow communication between networks called routing. Routing occur a network layer also called as layer three in the Open System Interconnection model (OSI) which consists of seven layers. When the computer is moved from one network to another, it needs to reconfigure a new IP address to communicate with the devices, however this delay is negligible in the traditional networks because it could involve that the computer is been switched off and restarted in a new location, therefore no continuation of communication is required. 4.2 Wireless Local Area Network (WLAN) WLANs provide flexibility to users through the use of radio waves instead of wires to carry data from one point to other. WLAN was originally standardized by the IEEE 802 LAN/MAN committee in 1980 [15]. A wireless commercial standard was introduced by 802 projects and was called b; this standard provides an 11 Mbps data rate to mobile users. The standard was developed to utilize the Industry, Scientific and Medical (ISM) band at 2.4 GHz shared medium. In this standard, since there is no control over network attachment, the total capacity of the medium is shared among users and decreases as the number of users increases, thus providing lower data rates. However, WLANs are able to switch data rates dynamically as users attach and detach; they also have the ability to switch speed based on the signal strength. 4.3 Wireless LAN Topologies WLANs are designed for flexibility and availability; they are composed of a mobile node or station which consists of a wireless network card and software installed to work with the WLAN. The Access Point (AP) a special type of device which acts as a bridge between the mobile station and the network. Wireless standards define two types of topologies, infrastructure 669

3 and ad-hoc. The simplest setup is ad-hoc, which is also called an Independent Basic Service Set (IBSS); in which all MNs communicate with each other as peer to peer, with no AP present between them, as shown in Figure 1: Independent Basic Service Set (IBSS) or Ad-Hoc. of mobility, in which a MN moves or changes its point of attachment between APs connected to different networks. Figure 3: Extended Service Set (ESS) Figure 1: Independent Basic Service Set (IBSS) or Ad-Hoc The Basic Service Set (BSS) is the foundation of the Infrastructure mode, in which there is at least one AP connected to the wired network and all MNs are connected to the AP access network services through this AP, as shown in Figure 2: Basic Service Set (BSS) To achieve successful mobility, a MN must perform Layer 2 (L2) and Layer 3 (L3) functions. L2 mainly depends on MAC address exchange between MN and AP. L3 depends on changing the IP address of the MN; however, L3 cannot be processed until the L2 process is finished successfully [16]. Since the L2 process is hardware dependent, it is mainly controlled by the manufacturer of the wireless network interface card (WNIC) and the driver; it may also depend on the signal strength or other environmental conditions. A passive scanning process at the WNIC driver level helps MN find APs within the range and switch to a certain AP when required. However, the process of L3 switchover requires IP change and infrastructure participation and configuration. This involves network devices to send signals to other devices when the MN moves (between old and new networks?). These devices are called access routers, and are connected through a common network; thus they are able to send MN information. To cater to mobility, the concept of the mobile IP protocol was standardized by IETF [3]. Figure 2: Basic Service Set (BSS) As mentioned before, wireless is a shared medium and it has a tendency to degrade data rates based on the number of users and signal strength; a network can deploy an Extended Service Set (ESS) in which more than one AP is connected to the wired network. This provides coverage of large areas and roaming between access points. This topology demonstrates connecting multiple BSS to a wired network. A user moves within the BSS and between BSS to perform mobility. This setup also provides users with adequate signal strength and load when the number of users increases, as shown in Figure 3: Extended Service Set (ESS). When users roam between BSS, they will find an AP and attempt to connect based on the signal strength available. ESS introduces the possibility of forwarding MN traffic from one BSS to another though a wired network by providing minimum disruption; this is because both APs are connected to the same wired network or a network switch. This introduces the concept 5 MOBILITY CONCEPT The development of real-time applications such as voice of IP (VoIP) and IP based data in the context of mobile devices (for example Laptops, 3G, 4G mobile phones and PDAs) and the new access technologies (such a Wireless LAN, GPRS, Bluetooth, ADSL and cable modems) demands mobility support at the network layer. The main aim is to allow mobiles nodes to keep communication with other hosts while roaming between different networks. Roaming occurs when a node moves from one access network to another; it constitutes physically moving from one network to another. Standard protocols like IPv4 and IPv6 will result in disconnection of ongoing communication. Mobile IP allows host to perceive minimum disruption during movement and then continue communication. As indicated in [2] traditional networks maintain host IP address, host identity with respect to its topological location. Host movement often result in a new location with respect to network IP topology, this results in a demand to acquire new IP address in order to route packets to host s new location. The new location could be a sub-network which is also referred to as grouping of hosts in a certain topological network which differs 670

4 at IP layer. Since host IP address is also used in transporting the data to upper layer of OSI model, it is likely that when the IP address changes the transport layer may disconnect (for example TCP sessions). Packets sent to the previous IP address are lost and host s previous peers are not able to communicate, because they are unaware of the new IP address assigned to the host. Since all the layers in OSI model depend on each other, the transport layer depends on IP layer, so many application do not react actively when the IP address changes. When the IP address changes application needs to restart the session, not only that the binding of host name to IP address in the Domain Name Server (DNS) also has to change based on the movement. Some examples of the applications which react due to the change in IP address are Voice over IP (VoIP), Virtual Private Networks (VPN), database applications, real audio and video streaming. Mobile IP is the savior of these problems by making the movement transparent to the upper layer like transport and session and application. The key target of Mobile IP are the people in motion, it does nothing to the stationary users who does not require movement. 6 MOBILE IP Mobile IP has been developed to provide mobility to the host which changes their point of attachment [4]; however the idea was that the application on the host can survive the handoff delays. Mobile IP provides mobility by introducing two IP addresses for the mobile hosts called as Mobile Node (MN), a static IP address called as home address (HoA) which is assigned by the home network, with this IP the host is recognized globally. The second IP called as care of address (CoA) which is a temporary IP address assigned by the visiting network until nodes stays in the foreign network. A special mechanism of mapping between these two IP address is used to send packets to the host s CoA to allow communication between previous peers and new peers. To communicate between two networks, devices such as routers are used and must be configured with special parameters to handle mobility functions. The router used at home network is called as Home Agent (HA) and the router at foreign network is called as Foreign Agent (FA). Any node either fixed or mobile communicating with this mobile node is called as correspondent node (CN). To understand this process more clear let s take an example; a user is connected to a home network by a home IP address HoA and downloading a file from the Internet on a wired LAN. Now the user wishes to leave the current place of attachment and would like to move away so he disconnects the cable from his computer, at this point the file would stop downloading in a traditional network. However if the place of attachment is installed with an wireless LAN, Mobile IP could allow the communication to continue and keep the file downloading. This is also true if the mobile node moves from one network to another and the communication between nodes continues as shown in Figure 4. Figure 4: Mobility 6.1 Handover Process Handoff process provides mechanism for users to roam, however in order to roam freely a user must not disconnect from the network. To achieve that network entities should communicate with each other and be able to transfer information while the mobile node (MN) is moving from one network to another, these entities are called as access routers [16]. They send packets to the mobile node during roaming which is possible if the IP address space is within the same hierarchy of the home network, but this is not possible because the MN may change its location to a different network with different address space. To perform the process of handover access router must be configured with additional feature which allow packets to move continuously to the foreign network keeping the MN in contact always. To achieve a successful Handoff process following requirements must be fulfilled. Link Detection: It depends mainly on the process where a MN may detect physical link parameters during movement; it could involve wireless signals to discover new wireless access point (WAP). This process depends on the wireless network card and its driver, passive scanning, wireless cell size overlaps and the speed of motion. However it may trigger the process of layer 3 handoff in which MN tries to gain access to the new network. Link establishment: Once the link is detected the MN tries to establish link with the WAP which involves exchange of MAC addresses hence working in the link layer (layer 2) parameters. Network Detection: After link establishment a MN tries to detect network parameters which involve network address space. This process allows the MN to find the network number which is also called as network prefix. This process will eventually allow MN to acquire a new IP address. IP address acquisition: In this process a MN acquire a new IP address from the Network. This process involves address auto configuration, which either by DHCP server or the link layer process. Duplicate address detection: This is a very important process during handoff to confirm that the IP address acquired is unique. In this process the MN announces its IP address on the network and check if that IP is being used by any other node on the network. However the chances of duplication of the IP address depends on the type of protocol used. 671

5 Access authorization: Access authorization allow MN to gain IP address from the network if successful, however if the MN is not allowed to access the network then IP address configuration process would not complete. This process could involve AAA system or if the network is wireless then provides multiple levels of authentication and authorization using many protocols. Mobility process depends a lot on the access authorization; it can easily break if the MN is not allowed to access the network thus dropping all the connections. Packet forwarding: Once the access is authorized and an IP address is configured packets starts to flow toward the MN in order to maintain communication. All the packet flow toward the IP address gained at the present network which could be a temporary IP address. Informing correspondent nodes about the new IP address: This can be done either using the home agent or directly informing the correspondent nodes depending on the protocol used but nevertheless it is a onetime process. 6.2 Mobile IP handoff Process: To achieve a successful handoff process a mobile node must perform following steps. Movement detection: A mobile node uses a mechanism to detect the change in the point of attachment; this could be a Layer 2 trigger or a signal from an access router. Registration: When the MN detects that it has moved to a foreign network it tries to obtain a care of address (CoA). This CoA should be registered in the MN s home agent (HA) in the home network. Tunneling: Tunneling is a default mechanism of Mobile IP to route packets to the MN s CoA by the HA. Home agent (HA) is responsible of intercepting all the packets destined to MN s HA and route them to the CoA, this is done by encapsulating packets by creating IP in IP payload configuration [17]. Binding: A binding is an association between the MN s CoA to HoA. A binding list is maintained in the access router or HA. A binding list is also used to inform the correspondent nodes (CN) about the new IP address of the mobile node. Once the MN returns back to the home network, its binding is deleted from the list. 7 PROBLEMS WITH MOBILE IP To achieve a successful handoff process, a mobile node goes through many steps as listed in section 6.2, however these steps also creates delays during movement which are listed below. Detection delays: This delay is associated with the detection of the new network; it may involve L2 or L3 signaling. However initially L2 signaling will be required to detect the change and then L3 will follow. Address Configuration delays: Address configuration delay involves the process of getting new IP address (CoA) and verifying its uniqueness. Registration Delays: Registration delay is a time required binding the CoA with HoA in the Home agent (HA); it could depend on the number of hops used to send the information. 8 MOBILE IPV4 Mobile IPv4 [4] was developed to overcome the problems addressed in Mobile IP. It also introduced the concept of foreign agent as an external body or the router. The process involves mobile node which is always addressed by it Home address (HoA) and while at foreign network obtains a care of address (CoA). This Care of Address can be obtained by the foreign agent or co-located (though a third party agent like DHCP). A mapping between HoA and CoA is maintained by the Home agent (HA). 8.1 Mobile IPv4 Operation [4] This section in detail mobile IPv4 handoff operation and signals associated with it. Mobility agent discovery: Foreign agents and home agents advertise their presence via Agent Advertisement messages. A mobile node may optionally solicit an Agent Advertisement message from any locally attached mobility agents through an Agent Solicitation message. Mobile node uses agents to determine if it is connected to the home network or foreign network. When the mobile node detects that it is located on its home network, it operates without mobility services. If returning to its home network from being registered elsewhere, the mobile node deregisters with its home agent, through exchange of a Registration Request and Registration Reply message with it as shown in Figure 5. Figure 5. Mobile IPv4 Handoff Process Address Configuration: When a mobile node detects that it has moved to a foreign network, it obtains a care-of address on the foreign network. The care-of address can either be determined from a foreign agent s advertisements (a foreign agent care-of address), or by some third party mechanism such as DHCP server [18]. Registration: The mobile node operating away from home then registers its new care-of address with its home agent through exchange of a Registration Request and Registration Reply message with it, possibly via a foreign agent. 672

6 Tunneling: All packets sent to the mobile node s home address are intercepted by the home agent and tunneled to the care of address (CoA). In the reverse direction, datagram sent by the mobile node are generally delivered to their destination using standard IP routing mechanisms, not necessarily passing through the home agent. When away from home, Mobile IP uses protocol tunneling to hide a mobile node s home address from intervening routers between its home network and its current location. The tunnel terminates at the mobile node s care-of address. The care-of address must be an address to which datagram can be delivered via conventional IP routing. At the care-of address, the original datagram is removed from the tunnel and delivered to the mobile node. 8.2 Problems with Mobile IPv4 Mobile IPv4 node goes through many steps to achieve a handoff process, however these process cause delays during the movement and results in latency not suitable for real-time applications. Agent Discovery Delays: The agent discovery delays are caused during the movement; it involves router advertisements in the foreign network. Generally a host requires minimum of three router advertisements if the interval is one second then the total time to discover the agent is 3 second which is not good for real-time applications. However if the mobile node detects the network change by IP address prefix advertisement then it is 0.5 second. Triangular Routing: Triangular routing involves process of sending all packets to correspondent node through the home agent. This delay the packets because of hops covered to reach the CN, therefore could disconnect the application by disturbing the upper layers. Handoff Problems: Handoff problems could cause packet loss, jitter and delays. In particular wireless networks such as , disturbance over the air, noise and radio frequency weakness could cause delays. Ingress Filtering: Ingress filtering is used to detect address spoofing in routers for outgoing traffic which can cause denial of service attack [19]. Routers only accept packets with source address which is topologically correct and valid prefix. In Mobile IPv4 process, the MN send the packets to the CN it uses it home address rather than CoA, therefore home agent router due to the ingress filtering reject these packets and does not allow communication to CN. Private Addressing Scheme: Due to the shortage of IPv4 address, a concept of private addressing scheme has been introduced to cope with the IP addresses [20]. Private IP addresses are not allowed to be routed to the internet; therefore a public IP address is required by the mobile node to move to different network. 8.3 Mobile IPv4 Optimization Mobile IPv4 extension has been developed in order to eliminate some of the handoff latencies, which are discussed in this section. Reverse Tunnel: It provides the solution to the Ingress filtering on the router. Basically reverse tunnel allows the mobile node to encapsulate the packets to CN by using the same tunnel used for forwarding the packets [21]. Route Optimization: Route Optimization solves the problem of triangular routing by providing direct path communication between MN and CN. This done by directly sending the binding update to CN informing the CoA address MN has configured [9]. Smooth Handoff to Foreign Agent: This process allow mobile to handoff to a new foreign agent smoothly by tunneling packets to MN s new care of address. The MN send binding update to the old FA during registration process with the new one, this process allows the packet flow to reach the MN with disconnecting. 9 MOBILE IPV6 REQUIREMENTS It is already discussed that to leverage mobility, Mobile IP is the main requirement. However we have also seen that there are many deficiencies in the protocol and efforts has been made to resolve these. Below are some of the requirements Mobile IPv6 poses to fulfill to satisfy mobility. Session Continuity: Session continuity allows mobile node to continue communication when it changes its IP address. An example could be given as mobile phone users keep talking while moving from one place to other even at high speed, their session does not break. Reachability: It the most important component of the mobility, just to find out where the node is and how to communicate with it. A mobile node keeps changing its location but it is important for other nodes to know the location and reach it through network. Transparency: Mobile IPv6 should not require any change in the host application layer; the operation should be transparent from the upper layers for a successful mobility. Lower Layer Independence: Mobile IPv6 should not be dependent on specific link layer technology. It should be able to use in all type of wireless/wired technologies. End to end signaling: Mobile IPv6 should be able to provide end-to-end communication, the hosts should be able to communicate and the access router should be able to route packets based on the source and destination addresses. Scalability: Mobile IPv6 should be able to provide scalable network communication, therefore should be able to allow local or global mobility. 10 MOBILE IPv6 HANDOVER A handoff process consists of layer 2 attachments after that discovering new routers, address configuration, movement detection and then IP registration which lead to layer 3 handoff process. During the handoff process a mobile node is not able to send or receive data from any other node in the home or foreign 673

7 networks; therefore some packets may be lost or delayed, this leads us to a concept of Handover Latency. MIPv6 handoff signaling process is shown in Figure 6. Figure 6. Mobile IPv6 Handoff Process Movement Detection: Mobile node detects that it has moved to a new network, this is done by the router or the mobile node. Routers send router advertisements (RA) at regular intervals to inform their presence. Router advertisements are multicast packets, RFC 3775 recommends interval value of 30msec to 70 msec. When the mobile node moves into a new network it listens to the RAs and detect that it has to form a new IP address. However if the Mobile node could not listens to the RA, it can force the router by sending Router Solicitation message to request RA by using neighbor discovery method [22]. Obtaining Care of Address (CoA): When a mobile node discovers the new router it forms a new IP address called as Care of Address (CoA). However to register this CoA it has perform Duplicate Address Detection to check its uniqueness [22]. Since the delay caused by the DAD process is high an alternate method of performing DAD is been standardized by IETF which is called Optimistic Duplicate Address Detection [23]. After confirming the address uniqueness the mobile node form the IP address either by stateless [24] or statefull address configuration through Dynamic Host Configuration Protocol version 6 (DHCPv6) [25]. Binding Management: The binding management performs binding acknowledgement, binding update and binding request process. Returning Home: The mobile node detects that it has returned home by listening to the RA from the router and configuring the network prefix. It than instruct the HA to remove the binding entry of the mobile node in order not to tunnel packets. susceptible to the delays and hence will provide poor communication Mobile IPv6 Extensions Fast Hanover for MIPv6 (FMIPv6) FMIPv6 [28], [29] is an extension to MIPv6 protocol. In this protocol a concept of Previous Access Router (PAR) and New Access Router (NAR) is proposed. This extension allows MN to connect to NAR while still communicating with PAR. Once the MN is connected to NAR, it requests to PAR to tunnel packets to NAR so that packets are not lost. It also allows the MN to configure old Care of Address (ocoa) and New Care of Address (ncoa). If the MN moves too fast between networks, then it may not be possible to produce tunnel between PAR and NAR which results in packet lost and also there is a dependency on the L2 triggers. FMIPv6 has two modes of operation called as Predictive and Reactive which are discussed below. Predictive Fast Handoff: In this process the MN obtains the ncoa before it actually moves to a new network. During this process the MN request to build tunnel to the NAR as shown in Figure 7. Reactive Fast Handoff: In this process the MN first moves in the new network then request the PAR to establish tunnel to NAR, during this time it also acquires ncoa as shown in Figure 8. Figure 7. Mobile IPv6 Fast Handover (Predictive) 10.1 MIPv6 Problems during Handover The original mobile IPv6 protocol was not designed to handle rapid change in network, however with all the modifications to previous protocols it did not provide promising results in terms of handover delays. In worst case scenario the MN will keep doing handovers rather than communication. Since the handover takes longer duration to complete, packets are lost or dropped specially for the real-time applications. Both connectionless UDP [26] and connection oriented TCP [27] protocols will be 674

8 Figure 8. Mobile IPv6 Fast Handover (Reactive) Hierarchical Mobile IPv6 (HMIPv6) A Mobile node may change its point of attachment so fast or frequent so that the access router may not be able to address issue. Although FMIPv6 is designed to handle that, however it also generates signaling overhead. Therefore to reduce signaling during handoff and handle fast movement Hierarchical Mobile IPv6 extension is developed by IETF [30]. It is aimed to provide localized mobility such as micro mobility to the MNs by introducing a network entity called as Mobility Anchor Point (MAP). MAP is used to control the mobility domain in which MN can move without any delays; this is achieved by the address hierarchy within the domain. In HMIPv6 a MN has two IP addresses, a regional Care of Address (RCoA) which identifies the global link or Internet and an on-link Care of Address (LCoA) which defines the local network. When the MN moves within the MAP domain, its LCoA is changed. When the MN moves to a different MAP domain then it has to change both LCoA and RCoA and register with HA to perform successful handover as shown in Figure 9 (Intra-MAP) and Figure 10 (Inter-MAP). Figure 9. Hierarchical Mobile IPv6 (Intra-MAP) HMIPv6 still need a further handover enhancement for supporting the real-time applications. FMIPv6 is the typical protocol to reduce the handover latency for Mobile IPv6 nodes. By introducing hybrid protocol which is a combination of FMIPv6 and HMIPv6it is possible to reduce the handoff delays. 11 HANDOVER IN WLAN An handover takes place when a Mobile Host changes its association from one Access Point to another which is also called as "re-association". This process consists of the following steps [32]: The MN realizes that a handoff is necessary due to degrading radio transmission environment for the current AP. The MN performs a scan to see what APs are available. The result of the scan is a list of APs together with physical layer information, such as signal strength. The MN chooses one of the APs and performs a join to synchronize its physical and MAC-layer timing parameters with the selected AP. The MN requests authentication with the new AP. For an "Open System", such authentication is a single round-trip message exchange with null authentication. The MN requests association or re-association with the new AP. A re-association request contains the MAC-layer address of the old AP, while a plain association request does not. If operating in accordance with i [33], the STA and AP would execute 802.1X EAP-on-LAN procedures to authenticate the association (step 3 would have executed in "Open System" mode). The new AP sends a Layer 2 Update frame on the local LAN segment to update the learning tables of any connected Ethernet bridges. Since networks depends on layer 2 process very much, therefore FMIPv6 protocol was designed to leverage information from the link layer to suite network. In this process a MN acquires the IP address before it actually moves to a new point of attachment. However MN has to register the IP address with the home agent (HA) therefore the round trip time cause delays which disturbs the real-time application communication as shown in Figure 11. Figure 10. Hierarchical Mobile IPv6 (Inter-MAP) Fast Handoff for HMIPv6 (F-HMIPv6) Fast Handover over HMIPv6 (F-HMIPv6) [31] networks is another scheme to reduce the handoff delay in MIPv6. The HMIPv6 was developed to reduce the signaling overhead and delay concerned with Binding Update in Mobile IPv6. Therefore Figure 11. Fast Handover in WLAN 675

9 12 DISCUSSION MIPv6 has been developed by Internet Engineering Task Force (IETF) to provide mobility to the hosts with the original concept of Mobile IP. Since Mobile IP was developed to provide mobility to users with fixed nodes, it did not pose any issues with the delays during change in network. However with the growth in the Wireless technologies, mobility became more important and hence users need to connect to the networks even if they move from physical location to another. Mobile IP suffered with the IP addressing scheme of IPv4, therefore it was not possible to provide IP address to each and every node so that they can communicate over the Internet. MIPv6 is an extension of IPv6 protocol, therefore it provide enough IP addresses so that almost every electronic device on earth can have an IP address which can communicate to other devices over the Internet. Mobile IPv6 provides host based mobility, which means it is the host which initiates the change in the network and inform the access router of leaving the location and entering the new location. It poses many delays during this process; these delays are not suitable for the real-time applications. To overcome these delays many extensions to the base MIPv6 protocol has been developed which includes, A Hierarchical Mobile IPv6 (HMIPv6), as proposed in [34], [30] to reduce signaling durations during handovers; to the corresponding node (CN) and to the HA. Fast Handovers in Mobile IPv6 method is another extension to MIPv6 to reduce the delays which is proposed in [28] [29] and. A similar idea of Resource Reservation Protocol is presented in [35], and Flow-based fast handover for Mobile IPv6, done by [36], is also a method reducing handover delay. A Hybrid handover mechanism is a combining of HMIPv6 and FMIPv6, such as [37], [38]. Handover Latency has been studied widely in WLANs; most of the research has been carried out to develop methods to reduce delays generally in any wireless media. However some studies like [39] [40] [41] [42] has been done to propose method of reducing delays during handover. However none of the above mentioned methods propose as when the new Access Point would detect the Mobile node, or how to make new Access Point detect that the Mobile node is going to attach with it. 13 CONCLUSION & SUGGESTED WORK Mobile Internet users demand all-the-time connectivity, even when they move from one network to another, which requires constant communication with the network devices. Mobility is possible using IPv4; however, the growing number of Internet users means that IPv4 is not the best choice because it cannot cope with the rapid growth and does not have all of the required physical and logical structures. In contrast, IPv6 seems to be able to address many IPv4 mobility issues. Both of these protocols have variants to provide mobility, such as MIPv4 and MIPv6. However, handover latency is an issue common to Internet mobility, and both IPv4 and IPv6 are facing it as a major challenge. Handover latency is the result of the time required to execute the handover algorithm. For example, in Mobile IPv6, each MN is tracked by it s HA. When an MN moves to a new network, a new IP address (CoA) will be provided. This process is common in both MIPv4 and MIPv6; however, to reduce signaling processes, MIPv6 does not have FA but still does not allow real-time communication to work. Many extensions to the base Mobile IPv6 has been developed which are discussed in the Discussion section but none of them has really suggested methods to reduce the latency in Wireless Local Area Networks. This research will provide a context aware method which will allow new access point (AP) to discover the Mobile node before it moves and thus allow the data to be transferred immediately when the node performs registration. A model will be developed in a simulation environment using OMNET++ and results will be shown graphically. 14 REFERENCES [1]. J. H Saltzer, D.P. Reed and D.D. Clark. "End to End Arguments in System Design". ACM Transaction in Computer Systems 2, Vol 4,. November, [2]. Postel, Jon. Internet Protocol DARPA Internet Program Protocol Specification. IETF RFC 791. September [3]. Perkins, Charles. "IP Mobility Support". IETF RFC October [4]. Perkins, Charles E. "IP Mobility Support for IPv4,". IETF RFC August [5]. Deering, Stephen E and Hinden, Robert M. "Intenet Prtocol Version 6". RFC December [6]. Hinden, Robert M and Deering, Stephen E. IP Version 6 Addressing Architecture. IETF RFC Fabruary [7]. Johnson, David B, Perkins, Charles E and Jari Arkko. "Mobility Support in IPv6,". IETF RFC June [8]. ISP List websites with an IPv6 address. ipv6tf.org. [Online] March [Cited: March 21, 2009.] [9]. Perkins, Charles E. Mobile Networking through Mobile IP. IEEE Internet Computing. s.l. : IEEE Computer Society, January [10]. WiMAX Forum. [Online] [11]. Soininen, Jonne. Transition Scenarios for 3GPP Networks. IETF RFC August [12]. Network Management Software. [Online] [13]. Wireshark. [Online] 676

10 [14]. OMNET++ Descrete Event Simulation System. [Online] [15]. IEEE 802 LAN/MAN Standard Committee. [Online] [16]. Koodli, Rajeev and Perkins, Charles. "Mobile Internetworking with IPv6 Concepts, Principals, and Practices". s.l. : Willey, [17]. Perkins, Charles. Minimal Encapsulation within IP. IETF RFC October [18]. Droms, Ralph. Dynamic Host Configuration Protocol. IETF RFC March [19]. Ferguson, Paul and Senie, Daniel. Network Ingress Filtering: Defeating Denial of Service Attacks which employ IP Source Address Spoofing. IETF RFC May [20]. Rekhter, Yakov, et al. Address Allocation for Private Internets. IETF RFC Fabruary [21]. Montenegro, Gabriel E. Reverse Tunneling for Mobile IP. IETF RFC January [22]. Narten, Thomas, Nordmark, Erik and Simpson, William Allen. Neighbor Discovery for IP Version 6. IETF RFC September [23]. Moore, Nick. Optimistic Duplicate Address Detection (DAD) for IPv6. IETF RFC April [24]. Thomson, Susan, Narten, Thomas and Jinmei, Tatuya. IPv6 Stateless Address Autoconfiguration. IETF RFC September [32]. McCann, Pete. Mobile IPv6 Fast Handovers for Networks. IETF RFC November [33]. Medium Access Control (MAC) Security Enhancements. IEEE Std i July [34]. Castelluccia, Claude. A Hierarchical Mobile IPv6 Proposal. SIRAC Project Report. November [35]. Plasto, Daniel. Fast RSVP handover in mobile IPv6. First Australian Undergraduate Students Computing Conference, [36]. Sulander, Miska, et al. Flow-Based Fast Handover Method for Mobile IPv6 Networks. IEEE Publication [37]. Hsieh, Robert, Zhou, Zhe Guang and Seneviratne, Aruna. A Seamless Handoff Architecture for Mobile IP. INFOCOM [38]. Kashihara, Shigeru, et al. End-to-End Seamless Handover Using Multi-Path Algorithm [39]. Aparicio, Albert Cabellos, et al. Measurment Based Analysis of the Handover in a WLAN MIPv6 Scenario [40]. Yading, Fang, et al. Research on Seamless Handover for WLAN with MIPv [41]. González, Francisco A, Pérez, Jesús A and Zárate, Victor H. HAMS: Layer 2 Handoff Accurate Measurment Strategy in WLANs [42]. Dimopoulou, Lila, Leoleis, Georgios and Venieris, Iakovos S. Fast handover Support in a WLAN Environment [25]. Bound, Jim, et al. Dynamic Host Configuration Protocol for IPv6. IETF RFC July [26]. Postel, Jon. User Datagram Protocol. August [27]. Postel, Jon. "Transmission Control Protocol". DARPA INTERNET PROGRAM. September [28]. Koodli, Rajeev. "Fast Handover for Mobile IPv6,". IETF RFC July [29]. Koodli, Rajeev. Mobile IPv6 Fast Handovers,. IETF RFC June [30]. Soliman, Hesham, et al. "Hierarchical Mobile IPv6 Mobility Management (HMIPv6),". IETF RFC August [31]. Jung, HeeYoung, et al. Fast Handover for Hierarchical MIPv6. IETF Internet Draft Expired. April