Information Science Institute, State Key Laboratory of Integrated Service Networks, Xidian University, Xi’an 710071, China
simple_xly@126.com
Show less
History+
Received
Accepted
Published
2010-02-10
2010-04-21
2010-12-05
Issue Date
Revised Date
2010-12-05
PDF
(353KB)
Abstract
Proxy mobile Internet protocol version 6 (PMIPv6) is a network-based mobility management protocol proposed by Internet Engineering Task Force (IETF), which allows nodes to remain service connectivity while moving around in the IPv6 Internet. PMIPv6 is different from the host-based mobility management protocols (mobile Internet protocol version (MIPv6), hierarchical mobile IPv6 (HMIPv6), fast handover for mobile IPv6 (FMIPv6), and fast handover in hierarchical mobile IPv6 (F-HMIPv6)), whose signaling are transferred among some network entities except mobile node (MN). This paper focuses on the analytical modeling of performance analysis for PMIPv6 and other protocols using IEEE802.16-based wireless metropolitan area networks as the wireless access network. The performances of these protocols are evaluated by some metrics like handover latency, service disruption time, and binding update cost. Numerical results show that PMIPv6 has better performance.
Internet Engineering Task Force (IETF) proposed a network-based mobility management protocol, proxy mobile Internet protocol version 6 (PMIPv6), in August 2008. PMIPv6 is a new method to solve IP mobility challenge. It is possible to support mobility for IPv6 nodes without host involvement by exchanging mobile IPv6 signaling message between a network node and a home agent. PMIPv6 is based on the mobile Internet protocol version (MIPv6) that has some advantages [1].
1) Reuse of home agent functionality and the message/format used in mobility signaling. MIPv6 is a mature protocol with several implementations that have undergone interoperability testing.
2) A common home agent would serve as the mobility agent for all types of IPv6 nodes.
Hierarchical mobile IPv6 (HMIPv6), fast handover for mobile IPv6 (FMIPv6), and fast handover in hierarchical mobile IPv6 (F-HMIPv6) are the extensions of MIPv6 that solve the problems of MIPv6 in different sides. MIPv6, HMIPv6, FMIPv6, and F-HMIPv6 are mobility management protocols that are based on the host.
Through the comparison of PMIPv6 and other protocols, we can see that the network-based protocol has advantage over other host-based protocols. The rest of this paper is organized as follows. Section 2 introduces PMIPv6 protocol and other protocols. Section 3 gives the mobility model and discusses the performance metrics of different protocols based on the mobility model. Section 4 analyzes the influence of parameters on performance of protocols. We finally conclude the paper in Sect. 5.
Mobility management protocols
PMIPv6
In PMIPv6 protocol, some new terms are given, which are to be interpreted as follows [1]:
1) PMIPv6 domain
PMIPv6 domain refers to a network where the mobility management of a mobile node (MN) is handled using the PMIPv6 protocol. There are two entities that are involved in PMIPv6 domain: local mobility anchor (LMA) and mobile access gateway (MAG).
2) LMA
LMA is the home agent for the mobile node in a PMIPv6 domain that is the entity to manage the mobile node’s binding state.
3) MAG
MAG is an entity in PMIPv6 domain that is responsible for tracking the mobile node’s movement. It has a function on an access router that manages the mobility-related signaling for a mobile node that is attached to its access link.
4) Proxy care-of address (PCoA)
PCoA is a globe address configured on the interface of MAG and is the endpoint of the tunnel between LMA and MAG. LMA registers it in the binding cache entry for the mobile node.
5) Proxy binding update (PBU)
PBU is a request message sent by MAG to finish the binding update process.
6) Proxy binding acknowledgment (PBA)
PBA is a response message sent by LMA to finish the binding update process.
Figure 1 gives the handover process of PMIPv6. PMIPv6’s IP layer handover is carried out after media access control (MAC) layer handover. First, MAG and LMA exchange PBU message and PBA message to establish a bi-directional tunnel. Then, MN sends a router solicitation (RS) message to the base station (BS), and BS replies a router advertisement (RA) message to MN. After that, the handover process is finished. MN uses the tunnel to deliver the data.
Other protocols
Figure 2 gives the handover process of different mobility management protocols, including MIPv6 [2], HMIPv6 [3], FMIPv6 [4], and F-HMIPv6 [5], and it also gives MAC layer handover process of IEEE802.16e [6].
Performance analysis
Network model
Figure 3 shows the network model, which includes MN, BS, access router (AR), mobility anchor point (MAP), home agent (HA), and correspondent node (CN). There are four MAPs, and each MAP has n ARs [7]. The coverage of MAP is called domain, and the coverage of AR is called cell. In other words, each domain has n cells. A BS connected to an AR has a wireless access interface for connecting MNs. HA and CN are connected to different MAPs. In this paper, we suppose MN is located in another MAP different from HA’s and CN’s MAPs. In Fig. 3, a, b, c, d, e, and f denote the hop counts between different entities.
We adopt fluid flow (FF) model in where the direction of an MN’s movement in an MAP domain is distributed uniformly in the range of (0, 2π). V is the average speed of MN. Let μc, μd be cell crossing rate and domain crossing rate, respectively. Furthermore, let μl be the cell crossing rate for that MN still stays in the same MAP domain. If we assume that each AR coverage area is SAR, the border crossing rate is given by [8]In this paper, we suppose that residence time in a cell and in a domain follow exponential distribution with parameters μc and μd, while session arrival process follows a Poisson distribution with rate λS. Hence, the average number of cell crossing and domain crossing can be obtained as follows:Let E(Nl) be the average number of cell crossing rate of an MN, which is in the same domain. We can get the expression [9]
Parameter analysis
Using the network model and mobility model, we analysis the important performance metrics, such as handover latency, service disruption time, and binding update cost. Let t1, t2 denote the unit transmission delay of a control packet in a wired, wireless link, respectively; c1, c2 the unit transmission cost of a control packet in a wired, wireless link, respectively; TX,Y the transmission time of a control packet between entities X and Y; CX,Y the transmission cost of a control packet between entities X and Y; and dX,Y the hop distance between X and Y. Although different control packets have different size, we assume that they have the same size for simplicity. Therefore, the transmission delay and transmission cost are proportional to the hop distance. We can get the following equations:where dX,Z is the hop distance of wired link, and dZ,Y is the hop distance of wireless link. We assume that Sc and Sd are the size of a control packet and a data packet. The relationship of Sc and Sd is thatTherefore, the transmission delay and cost of a data packet is k greater than a control packet.
Handover latency
Handover latency is defined here as the interval between the time MN sends mobile station handover request (MOB_MSHO-REQ) message and the time MN can send or receive data packets. We assume that the propagation time (10-3 ms), processing, and queuing delays are much smaller than the transmission time, so they are negligible.
For PMIPv6, MN starts IP layer handover after MAC layer handover. The handover latency can be expressed as follows:
Service disruption time
Service disruption time is defined as the period when MN cannot send or receive data packets. It usually lasts from when MN disconnects from the serving BS until it can send or receive the data packet in the new link. MN cannot receive data packets when it sends a mobile handover indication (MOB_HO-IND) message to the serving BS. Then, MN performs the IEEE802.16e network re-entry procedure and sends a fast neighbor advertisement (FNA) message. The service disruption time of PMIPv6 is given by the following equation:
Binding update cost
Depending on the type of movement and mobility management protocol, two kinds of binding updates can be performed: local binding update (Cl) and globe binding update (Cg) [10]. For MIPv6 and FMIPv6, globe binding update is performed regardless of movement type. For HMIPv6, F-HMIPv6, and PMIPv6, globe binding update is performed when MN moves out an MAP domain. The binding update cost is expressed as follows:
Supposing that SMR= λS/μc, the binding update cost becomes
For PMIPv6, the binding update cost is computed as follows:
For other protocols, the handover latency, service disruption time, and binding update cost can be obtained as follows:
Numerical results
In this section, we use the parameters in Table 1 to calculate the performance metrics.
Figures 4 and 5 show the variation of handover latency and service disruption time as the wireless link latency changes. Handover latency and service disruption time increase linearly as the wireless link latency changes. In Fig. 4, when MN moves in an MAP domain, the relationship of different protocols’ handover latency areWe can see that the duplicate address detection (DAD) process takes a long time in the whole handover process. PMIPv6 has the least handover latency because it has omitted the DAD process. In Fig. 5, when MN moves in a MAP domain, the relationships of different protocols’ service disruption time areMIPv6 and HMIPv6 take IP layer handover after MAC layer handover has finished, so the service disruption time is large.
Figure 6 demonstrates binding update cost as a function of MN’s velocity. The lower the MN’s velocity is, the less the binding update cost is. Figure 7 illustrates the variation of binding update cost as the SMR changes. When SMR is small, λS is less than μc. The larger μc causes frequent cell crossing, so the binding update is large. The relationships of different protocols’ binding update cost are
Conclusion
In this paper, we give the performance analysis of different mobility protocols (MIPv6, HMIPv6, FMIPv6, F-HMIPv6, and PMIPv6) in terms of handover latency, service disruption time, and binding update cost. Numerical results show that the PMIPv6 has least handover latency, service disruption time, and binding update cost.
JohnsonD, PerkinsC, ArkkoJ. Mobility support in IPv6. <patent>IETF RFC 3775</patent> 2004
[3]
SolimanH, CastellucciaC, El MalkiK, BellierL. Hierarchical mobile IPv6 mobility management (HMIPv6). <patent>IETF RFC 4140</patent> 2004
[4]
KoodliR. Fast handovers for mobile IPv6. <patent>IETF RFC 4068</patent> 2005
[5]
JungH Y. Fast handover for hierarchical MIPv6 (F-HMIPv6). <patent>IETF draft edition</patent> 2005
[6]
IEEE standards. IEEE standard for local and metropolitan area networks — Part 21: Media independent handover services. New York: IEEE, 2009
[7]
LeeJ H, ChungT M, GundavelliS. A comparative signaling cost analysis of hierarchical mobile IPv6 and proxy mobile IPv6. In: Proceedings of the 19th IEEE International Symposium on Personal, Indoor and Mobile Radio Communications. 2008, 1-6
[8]
BaumannF V, NiemegeersI G. An evaluation of location management procedures. In: Proceedings of the 3rd Annual International Conference on Universal Personal Communications. 1994, 359-364
[9]
PackS, KwonT, ChoiY. A performance comparison of mobility anchor point selection schemes in hierarchical mobile IPv6 networks. Computer Networks, 2007, 51(6): 1630-1642
[10]
PackS, ChoiY. Performance analysis of hierarchical mobile IPv6 in IP-based cellular networks. In: Proceedings of the 14th IEEE International Symposium on Personal, Indoor and Mobile Radio Communications. 2003, 3: 2818-2822
RIGHTS & PERMISSIONS
Higher Education Press and Springer-Verlag Berlin Heidelberg
AI Summary 中Eng×
Note: Please be aware that the following content is generated by artificial intelligence. This website is not responsible for any consequences arising from the use of this content.