On a moving direction pattern based MAP selection model for HMIPv6 networks

Abstract

In a large-scale mobile IPv6 network, usually there are several coexisting mobility anchor points (MAPs) for networking robustness and traffic sharing. Therefore, it is a challenging issue for an arriving mobile node to choose the most appropriate MAP to bind. This task must be carried out by considering the issues of load balancing, binding update and packet delivery cost minimization. This paper proposes a novel MAP selection scheme for hierarchical mobile IPv6 networks to allow a mobile node to discover the most appropriate MAP when there are multiple coexisting MAPs. This scheme is an enhancement to the adaptive MAP selection scheme. The proposed scheme improve the overall performance due to the consideration of the movement (direction) pattern of mobile nodes. Simulation results show that this scheme outperforms the existing cost models in terms of total binding update and packet delivery costs, ensuring a level of load balance similar to adaptive MAP selection scheme.

Introduction

Mobile communication has become an indispensable part of our daily life due to the widespread use of portable computers and handheld devices, such as PDAs and smart phones, etc. While roaming between different IP networks, mobile users access the Internet to retrieve emails and other information, and communicate with people via video conferencing. Recently, a growing number of mobile applications require mobile users to transmit multimedia data in wireless environments at a relatively high transmission rate.

In order to transmit time-sensitive or real-time multimedia data in a mobile wireless environment, seamless handoff must be provided. In order to facilitate mobile roaming between different wireless or wired access networks, various mobile IP (MIP) protocols were introduced by the Internet Engineering Task Force (IETF). In the mobile IPv4 (MIPv4) protocol [1], [2], [3], [4], registration can be done in some specialized routers such as foreign agents (FAs) or home agents (HAs). On the other hand, mobile IPv6 (MIPv6) [5] does not require dedicated routers to act as FAs. Both protocols allow a mobile node (MN) to maintain a permanent IP address (viz., the home address) while visiting different networks. The HA redirects packets to the MN while the MN is away from its home network. Whenever an MN changes its point of attachment, both protocols require the MN to update its HA with its new location and all correspondent nodes (CNs) communicating with the MN. Even if the MN roams between subnets within the same domain, the MN must send a binding update (BU) to the HA, which usually resides far from the location of the MN.

To reduce the signaling overhead caused by the handover process, the concept of mobility anchor point (MAP) was introduced in the hierarchical MIPv6 (HMIPv6) protocol [6] and its functionality was extended in many follow-up studies [7], [8]. An MAP serves as a local HA in a foreign network. Whenever an MN moves to a new subnet within the same domain, it sends a BU to the MAP, rather than the HA. This reduces the signaling overhead significantly, since the MAP is usually much closer to the MN than the HA. Within a large-scale network infrastructure, there are usually several coexisting MAPs to improve robustness and enable traffic sharing. As a result, in a given MAP domain, an arriving MN has the task of determining and binding to the most appropriate MAP. When the MN must select the proper MAP in a given domain, selection stability, load balancing among MAPs, BU and packet delivery (PD) costs are important criteria for achieving the optimal performance.

While acknowledging the previous schemes [11], [12], [13], [14], [15], [16], [17], [18], [19] which were found to be efficient, we need to point out that they only focused on one or two influencing factors for determining the most suitable MAP. As a result, many previous schemes work effectively only in some specific cases. However, in general, there might exist some more suitable MAPs than those proposed by the aforementioned MAP selection procedures.

The contributions of this paper are summarized as follows. We propose an improved MAP selection scheme, which is an enhancement to the most recently reported scheme reported in [15]. The proposed scheme can discover a better MAP in terms of total BU and PD costs, while still maintaining the MAP load balance and outperforming the existing schemes. Compared with the existing methods, the proposed scheme takes into account almost all important factors which have impact on the MAP selection process, including (1) the handoff frequency of an MN; (2) the direction of movement of an MN; (3) the expected BU costs (depending on the selected MAP); and (4) the expected packet delivery costs (depending on the selected MAP). In particular, we consider the movement patterns of an MN in this paper. These considerations affect the processing load of an MAP and selection stability.

The remainder of this paper is outlined as follows. Section 2 discusses the related works and the problems with existing MAP selection schemes. Section 3 proposes our MAP selection scheme as an effort to improve the existing schemes. In Section 4, we investigate the proposed scheme via analysis and show the performance of the proposed scheme, followed by the conclusions and future works given in Section 5.