Considerations of IP multicast for load balancing in proxy mobile IPv6 networks

Nguyen, Tien-Thinh; Bonnet, Christian
Computer Networks, 30 July 2014, Elsevier

Proxy Mobile IPv6 (PMIPv6), taking advantages of the network-based mobility management, enables mobility support for the mobile nodes (MNs) without requiring their involvement in mobility signaling. However, as a centralized mobility management, PMIPv6 relies on a central mobility entity, i.e., Local Mobility Anchor (LMA) to provide the mobility support. The LMA is responsible for maintaining the mobile node's (MN) reachability state and forwarding the traffic from/to the current location of the MN. As mobile traffic demand rapidly increases, it is easy to make the LMA a bottleneck and a single point of failure. Therefore, load balancing (LB) mechanism among LMAs is a promising solution for these issues. Although previous studies proposed several solutions for distributing the load among the LMAs, none of them considers the multicast service. From the fact that the multicast service is expected to be widely used for delivering multimedia traffic (which will account for the majority of mobile traffic), it can also be considered as a crucial load factor. As a result, the efficiency of the existing solutions may be degraded when considering multicast. Furthermore, applying the existing LB mechanisms can raise several issues for not only the ongoing unicast sessions but also the multicast ones. To tackle these issues, this paper proposes a new LB solution which mainly focuses on the multicast service. The experiments and the numerical results show that this solution helps to better distribute the load among the LMAs while greatly reducing the multicast service disruption as well as avoiding the influence on the ongoing unicast sessions. In addition, the proposed solution can co-operate with the existing proposals to improve the performance of the network.

Communication systems
Eurecom Ref:
© Elsevier. Personal use of this material is permitted. The definitive version of this paper was published in Computer Networks, 30 July 2014, Elsevier and is available at :