An adaptive redundancy protocol for mesh based multicasting

Abstract

Mesh-based multicast routing protocols for multi-hop wireless ad hoc networks (MANETs) build multiple paths from senders to receivers. Higher redundancy results in higher reliability because packets can be delivered even in the presence of links breaking. However, in less dynamic environments, in which links break less frequently, the additional redundancy may not be needed in terms reliability and may significantly increase overhead. This paper investigates the tradeoffs between reliability and efficiency in mesh-based multicast protocols for MANETs. We introduce an adaptive mesh-based multicast mechanism that controls mesh redundancy based on link reliability in the neighborhood of the node. Mesh redundancy is measured by the number of paths from each receiver to the core of the group’s mesh, which serves as the address of the group. We introduce a metric called Mesh Reliability Index (MRI), which allows nodes to estimate the reliability of the mesh in their neighborhood, and determine whether redundancy needs to be increased or decreased. Through simulations, we compare the performance of the adaptive mesh-building protocol against the non-adaptive version (which builds the mesh with maximum redundancy), and against ODMRP for a wide range of scenarios with varying mobility, group members, number of senders, traffic load, number of multicast groups and terrain size. Our results show that adjusting mesh redundancy based on link reliability can maintain high packet delivery ratios with less overhead compared to non-adaptive mesh-based multicast protocols.