Intelligent Multicast Tree Construction Protocol with Optimal Bandwidth Allocation for WSNs

摘要

This paper addresses the problem of multisession multicast tree construction with bandwidth and rate allocation in wireless sensor networks. Previous work has shown that when the goal is to find multicast routing tree, the problem becomes NP-complete. In this work, we present a heuristic Multisession Multicast Routing and Bandwidth Allocation protocol, termed MMBA, that makes use of fuzzy logic to evaluate nodes' and network conditions during the multicast tree construction process. Rate assignment is optimized in order to be able to accept as many data streams (i.e., sessions) at the highest possible data rate in the sensor network as possible, allowing source nodes to transmit at maximum available rate, while maximizing the overall network throughput and utility. We conduct extensive evaluations to study the performance of the proposed protocol compared to existing approaches such as shortest path, Steiner and minimum transmission tree. Simulation results show that our protocol effectively improves the network throughput and utilization, while conserving per node energy consumption.