Demand-scalable geographic multicasting in wireless sensor networks

In this paper, we focus on the challenge of demand-scalable multicast routing in wireless sensor networks. Due to the ad-hoc nature of the placement of the sensor nodes as well as the variations in the available power of the nodes, centralized or stateful routing schemes are not applicable. Thus, in this paper, we first introduce a Geographic Multicast routing Protocol (GMP) for wireless sensor networks.¹ The protocol is fully distributed and stateless. Given a set of the destinations, the transmitting node first constructs a virtual Euclidean Steiner tree rooted at itself and including the destinations, using a novel and highly efficient reduction ratio heuristic (called rrSTR). The simulation results on NS2 show that GMP requires 25% less hops and energy than the existing Position Based Multicasting, PBM, Location-Guided Steiner trees, LGS, approaches. The GMP algorithm as well as LGS and PBM all assume that each recipient receives the same copy of the multicast message. In reality, however, especially when the transmission includes streamed media, different recipients have different demands (in terms of the frequency of packets or the quality of media). Thus, in this paper, we investigate the suitability of the geographic multicasting schemes for situations where scalable transmission paths can save power. In particular, we propose intuitive mechanisms to extend the three schemes to cases where the data transmission can scale based on the demand. This leads to three new weighted multicast routing algorithms: wGMP, wLGS, and wPBM. The results show that the wGMP algorithm provides the best opportunities for scalability due to its flexible self-correcting decision making process, while other schemes, such as wLGS and wPBM are not directly suitable for scalable multicasting, due to their naively greedy structures.