Receiver-based Protocol Enhancements for Wireless Ad-Hoc and Sensor Networks

Receiver-based protocols have been proposed as a means of allowing communication when nodes do not maintain any state information. In receiver-based protocols, receivers contend to be the next-hop router of a packet, and the transmitter selects the “best” receiver under a given optimality criteria to become the next hop for transmission. Receiver-based protocols are unique in wireless ad hoc and sensor networks in that they remove the need for costly state maintenance, and they do not require any synchronization of nodes’ sleep-awake schedules for duty cycling. In this thesis, I investigate the advantages of receiver-based protocols in multicast and convergecast communication, which are two important communication patterns in wireless ad hoc and sensor networks. Furthermore, I develop several enhancements to existing receiver-based protocols to improve performance. For multicast communication, I use the RBMulticast protocol, which is a stateless receiverbased multicast protocol that simply uses a list of the multicast members’ (e.g., sinks’) addresses, embedded in packet headers, to enable receivers to decide the best way to forward the multicast traffic. Using simulations, I investigate the performance of RBMulticast in terms of packet delivery ratio, average latency and average overhead, and I develop a new retransmission scheme to enhance the performance of RBMulticast. For receiver-based protocols using convergecast communication, I develop novel duty cycle assignment algorithms. I model the expected energy consumption of nodes utilizing receiver-based protocols as a function of their duty cycle and expected traffic load, which can be derived from their distance to the sink node. Using this analysis, I determine the duty cycle that minimizes the expected energy dissipation for a given node distance to the sink. Moreover, I propose an adaptation for the derived distance-based duty cycle based on local observed traffic. Simulation results indicate that both methods have advantages over a traditional fixed duty cycle approach.