MPR-based broadcasting in ad hoc and wireless sensor networks with a realistic environment

Summary Most existing broadcasting protocols for ad hoc and wireless sensor networks use the standard Unit Disk Model (UDM) to represent the physical layer. Whenever this model does not take into account the fluctuations of radio signal, these protocols must be modified to adapt to transmissions in a real environment. In this paper, we apply the Log-Normal Shadowing Model (LNSM) to represent a realistic simulation environment and focus our study particularly on the performance of broadcasting Multipoint Relay Protocol (MPR). Thus, using LNSM we conduct extensive simulations to illustrate the effects of radio signal fluctuations on protocol performance. Unfortunately, our findings show that fluctuation presence has a significant impact on protocol performance. Hence to improve these performances, within this framework we propose two schemes: the first one optimizes the probability of successful reception by the selected neighborhood as relay nodes, and the second uses one probability threshold to select the relay nodes and another to maximize the probability of correct reception by two-hop neighbors of the source node. Finally, simulation results are presented, showing that our schemes provide a better performance over the ideal model.