Many-to-one deliverability of greedy routing in 2-D wireless sensor networks

In this paper, we study deliverability of greedy routing in wireless sensor networks, where nodes are distributed over a disk area according to a homogeneous Poisson point process. In our work, we model the level of deliverability of a sensor network as the probability that all sensor nodes can successfully send their data to a base station, which is named probability of guaranteed delivery. We study the relationship between the critical transmission power of sensor nodes and the probability of guaranteed delivery, such that when all sensor nodes transmit with a higher power than the critical transmission power, the sensor network can reach the desired probability of guaranteed delivery. We identify two very tight analytical upper bounds on the critical transmission power for the idealistic u-disk model and the realistic log-normal shadowing model respectively. The correctness and tightness of these two upper bounds are verified by extensive simulations.