On Hopping Strategies for Autonomous Wireless Networks

The transmission capacity (TmC) of an ad hoc network measures the area spectral efficiency (in bits/sec × Hz×m²) of successful transmissions as a function of the required transmission rate in single-hop links ¿, under the assumption that the average density of active links ¿_at in the network is given. In reality, however, the probability that a node wishing to transmit becomes active is conditioned on the availability of a receiving peer. Consequently, ¿_at is not a given parameter but rather a function of topological parameters. In this paper, we employ stochastic-geometric tools to obtain an expression of ¿_at as a function of the transmission range d and network density ¿, and apply the result to evaluate the transmission capacity of autonomous interference-limited networks. We then use the TmC to study the impact of closest-neighbor, furthest-neighbor and random-neighbor hopping strategies on the performance of such networks. It is shown that amongst these alternatives, the closest-neighbor strategy always achieves the highest transmission capacity. Furthermore, it is found that the advantage of closest-neighbor hopping is more significant in networks with high densities, large transmission ranges and/or higher required rates, where interference is the dominant limiting factor. In noninterference-limited networks, however, the three strategies are equivalent.