Performance of neighbor discovery algorithms in mobile ad hoc self-configuring networks with directional antennas

Abstract

One of the key performance measures for a neighbor discovery algorithm in military applications with directional antennas is how long it takes to discover the neighbors given that nodes do not have priori knowledge about where their neighbors are. One general impression in neighbor discovery is that algorithms based on all directional (directional transmission and directional receive) takes longer time to discover all neighbors than those algorithms partially or entirely based on omni-directional antennas. In this paper, we prove that this may not be the case. We first describe a generic random neighbor discovery algorithm (GRA), which uses omni-directional antennas for reception. Then we derive analytical expressions for the average number of slots required to discover all (or any one of) the neighbors for GRA and for an all-directional, scan based algorithm (omni-directional antennas are not used at all). Analytic and simulation results are presented to compare the two algorithms in terms of the expected number of slots required to discover any one or all of the neighbors. Based on the numerical results, we conclude that scan-based all-directional algorithm gives better performance than the algorithm which uses omni-antennas. If designed properly, all-directional neighbor discovery algorithms can discover neighbors in less time than those algorithms that use omnidirectional antennas. The advantages and disadvantages of each algorithm are discussed as well.