On mobility-capacity-delay trade-off in wireless ad hoc networks

We show that there is a trade off among mobility, capacity, and delay in ad hoc networks. More specifically, we consider two schemes for the mobility of nodes in ad hoc networks. We divide the entire network into cells whose sizes can vary with the total number of nodes or whose size is independent of the number of nodes. By restricting the movement of nodes within these cells, we calculate throughput and delay for randomly chosen pairs of source-destination nodes, and show that mobility is an entity that can be exchanged with capacity and delay. We also investigate the effect of directional antennas in a static network in which packet relaying is done through the closest neighbor, and verify that this approach attains better throughput than static networks employing omnidirectional antennas.