Impact of Topology Control on End to End Performance for Directional MANETs

Recently, there have been increasing interests in designing topology control algorithms for MANET. However, almost all work focuses on the objective of reducing the transmission power of nodes while preserving the network connectivity. Therefore, it is still an open issue whether the potential benefits of topology control will eventually lead to improvement in network performance, and, if so, which performance metrics may be improved. In this paper, we study the impact of the benefit of reducing broadcast overhead of topology control on network performance by integrating a topology control algorithm with a directional antenna-based MAC protocol for MANET. Simulation results show that, somewhat unexpected, topology control does not significantly improve the throughput; besides, topology control significantly increases the delay. Three possible reasons to explain this are as follows (i) directional antenna inherently reduces interference to the extent that additional improvement enabled by topology control is marginal at best; (ii) any savings on reduced overhead (as a result of having fewer neighbors) might be offset by increased path lengths; (iii) mobility makes it difficult to select and keep "good" neighbors. Therefore, we have doubts on topology control's potential benefits for end-to-end performance, especially for directional antenna-based MANETs