A Geometric Model to Reduce Interference Computations in Simulation of Large Ad hoc Networks

Abstract

Simulation of wireless networks has recently drawn considerable attention in the area of abstraction. Simulation is a precious tool used to model complex systems where the desired network size is large in scale. For large mobile ad hoc networks, the most computation intensive tasks in simulation are computing interference and determining which receivers are in range of a transmitter. In both cases, O(N2) physical layer calculations are required for a wireless system of N nodes, which scales poorly. In this paper, a geometric model, minimal rectangular coverage area is devised for the optimization of the complexity of interference computations in simulations of wireless mobile ad hoc networks. This method lessens the number of unaffected nodes by considering less area as affected by transmission range, which exists outside the transmission range of a transmitter. However, the experimental results suggest that this geometric model reduce the affected coverage area 12.5% - 78.15% than existing grid based algorithm used in current version of NS2. This paper also discusses about the efficiency considerations of the algorithm in detail.