Analysis of the Simulated Aggregate Interference in Random Ad-hoc networks

In this paper, we propose a new method for bias correction in the simulation of random wireless ad-hoc networks (WANETs), when the distribution of the node locations is modeled as a Poisson-Point-Process (PPP). The aggregate interference is the main limiting factor in WANETs, and dominates the achievable rate and thus also the network capacity. In the proposed method, a bias correction constant is added to the aggregate interference that is measured in each simulation iteration. The value of the constant is derived through stochastic geometry analysis. We prove that the proposed method can reduce the computational complexity by several orders of magnitude, while producing more accurate simulation results. This improved accuracy is also demonstrated by simulations. As an example, we prove that a bias corrected simulation with only 100 transmitters is sufficient to estimate the aggregate interference with an accuracy of 1%.