On the performance of asynchronous ad-hoc networks using interference alignment with renewal process

The performance of ad-hoc networks is greatly affected by interference, particularly interference from nearby nodes. Based on stochastic geometry, this work studies a form of multiple-input-multiple-output (interference alignment (IA) that eliminates interference from transmitters within a range and treats the remaining interference as a shot noise process. Adapting to the bursty nature of ad-hoc network traffic, the authors introduce a novel distributed ad-hoc network with renewal process, whereby cooperation between interferer and receiver is implemented in one-way process, which results from the difficulties for the node to change the beamforming or receiving matrix whereas communicating with a proposed transmitter. In addition, this scheme takes advantage of IA, leaving the desired signal with higher degree of freedom (DoF) compared with the partially zero-forcing method. Outage probability and transmission capacity are derived with and without Rayleigh fading. Simulation results show the effect of IA in ad-hoc network, including the incremental in the DoF of the desired signal as time goes by. Monte Carlo simulations are implemented and show that IA outperforms successive interference cancellation by at least 10 dB in terms of the signal-to-interference ratio with large path-loss exponent.