Self-tuned distributed monitoring of multi-channel wireless networks using Gibbs sampler

Wireless side monitoring employing distributed sniffers has been shown to complement wired side monitoring using Simple Network Management Protocol (SNMP) and base station logs, since it reveals detailed PHY and MAC behaviors, as well as timing information. Due to hardware limitations, wireless sniffers typically can only collect information on one channel at a time. Distributed algorithms are desirable to determine the optimal channel allocation of sniffer nodes to maximize the information collected. In this paper, we propose Gibbs sampler based algorithms for robust distributed monitoring of multi-channel wireless networks. Among several variants of the base Gibbs sampling approach, we find that most algorithms suffer from high sensitivity to parameter selection. In contrast, Gibbs sampling using a thermodynamic schedule is self-tuned and can adapt to different network configurations. Simulation studies show that the proposed algorithms can achieve faster convergence rate and have higher chance of reaching global optima than traditional Gibbs