Fastest mixing Markov chain on symmetric K-partite sensor networks

In this paper analytical solution of fastest mixing Markov chain problem over a sensor network with K-partite topology is provided. The solution procedure consists of Stratification of sensor network's connectivity graph and semidefinite programming. The studied topologies are evaluated in terms of asymptotic and per step convergence rates. The obtained optimal transition probabilities have been compared with those obtained from Metropolis-Hasting method by comparing mixing time improvements numerically.