Stochastic Learning for the SET K-COVER Problem in Heterogeneous Wireless Sensor Networks

This paper addresses the SET K-COVER problem by proposing a memorial mixed-response algorithm (MMRA) from the perspective of learning in games, where each sensor node is viewed as a non-greedy game player with a limited memory and updates its action following a finite memory and a mixed response rule. We prove that our MMRA converges to a convention of Nash equilibria. Moreover, a tradeoff between solution efficiency and computation time could be achieved via the adjustment of the amount of randomness. Comparison results against typical distributed methods demonstrate the advantage of our methodology.