Game Theoretic Power Allocation for Sparsely Distributed Clusters of Wireless Sensor Networks

Sparsely distributed clusters of wireless sensors with applications ranging from state-wide wireless bridge monitoring to animal herds tracking have gained recent attention due to their special characteristics and challenging resource allocation problems. A novel optimal power allocation paradigm based on game theory is proposed to minimize intra-cluster interference. Cost functions are assigned to the transmission power, estimate of the aggregate interference generated by the cluster on the overall network based on worst case interference estimates, and the signal to interference and noise ratio at the node level. The algorithm introduced in this thesis, called Game Theoretic Power Allocation in Sparsely Distributed Clusters of Wireless Sensors (GPAS), expands the existing research on the Greedy Asynchronous Distributed Interference Avoidance algorithm (GADIA) [1] by optimizing power allocation in addition to channel usage. With the addition of power management using the GPAS algorithm, sparsely distributed clusters use less power and avoid interfering with each other. This increases the life of nodes on the mote level and increases the ability of clusters to communicate by reducing otherwise wasted broadcasting power. This allows clusters to achieve a signal to interference and noise ratio that is suitable, while not flooding the limited resources of the channel