On adaptive gain tuning for wireless Ad Hoc networks

We have proposed an adaptive gain scheme for resource allocation in wireless ad hoc networks based on single neural node and virtual price concept. The degree of congestion on a wireless link can be symbolized an aggregate virtual price along a route, which depends mostly on the wireless nodepsilas queue occupancy. This approach adopts an adaptive approach to control the transmission rate at the source, named ASR controller. Our object is to optimize the resource allocation dynamically through the transmission rate adjustment and stabilized the queue size of wireless links. The basic control theory approach for the algorithm is been presented. The performance of simulation indicates that the proposed scheme can achieve the goal of optimal channel capacity sharing among multi-hop flows and sharing bandwidth fairly.