Channel Assignment Scheme based on Learning Automata for Clustered Wireless Ad-Hoc Networks
Shiva Shabanzadeh Dehkordi, Javad Akbari Torkestani

In this paper we have design a learning automata-based scheme for medium access scheduling in clustered wireless ad-hoc networks. In this scheme, the collision-free intra-cluster communications are organized by the cluster-heads using learning automata rules. The advantage of applying learning automatons in this scheme is that each cluster-head learns the traffic parameters of its own cluster members. Each cluster-head monitors the intra-cluster transmissions and coordinates these transmissions to avoid collisions. In the proposed polling scheme, a portion of bandwidth is assigned to each cluster member Commensurate with its need such as traffic load. The simulation results show that the proposed polling scheme outperforms the existing methods in term of almost all metrics of interest.

Full Text: PDF