Abstract: To solve the stability issue of clustered mobile ad hoc networks, a weighted clustering algorithm is proposed, based on the theory of the network flow balance. The algorithm chooses cluster heads relying on the adaptive weight, and dynamically clusters the mobile nodes distributed randomly and clinging to the network conditions and nodes’ state. Clusters are optimized according to the scalable degree. Simulation results show that, in the condition of maintaining the former expenses, the algorithm could balance the load of cluster heads, and prevent the cluster heads from becoming the bottleneck effectively.

Key words: ad hoc networks, architecture, clustering algorithm, flow balance, dynamic adaptability