An Evolutionary Game Theoretic Approach for Stable and Optimized Clustering in VANETs

Discovering and maintaining efficient routes for data dissemination in vehicular ad hoc networks (VANETs) has proven to be a very challenging problem. Clustering is one of the control protocols used to provide efficient and stable routes for data dissemination. However, the rapid changes in network topology in VANETs creates frequent cluster reformation, which can seriously affect route stability. We propose a novel evolutionary game theoretic (EGT) framework to automate the clustering of nodes and nominations of cluster heads, to achieve cluster stability in VANETs. The equilibrium point is proven analytically and the stability is also tested using Lyapunov function. The performance of the proposed evolutionary game is empirically investigated with different cost functions using static and mobile scenarios. The simulation results demonstrate the effectiveness and robustness of our proposed EGT approach for different populations and speeds, thus reducing the overhead of frequent cluster reformation in VANETs.