A Genetic-Based Bayesian Framework for Stateless Group Key Management in Mobile Ad Hoc Networks

Abstract

This paper addresses the issue in managing a group key among dynamic group of nodes in mobile ad hoc networks (MANETs), where the participants frequently miss the group key update, commonly known as rekeying. In this paper, we propose a broadcast stateless and distributed group key management (GKM) framework: genetic-based Bayesian networks group key agreement (GBKA) scheme, for supporting dynamic rekeying mechanism in MANET. The proposed framework includes two main functionalities: First to initialize the construction of group key and second is to optimize the keys to maintain the group communication with less overhead, whenever membership changes. The genetic algorithm selects individuals with high genetic characteristics that reflect the node’s ability to adapt and survive the MANET environments. The rekeying scheme operates on genetic computation over finite area with Bayesian network modelling. Whenever the group membership changes, the public messages are updated accordingly with an unchanged private key. Our proposed scheme is very efficient with hash and Lagrange interpolation polynomial implementation. The simulation results show that the proposed group key management scheme achieved higher performance and security compared with other existing key management schemes.