Fuzzy-based probabilistic broadcasting in mobile ad hoc networks

Broadcasting in Mobile Ad Hoc Networks is a fundamental process for nodes to maintain network connectivity and stability, providing support to management procedures in order to exchange control information and aiding routing protocols to keep up to date records of routes to all possible destinations. The most commonly used broadcast scheme in MANETs is Simple Flooding (SF) due to its simplicity. SF is found to be neither resource efficient nor reliable, especially in dense and highly populated networks. Various adaptive schemes have been proposed to alleviate the effects caused by SF that take into account network conditions and other limitations. Adaptivity is most commonly implemented with the support of hello protocols. It is critical to decide upon the best frequency value to be used by the hello protocol in order to avoid introducing unnecessary control overhead. In this paper, we demonstrate the need for a hello protocol with frequency adaptivity compared to either fixed or plain dynamic frequency implementations as part of an adaptive probabilistic broadcast scheme. Furthermore, we introduce the use of Fuzzy Logic to control the frequency of a hello protocol and thus to achieve optimum results for a novel broadcast algorithm that bases its rebroadcast decision on a probability model. Our simulation results clearly demonstrate the superiority of our proposed scheme when Fuzzy Logic is incorporated by the hello protocol used to implement adaptivity against fixed and dynamic frequency alteration approaches.