Modeling and performance reliability assessment of cluster-based mobile Ad Hoc Networks

Mobile Ad-hoc Networks (MANETs) with a conventional flat network architecture are constrained in communication efficiency, and cluster-based network models have untapped potential to solve this problem with their information aggregation and long-distance relay capabilities. Aware of the potential, this study establishes a cluster-based MANET (C-MANET) model and proposes a corresponding performance reliability assessment framework. Specifically, the modeling process includes 1) weighted probability-based cluster head selection, 2) leadership-driven clustering, and 3) cluster update. Subsequently, a link stability-oriented reliability assessment framework for the proposed C-MANET model is developed in three steps: 1) evaluate the probability of path stability, 2) assess the rate of transmission success, and 3) analyze the reliability of clusters. The modeling method and the assessment framework are validated in a case of unmanned vehicle clusters, yielding three key findings: 1) path stability probability drops over time at a decreasing rate; 2) the quantity of member nodes demonstrates greater impact on path stability than the quantity of cluster heads during inter-node communication; 3) cluster head communication range and cluster head ratio significantly influence both intra-cluster and inter-cluster performance reliability. This research provides a systematic methodology for modeling C-MANETs and evaluating their performance reliability.