Throughput - and Cost - Aware Node Relocation for MANET Resiliency Under Jamming Attacks

As the use of Mobile Ad Hoc Networks (MANETs) increases, so does the number of cyber attacks to disrupt their operation. In particular, jamming attacks can prevent critical network nodes to communicate and relay information, which seriously disrupts network performance. In this paper, we propose Reactive Sequential Relocation (RSR), an efficient relocation strategy to mitigate the impact of jamming attacks by physically moving critical nodes in the network outside the attacked area, optimizing throughput, network resiliency, and relocation cost (i.e., relocation distance). We formulate an associated joint relocation problem Reactive Joint Relocation (RJR), show that it is NP-hard (non-deterministic polynomial-time hardness), and introduce a heuristic solution in RSR. Extensive simulation studies show that RSR solves the RJR problem efficiently while simultaneously leveraging centrality measures, maximum global throughput, and mean distance to generate the most optimal final relocation outcome. Compared to baseline methods such as Random Relocation (RR) and RJR, RSR improves the maximum global throughput by up to 206% on average while decreasing the optimization runtime by up to fivefold and decreasing the average relocation distance by up to 51%.