Routing for Opportunistic Networks Based on Probabilistic Erasure Coding

We investigate the problem of routing in opportunistic networks and propose a novel routing algorithm, which combines probabilistic routing with erasure coding. Erasure coding generates large amounts of code blocks with fixed overhead; we apply a sophisticated but realistic method to allocate the generated code blocks to nodes that relies on a probabilistic metric for evaluating node potential. Our goal is to enhance further the robustness of the erasurecoding based forwarding in worst-case delays but also in small-delay scenarios. We detail our algorithm and evaluate its performance against other well-known routing algorithms. We exploit scenarios with adequate resource storage as well as scenarios with limited storage capacity. In both cases, our algorithm yields promising results.