GODDeS: Globally ∈-Optimal Routing Via Distributed Decision-theoretic Self-organization

摘要: This paper introduces GODDeS: a fully distributed self-organizing decision-theoretic routing algorithm designed to effectively exploit high quality paths in lossy ad-hoc wireless environments, typically with large number of nodes. The problem is modeled as an optimal control for decentralized Markov Decision Process, links characterized by locally known packet drop probabilities that either remain constant on average or change slowly. equivalence this optimization performance maximization explicitly constructed probabilistic automata allows us apply the theory quantitative measures regular languages, and design highly efficient solution approach attempts minimize source-to-sink across network. Theoretical results provide rigorous guarantees global performance, showing achieves near-global optimality, polynomial time. It also argued GODDeS significantly congestion-aware, exploits multi-path routes optimally. development supported high-fidelity network simulations.