Multi-hop wireless networks, such as ad-hoc and mesh networks, suffer from inherent topology dynamics due to unstable wireless links and node mobility. Stable addressing, as needed for reliable routing, in such evolving, challenging network conditions is thus a difficult task. Efficient multi-hop wireless communication in these networks then requires a fully decentralized, scalable routable addressing scheme that embraces network dynamics and dynamically recovers from failures. This paper explores the feasibility and limits of Mobile Probabilistic Addressing (MPA), a novel addressing approach in IEEE 802.11-based multi-hop wireless networks. MPA is based on a probabilistic addressing paradigm that derives statistical distributions of hop distances between nodes to i) assign fuzzy routable regions to nodes instead of discrete addresses, and ii) provide a distributed storage service to store and retrieve node addresses. We evaluate MPA in simulation and in an 802.11 wireless mesh network of 51 nodes. Our results highlight the graceful topology maintenance and recovery of MPA in challenging networking conditions due to node mobility and unstable link conditions. Precisely, we observe that, when compared with the state-of-the-art, our proposed mechanism achieves an order of magnitude fewer address changes in the network translating into less overhead traffic and high packet success.
展开▼
