Opportunistic packet scheduling and routing in wireless mesh networks

Opportunistic routing has emerged as a promising approach to cope with unreliable and unpredictable wireless links in static multi-hop wireless networks. Most existing opportunistic schemes force strict timing between the potential forwarders in order to reduce the number of redundant packet copies generated during the forwarding process. Alternative opportunistic schemes have been also proposed, which adopt a randomized forwarding process to better exploit path diversity. However, both approaches suffer from performance degradations when the number of flows increases. To address this issue, in this paper we describe PacketOPP, a novel opportunistic routing protocol that combines randomized opportunistic forwarding with opportunistic packet scheduling to improve the support of multiple simultaneous flows. Specifically, PacketOPP implements a packet scheduler to award a higher priority to the packets that are expected to deliver the maximum opportunistic gain. Using extensive ns-2 simulations we show that PacketOPP outperforms state-of-the-art non-scheduled opportunistic routing protocols, such as ROMER and MaxOPP, as well as a traditional shortest path routing protocol.