QoS-aware distributed adaptive cooperative routing in wireless sensor networks

In this paper, we investigate a cooperative routing problem in time-varying Wireless Sensor Networks (WSNs) targeting the achievement of quality-of-service guarantees in delay and reliability domains. We develop a distributed adaptive cooperative routing protocol, called DACR, that exploits cooperative communication on top of delay- and energy-aware end-to-end routes and optimizes the trade-off between the reliability and delay through Lexicographic Optimization at each hop. We employ a lightweight reinforcement learning method to update the routing nodes with knowledge of expected performances that could be provided by the candidate relay nodes, helping to determine the optimal relay with the least overhead. The decision of selecting a transmission mode (i.e., direct or relayed transmission) at each hop is taken adaptively so that the reliability is maximized. The performances of our DACR have been evaluated through ns-2 simulations for a wide range of link failure rates and data traffic generation rates and the results show that the DACR outperforms a number of state-of-the-art protocols.