Reliable congestion-aware information transport in wireless sensor networks

Abstract

Wireless sensor networks (WSNs) constitute the transportation conduit for the results of the in-network processing of the raw data. In WSNs the sensor node and communication level perturbations are often the norm than the exception due to varied environmental conditions. Consequently, the diverse applications supported by WSNs stipulate their individual (and varied) requirements for WSN information transport reliability in order to meet their specific responsiveness needs. The use of an approach that guarantees the highest reliability level for information delivery is not a realistic option as this over-provisioning wastes key resources such as energy or bandwidth. In this paper, we present a new approach termed ReCAIT that targets 'congestion-aware' reliable information transport in WSNs to provide application-specific tunable reliability and thus avoids over-provisioning. To provide tunable reliability, ReCAIT efficiently integrates probabilistic adaptive retransmissions, hybrid acknowledgement and retransmission timer management. ReCAIT proactively alleviates the congestion by transporting information on multiple paths. If congestion persists ReCAIT's back-pressure mechanism triggers the information rate control. Our simulation results show that ReCAIT provides tunable reliability and mitigate congestion, which maximises the efficiency of information transport in terms of reduced number of transmissions.