Fuzzy inference based delay and channel aware communication in low-power sensor networks

Traditionally, transmission power is adjusted dynamically to overcome unreliability over lossy links in energy-constrained Wireless Sensor Networks (WSNs). The network node increases its transmission power to achieve immunity against link errors or lowers the power to save energy and prevent interference. Through systematic analysis, we illuminate that it adversely affects the channel contention, network throughput and energy consumption at network scale. Therefore, we implement a novel but effective Fuzzy Inference based Delay and Channel aware Communication (FI-DACC) mechanism at WSN nodes which employs lightweight Forward Error Correction (FEC) codes at low transmission power. Besides selecting an adequate FEC code, the proposed strategy also tunes MAC layer back off duration to prevent packet collisions. To simplify the decision process, a cascaded structure of two fuzzy logic controllers is formulated that undertakes heterogeneous parameters into account. We integrate our solution into IEEE802.15.4 based energy constrained WSN scenarios and evaluate the scheme from real-time packet delivery viewpoint. Results indicate that the proposed approach enhances network real-time capacity with low energy overheads as compared to alternative schemes, for e.g. Real-time Power Routing (RPAR), adaptive power control, hybrid, Interference-aware Transmission Power Control (I-TPC) [4].