Transmission Scheduling Scheme Avoiding Multi-Packet Excessive Interference in Underwater Acoustic Sensor Networks

Scheduling-based Medium Access Control (MAC) protocols significantly augment the ability for simultaneous transmissions while mitigating signal interference, thereby elevating the bandwidth utilization efficiency crucial for data-collection-oriented underwater acoustic sensor networks (UASNs). However, existing scheduling-based MAC protocols for UASNs primarily analyze the interference between pairwise links or nodes, neglecting the cumulative interference effect fostered by the concurrent transmissions of multiple nodes. Besides, the existing multi-node interference models used in Radio Frequency (RF) based wireless networks only describe the total strength of cumulative interference signals and do not consider scenarios in which the interference from multiple nodes does not align at the receiver. This paper introduces an innovative interference model to extensively quantify the non-aligned overlapping interference from multiple packets on a single valid packet in segments. Based on this segmented interference model, we establish a transmission time constraint that can avoid excessive interference in high-interference segments and utilize low-interference segments for parallel reception, thereby enhancing the degree of time reuse. The optimization of transmission slot scheduling is modeled as a sequential decision-making process, wherein the action space is delineated by considering the constraints related to multi-packet interference. Furthermore, the Reinforced Packet Level Slot Scheduling (R-PLSS) algorithm leveraging the principles of approximate dynamic programming is proposed to allocate transmission slots within each frame. The simulation results demonstrate that the R-PLSS algorithm avoids the practical packet demodulation failure caused by excessive interference from multiple packets and can comprehensively improve the channel utilization efficiency.