A load-adaptive fair access protocol for MAC in underwater acoustic sensor networks
Wenbo Zhang,Xin Wang,Guangjie Han,Yan Peng,Mohsen Guizani,Jingyi Sun
DOI: https://doi.org/10.1016/j.jnca.2020.102867
IF: 7.574
2021-01-01
Journal of Network and Computer Applications
Abstract:<p>To address time unfairness access problems in the dynamic channel, an underwater adaptive contention window (CW) adjustment backoff algorithm (QL-UACW) based on Q-learning is presented in this paper. That algorithm employs the concept of reinforcement learning to adaptively adjust the competition window by autonomously learning the network communication process. Based on QL-UACW and in order to address space unfairness access issues of nodes location, a load adaptive hybrid media access control (MAC) protocol with competition for the same directional nodes and time slot allocation for different directional nodes called SDHA is also presented. The protocol divides the nodes into co-nodes and counter-nodes according to the relative communication directions of the nodes. The co-nodes use competition protocol and the counter-nodes use time slot allocation protocol to solve the hidden terminal problem of node communication in order to prevent space access unfairness. Theoretical and simulation analysis show that the QL-UACW algorithm effectively improves the fairness of node access channels, reduces the collision rate of data frames, and increases network throughput. Additionally, the SDHA protocol is proven to provide advantages in normalized network throughput, average end-to-end delay, and the success rate of data frame transmission.</p>
computer science, interdisciplinary applications, software engineering, hardware & architecture