Jointly Power Control and Scheduling in Underwater Wireless Sensor Networks With Oblivious Freshness Requirement

This paper considers an underwater wireless sensor network where a sink station collects time-sensitive information from multiple sensors. For timely monitoring, the sink station aims to maximize the data freshness of the entire network. The difficulty includes time-varying channel states, limited transmission bandwidth and power consumption caused by underwater acoustic communication. Moreover, due to the time-varying service requirements, the importance of data is unknown until it is received and processed by the sink station. To overcome these difficulties, we characterize the data freshness at the terminal through a set of non-decreasing functions with respect to the popular metric Age of Information (AoI). To save the energy consumption, each sensor will transmit with different power to combat the different channel states. Then, we relax the bandwidth constraint and resort to the online learning framework with Lyapunov drift analysis to design a jointly scheduling and power control algorithm based on historical observations. The algorithm is proven to achieve the sub-linear expected performance for both cumulative age regret and bandwidth violation. Finally, we propose the truncated scheduling strategy to satisfy the hard bandwidth constraint. Simulation results validate the performance of the proposed algorithms compared with the optimal offline algorithm with complete information.