Globally optimizing network utility with spatiotemporally-coupled constraint in rechargeable sensor networks

Energy harvesting is emerging as a promising technology to extend the lifetime of wireless sensor networks, which is referred to as rechargeable sensor networks (RSNs), This paper is concerned with the network utility maximization problem of static-routing RSNs with limited battery capacity. In specific, to calculate the energy consumption rate, one node's sampling rate is coupled with some other nodes. Moreover, because the energy consumption rate is constrained by the energy harvesting rate and current battery level to avoid depletion or overcharge of the battery, it is also coupled across the time horizon. In this paper, we will globally optimize the network utility. By dual decomposition, we decouple it into separable subproblems, which can be distributively solved without the need to coordinate with other nodes or time horizon. We then propose a distributed algorithm in the context of joint rate and battery control, which converges to the globally optimal solution. Numerical results, based on the real solar data, demonstrate that the proposed algorithm always achieves higher network utility than existing approaches.