Joint energy allocation for sensing and transmission in rechargeable wireless sensor networks

• Main Author: Mao, Shaobo
• Language: English
• Published: University of British Columbia 2012
• Online Access: http://hdl.handle.net/2429/42832

The energy management policy of a rechargeable wireless sensor network (WSN) needs to take into account the energy harvesting process, and is thus different from that of a traditional WSN powered by non-rechargeable batteries. In this thesis, we study the energy allocation for sensing and transmission in an energy harvesting sensor node with a rechargeable battery. The sensor aims to maximize the expected total amount of data transmitted subject to time-varying energy harvesting rate, energy availability in the battery, data availability in the data buffer, and channel fading. In this thesis, we first consider the energy allocation problem that assumes a fixed sensor lifetime. Then, we extend the energy allocation problem by taking into account the randomness of the senor lifetime. In the first part of this thesis, we study the joint energy allocation for sensing and transmission in an energy harvesting sensor node with a fixed sensor lifetime. We formulate the energy allocation problem as a finite-horizon Markov decision process (MDP) and propose an optimal energy allocation (OEA) algorithm using backward induction. We conduct simulations to compare the performance between our proposed OEA algorithm and the channel-aware energy allocation (CAEA) algorithm extended from [1]. Simulation results show that the OEA algorithm can transmit a much larger amount of data over a finite horizon than the CAEA algorithm under different settings. In the second part of this thesis, we extend the joint energy allocation problem by taking into account the randomness of the sensor lifetime, and formulate the problem as an infinite-horizon discounted MDP. We propose an optimal stationary energy allocation (OSEA) algorithm using the value iteration. We then consider a special case with infinite data backlog and prove that the optimal transmission energy allocation (OTEA) policy is monotone with respect to the amount of battery energy available. Finally, we conduct extensive simulations to compare the performance of the OSEA, OTEA, and CAEA algorithms. Results show that the OSEA algorithm transmits the largest amount of data, and the OTEA algorithm can achieve a near-optimal performance.