In distributed wireless sensor networks (WSNs), the energy restriction of sensor nodes is always the bottleneck and limits the network lifetime. Emerging energy harvesting techniques can be potential solutions. The energy obtained from the renewable energy source, on the other hand, is not stable. In this paper, we consider the coefficients estimation problem in a WSN with renewable energy source. The limited and unstable power supply prevents sensor nodes from transmitting all collected data to the fusion center. We develop an optimal data censoring strategy for the estimation problem. Firstly, the behavior of the incoming renewable energy source is learned from historical data through Long Short Term Memory (LSTM) networks. Then the amount of available energy for the next period is predicted. The optimal data censoring strategy is developed and updated based on the available energy. We show that the proposed algorithm can fully enjoy the benefit of renewable energy with extended network lifetime as well as satisfactory estimation performance.
Data Censoring in Renewable Energy Enabled Wireless Sensor Networks
2019-09-01
646824 byte
Conference paper
Electronic Resource
English
Fusion of Decisions in Multi-hop Wireless Sensor Networks with Three-Level Censoring Scheme
Springer Verlag | 2014
|Sensor Fusion with Censoring Limits
AIAA | 2020
|SENSOR FUSION WITH CENSORING LIMITS
TIBKAT | 2020
|Fusion of Decisions in Multi-hop Wireless Sensor Networks with Three-Level Censoring Scheme
British Library Conference Proceedings | 2014
|