Kalman Filtering Based Adaptive Transfer in Energy Harvesting IoT Networks

In this paper, we investigate an energy prediction algorithm based on Kalman filtering in energy harvesting IoT networks. The IoT nodes harvest renewable energy from nature and powered by green energy only. Owing to the space-time instability and non-uniformity of renewable energy, the IoT nodes may have insufficient energy supply. An unresolved challenge is accurately predicting the available renewable energy, and developing low complexity solutions that incorporate a lossless transfer guarantee. With this in mind, we propose the energy prediction algorithm based on Kalman filtering to bridge the gap between lossless transfer and unstable renewable energy. The energy prediction is performed at the access point in order to dynamically adjust the number of bits to be sent, and the data loss due to receiver energy depletion will be improved. In addition, real solar and wind energy profiles are exploited by simulations. The simulation results show that the proposed energy prediction algorithm can improve transmission efficiency in terms of the rate of drop bits and the number of time slots needed to transmit a given payload, and reduce the wastes of harvested renewable energy.