Joint Compression, Channel Coding, and Retransmission for Data Fidelity With Energy Harvesting
Chiara Pielli,Čedomir Stefanović,Petar Popovski,Michele Zorzi,Cedomir Stefanovic
DOI: https://doi.org/10.1109/tcomm.2017.2785323
IF: 6.166
2018-04-01
IEEE Transactions on Communications
Abstract:We consider a monitoring application where sensors periodically report data to a common receiver using time division multiplexing. The sensors are constrained by the limited and unpredictable energy availability provided by energy harvesting (EH), and by the channel impairments. To maximize the quality of the reported data, the packets transmitted contain newly generated data blocks together with up to $r - 1$ previously unsuccessfully delivered ones, where $r$ is a design parameter. These data blocks are compressed, concatenated, and encoded with a channel code. The scheme applies lossy compression, such that the fidelity of the individual blocks is traded off with the reliability provided by the channel code. We show that the proposed strategy outperforms the one in which retransmissions are not allowed. We also investigate the tradeoff between the value of $r$ , the compression and the coding rates, under the constraints of the energy availability, and, once $r$ has been decided, use a Markov decision process (MDP) to optimize the compression/coding rates. Finally, we implement a reinforcement learning algorithm, through which devices can learn the optimal transmission policy without knowing a priori the statistics of the EH process, and show that it indeed reaches the performance obtained via MDP.
telecommunications,engineering, electrical & electronic