Power optimization for multiple QoS, delay, and BER classes relying on finite-delay information theory

Chen Gong,Qian Gao,Lajos Hanzo,Zhengyuan Xu
DOI: https://doi.org/10.1007/s41650-017-0003-0
2017-01-01
Abstract:Future communication systems will include diffierent types of messages requiring diffierent transmission rates, packet lengths, and service qualities. We address the power-optimization issues of communication systems conveying multiple message types based on finite-delay information theory. Given both the normalized transmission rate and the packet length of a system, the actual residual decoding error rate is a function of the transmission power. We propose a generalized power allocation framework for multiple message types. Two diffierent optimization cost functions are adopted: the number of service-quality violations encountered and the sum log ratio of the residual decoding error rate. We provide the optimal analytical solution for the former cost function and a heuristic solution based on a genetic algorithm for the latter one. Finally, the performance of the proposed solutions are evaluated numerically.
What problem does this paper attempt to address?