Energy efficient resource allocation for quantity of information delivery in parallel channels

International audienceThis paper deals with the problem of minimising the energy required to transmit a certain amount of information bits. If there is no constraint, in a Gaussian channel, the solution leads us to the use an infinite transmission time or an infinite bandwidth. In particular, the minimum required signal-to-noise ratio to transmit one bit is equal to −1.59 dB. To allow the development of new green communication services, a finite bandwidth and a finite time have to be used in practice. Thus, we focus on the transmission of a given number of bits over a set of parallel Gaussian channels when there is an energy constraint, and we have the goal of minimising either the highest transmission time (amongst the channels) or the average channel occupancy time. This is a resource allocation problem that is formulated by targeting a certain energy consumption factor, herein defined as the ratio between the energy required in the asymptotic regime and the energy required to transmit a certain quantity of information with limited (time, bandwidth) resources. The Pareto front for the joint optimization of the maximum transmission time and the average channel occupancy time is also derived. Several explicative numerical results of the derived resource allocation algorithms for energy efficient communications are also reported