Energy Consumption Analysis for Adaptive Transmission of Big Data Over Fading Channels: A Statistical Characterization

We study the energy consumption of the adaptive transmission of big data over fading channels. Transmission of big data usually lasts over multiple channel coherence time periods. With adaptive transmission, the transmission rate and/or power is adaptively adjusted according to the channel realization. The energy comsumption varies with channel realizations. We propose a novel analytical framework to statistically evaluate the energy consumption of adaptive big data transmission. For the slow fading scenario, we derive the exact probability density function (PDF) and cumulative distribution function (CDF) of energy consumed over Markov channels. For the fast fading cases, we apply the statistical mixture model to estimate the PDF of energy consumption for wireless transmission of big data adaptively. Selected numerical results are presented to illustrate and to validate the mathematical formulations. These analytical results will greatly benefit the study of big data applications in the wireless environment.