Energy performance evaluation revisited: Methodology, models and results

In this paper, we revisit the radio network energy performance evaluation methodology, commonly known as E3F, that was developed in the European EARTH project during 2010-2011. We propose updates to the large-scale deployment and long-term traffic models in order to reflect the situation in a 5G time frame, and also an improvement in the sense that the baseline network is dimensioned to exactly meet the traffic and performance requirements. As an example, we apply the methodology in order to dimension and evaluate the energy performance of a state-of-the-art country-wide LTE network. The inter site distances of the resulting network range from 380 m in super dense urban (serving 20,000 citizens/km2) to 4850 m in rural areas (serving 100 citizens/km2). The resulting average power per area unit is 0.14 kW/km2, which is significantly lower than what was found in the EARTH project. This can be explained by two factors: i) The present study assumes more efficient base stations, and ii) the baseline network of the original E3F was over-dimensioned. Nevertheless, the conclusion from the EARTH project, that the absolute majority of energy consumed in a radio network is spent on providing coverage, still remains.