Optimizing wireless access networks towards power consumption: Influence of the optimization algorithm

Abstract

Abstract-Nowadays, wireless access networks are already amongst the top power consumers in the ICT (Information and Communication Technology) sector. As it expected that these networks will further expand in the future due to the extreme growth in mobile devices and the high bit rate demand of the applications running on these devices, it is important to consider power consumption as a key parameter in the network design phase. In this paper, two optimization algorithms are proposed: a capacity-based heuristic which aims to reduce power consumption by responding to the instantaneous bit rate demand by the user and an evolutionary opposition-based learning algorithm focusing on the joint-optimization of power consumption and geometrical coverage. Applying both algorithms on a realistic suburban case in Ghent, Belgium, show that both algorithms are able to design an LTE-A network consuming only 24% and 29%, respectively, of the power consumed by the reference scenario which is representative for today's networks. The evolutionary algorithm outperforms the capacity-based algorithm by obtaining a 5% lower power consumption, while the capacity-based heuristic has a 2 to 3% higher coverage. Future research in joint-optimization algorithms of energy and network performance is definitely needed.

Multi-objective optimization of cognitive radio networks

“…Hence, the global exposure EG is defined as a weighted average of the mean electric field E50 and the 95-percentile of the field strength E95 over the covered area [25], in order to optimize median and maximal exposure values. As in [23], we consider an equally weighted E50 and E95. Hence, the EG can be described by the following equation:…”

“…Algorithm 1 describes the network optimization algorithm that will be used for minimizing the network power consumption, spectrum usage, and exposure (i.e., goal KPIs). The algorithm is heuristic and capacity-based [23]. Hence, we cannot guarantee an absolute optimal network solution but a solution that is good enough for solving the optimization problem.…”

“…The lower the number of users per square kilometre, the more diverse the possible distributions of those users, and the higher the number of possible solutions. Notice that interference and spectrum management is not included in the base algorithm described in [23,24]. Hence, modifications were performed for the multi-objective optimization goals for this work.…”

“…Besides spectrum usage efficiency, power consumption, and exposure of human beings to radio radiation are important network KPIs. These indicators are closely related to the environmental footprint of Information and Communication Technologies (ICT) [23,24]. Hence, to achieve environment-friendly wireless networks, it is also required to optimize power consumption and exposure.…”

A Novel Design Approach for NB-IoT Networks Using Hybrid Teaching-learning Optimization

“…In [5] a capacity-based heuristic is presented for energyefficient network. Moreover, the application of evolutionary algorithms (EAs) to LTE network optimization is also addressed in previous works [6], [7], [8]. In this paper, we consider LTE networks with massive deployment of NB-IoT devices.…”