Research On Green Telecommunication Technologies Of Femtocell Networks

To cope with the explosive growth of wireless data service, the scale of wireless infrastructure grows sharply, which causes huge amount of energy consumption. A latest report by European Union shows that mobilecommunications contribute15%-20%of the entire information and communicationstechnologies (ICT) industry energy, which has become the5th energy consumption industry worldwide. On the other hand, as an important component of mobile communication network, the amount of femtocellincrease progressively.A market report by Informa Telecoms&Media shows up to17million femtocell base station (FBS) have been produced in2013year-end and more than85million FBS will be deployed in2016year-end, which will consume huge energy source. According tothe above analysis, the research of energy efficiency telecommunication techniques of femtocell networks will play a very important role in energy-saving.In this paper, we firstly overview and analyze the research status of green telecommunication technologies in femtocell networks. To resolve the existing problems of interference management, high complexity of resource allocationand quality of service (QoS) provisioning in energy efficiency femtocell networks, we propose three types of energy efficiency model:interference mitigation based instantaneous energy efficiency resource allocation model; exponentially weighted moving average (EWMA) based energy efficiency resource allocation model and delay QoS provisioning based energy efficiency power control model. Based on above energy efficiency models, three energy-saving strategies are proposed. The main work and innovation can be summarized asfollows: 1) To improve the energy efficiency of two-tier femtocell network and mitigate con-channel interference, a convex-pricing based distributed energy efficiency resource allocation strategy is proposed. Firstly, a convex-pricing based energy efficiency model is built, based on which we formulate the energy efficiency resource allocation optimization problem. Secondly, due to the proposed problem is NP-hard, to reduce the complexity while maintaining good performance, we decompose this proposed problem into two sub-problems:power control and subchannel allocation. Assuming user’s transmit power has been equally allocated to all subchannels, we get an energy efficiency subchannel allocation scheme; after that, the power control problem is" formulated as a super-modular game and get Pareto optimal solution of transmit power. At last, we propose a convex-pricing based distributed resource allocation (CPDRA) algorithm. Numerical results show the proposed CPDRA algorithm is effective in increasing energy-efficient performance compared with Round-Robin Scheduling (RRS) and an Energy Efficient Water Filling Algorithm (EEWFA), while protecting neighbour-cell users.2) Instantaneous energy efficiency resource allocation model is widely used in current academia,however, the model can not get the closed-form solution of subchannel allocation scheme and power control.According to above analysis, a fair low complexity energy-saving strategy is proposed. Firstly, an exponentially weighted moving average (EWMA) based energy efficiency resource allocation model is built, and we formulate the energy efficiency resource allocation problem. Secondly, to resolve this problem, we get a closed-form solution of transmit power in sparsely deployed femtocell network and a closed-form near optimal solution of transmit power in densely deployed femtocell scenario. Considering fairness, we maximize the geometric mean of the energy efficiency of all users to avoid that someone can not be served by femtocell base station, and acquire a fair subchannel allocation scheme. At last, a fair low complexity distributed resource allocation algorithm has been proposed. Simulation results show that the proposed algorithm has a lower complexity with slight loss of energy efficiencycompared with Round-Robin Scheduling (RRS) and a noncooperative energy-efficient power optimization (NEEPO) algorithm, while providing user’s fairness.3) Due to the fact that the existing works are barely considering delay provisioning, an energy-efficiency power control strategy based on delay guarantee is proposed. Firstly, considering the delay provisioning of real-time service,an energy efficiency model with statistical QoS guarantee is proposed, and we formulate the power control problem in the uplink of two-tier femtocell networks. Secondly, for the given delay requirement, this proposed problem is modeled as a Stackelberg game. In the game, macro-user acts as a leader and knows all other femto-users’ power strategies, while femto-users are followers and only know leader’s strategy. In the game procedure, leader selects transmit power strategy firstly and followers move subsequently based on leader’s strategy until reaching steady state. At last, a Q-learning based energy efficiency power control algorithm with delay provisioning is proposed. Simulation results show the proposed algorithm can improve the energy efficiency while providing delay guarantee, and the convergence speed is faster compared with a conjecture based multi-agent Q-learning (CMAQL) algorithm.