With the increasing dependence on the mobile intelligent terminal, more and more wireless multimedia application has come into people’s perspective, and the requirement of the wireless mobile data services has increased tremendously. In addition, according to the statistics from the International Telecommunication Union, the energy consumption and the proportion of CO2 emissions of the mobile communication industry are increasing year by year. Therefore, how to guarantee the users’ demand, decrease the network energy consumption and improve the network energy efficiency through the effective utilization of the existing wireless spectrum resource, have become key problems needed to be solved of the next generation of wireless network optimization design.
In order to meet the targets of 1000 times data rate improvement and 10 times energy consumption decrease of the next generation of wireless network, academia and industry have devoted much effort on investigating the key technologies. For the ultra dense heterogeneous network (UDNs), deployed with a large number of Co-Channel and low power small cells under the coverage of macrocell, shorten the distance between the user and its associated base station, improve the users’ quality of service and the network throughput, has become an important component of the next generation wireless network. However, because of the large-scale deployment of small cells, the Co-Channel interference becomes more and more complexity and the energy consumption of the network increases tremendously, which limits the improvement of energy efficiency.
In this dissertation, the energy efficiency problem of UDN is deeply studied. The main contributions of the dissertation are listed as follows:
1. Energy-efficient resource allocation and cell association algorithm for ultra dense heterogeneous networks. Aiming at improving the current studies which didn’t consider the alteration of cell association, the resource allocation of macrocell and the users’ data rate requirement, this chapter investigates the downlink energy efficiency problem in ultra dense heterogeneous network. Under the constraint of the users’ quality of service, a network energy efficiency maximization problem by jointly optimizing the frequency resource partitioning of macrocells, transmission power and cell association bias of small cells is formulated. This problem is proven to be an NP-hard problem which cannot be worked out by direct methods in polynomial time. And then, considering the alteration of users’ serving cell in the optimization process and the characteristic of energy efficiency, it is difficult for the existing resource allocation and cell association algorithms to find the optimal solution set. An modified particle swarm optimization based resource allocation and cell association algorithm is employed to solve the proposed joint optimization problem. In modified particle swarm optimization, local search and multi-restart are introduced to improve the performance of tranditional particle swarm optimization and guarantee the convergence and the optimality of the proposed algorithm. At last, numerical simulations are conducted and the simulation results reveal that, compare to the existing resource allocation algorithms, the proposed algorithm can obtain better energy efficiency and system throughput performance with lower energy consumption.
2. Carrier aggregation based energy-efficient dynamic resource allocation algorithm for heterogeneous network. Aiming at handling the strong interference problem caused by macrocell to users in the small cell edge extension region, and the performance degradation and excessive energy consumption caused by the poorly arranged carrier configurations, in this chapter, a novel dynamic resource allocation scheme based on carrier aggregation to alleviate the downlink interference from macrocells to users located in the cell range expansion area, the performance degradation and the excessive energy consumption is proposed. By jointly adjusting the carrier configurations and transmission power of macrocell on the signal channel, the proposed algorithm can avoid the interference of users in the small cell edge extension region, increase the network throughput, and decrease the network energy consumption. Simulation results show that, compared to the existing resource allocation schemes, the proposed scheme can significantly boost the picocell throughput and increase the throughput of macrocell while save the transmission power of macrocell. At the same time of the overall throughput promotion, the energy consumption of the network is reduced, and hence the energy efficiency performance is also be improved. The improvement of the energy efficiency is attributable to the rational allocation of component carrier and the dynamic adjustment of the transmission power.
3. Energy efficient cell-association bias adjustment algorithm for UDN. Aiming at the problems of the network load imbalance, network throughput limitation and the network energy cost, caused by traditional cell association algorithms, in this chapter, under the constraint of users’ data rate, we propose a novel cell association adjustment scheme by optimizing the cell-association bias configuration which finally alters the cell association relationship, increases network throughput, turns off the extra small cells that have no users, and hence improve the network energy efficiency. Considering the non-convexity of the energy efficiency optimization problem and the coupled relationship between cell association and scheduling during the optimization process, it is difficult to achieve an optimal cell-association bias solution. In this chapter, we first propose an energy-efficient centralized Gibbs Sampling based cell-association bias adjustment (CGSCA) algorithm. In CGSCA, global information such as channel state information, cell association information, and network energy consumption information need to be collected. Then, considering the overhead of the message exchange and the implementation complexity for CGSCA to obtain the global information in UDNs, we propose an energy-efficient distributed Gibbs Sampling based cell-association bias adjustment (DGSCA) algorithm with a lower message-exchange overhead and implementation complexity. At last, we analyze the implementation complexities (e.g., computation complexity and communication complexity) of the proposed two algorithms and other existing algorithms. Simulations are conducted and the results show that, compared to other existing algorithms, CGSCA and DGSCA have faster convergence speed, and higher performance gain of the energy efficiency and throughput. In addition, the importance of the users’ data rate constraint in optimizing the energy efficiency is analyzed. The energy efficiency performance of different algorithms with different number of small cell and the number of sleeping small cells as the increasing number of small cells are compared.
4. Energy-efficient cell association algorithm with backhaul link constraint for cooperative UDN. Different from existing works which assume the data rate of fronthaul link is higher than that of the backhaul link, ignore the resource allocation and energy consumption of backhaul link, aiming at the problem of backhaul link congestion caused by coordinated multiple points (CoMP) joint transmission (JT), in this chapter, we focus on the scenario of the constrained backhaul link data rate in UDN and handle the network energy efficiency optimization problem under CoMP JT. Considering the scenario of the constrained backhaul link data rate where the backhaul link data rate cannot meet the forward link data rate, this chapter propose a network energy efficiency optimization algorithm under CoMP JT by jointly optimize the cell association and backhaul link resource allocation of the network. To be specific, the formulated system energy efficiency optimization problem is solved by jointly adjusting the sleep/on indicator, the cell association matrix and the backhaul link data rate. The sleep/on indicator, cell association and backhaul link data rate allocation are tightly coupled with the resource consumption of small cell and the backhaul link, which makes the closed-form of the optimal solution difficult to be obtained, under such circumstance, decomposition method is utilized to convert the original mixed integrate nonlinear fraction programming problem into two subproblems. The first subproblem is an energy consumption minimization problem accomplished by optimizing sleep/on indicator and cell association matrix under the users’ data rate constraint. This problem is a 0-1 nonlinear programming problem and the traditional cell association algorithm is difficult to be applied due to the large network scale. So we convert it by means of the penalty function and propose an improved modified particle swarm optimization algorithm to solve it. Based on the obtained optimal sleep/on indicator and cell association matrix, the upper and lower bound of the data rate constraint of each small cell are derived. The second subproblem is maximizing the minimum achievable energy efficiency optimization problem which is a linear fractional programming by optimizing backhaul link data rate under the derived upper bound and lower bound. Due to the positive denominator of the second subproblem, it is actually a quasiconvex optimization problem which can be transformed into an equivalent liner programming problem. Then a linear programming algorithm is proposed to solve it. Considering the fractional form of energy efficiency, the decomposition and transformation are not always the best. So finally we propose a complementary small cell turned on algorithm to converge the result to the optimal solution by turning on the sleeping cell to achieve the best energy efficiency. Numerical simulations are conducted and the results show that the proposed algorithm has a better energy efficiency and throughput than existing algorithms with the increasing number of small cells and users, respectively. Although the backhaul link energy consumption is higher than the existing algorithms due to CoMP JT, the overall energy consumption of the network is lower.