| dc.description.abstract | The development of wireless technologies has been stimulated by the ever increasing network capacity and the diversity of users' quality of service (QoS) requirements. It is widely anticipated that next-generation wireless networks should be capable of integrating wireless networks with various network architectures and wireless access technologies to provide diverse high-quality ubiquitous wireless accesses for users. However, the existing wireless network architecture may not be able to satisfy explosive wireless access request. Moreover, with the increasing awareness of environmental protection, significant growth of energy consumption caused by the massive traffic demand consequently raises the carbon emission footprint. The emerging of green energy technologies, e.g., solar panel and wind turbine, has provided a promising methodology to sustain operations and management of next-generation wireless networks by powering wireless network devices with eco-friendly green energy.
In this thesis, we propose a sustainable wireless network solution as the prototype of next-generation wireless networks to fulfill various QoS requirements of users with harvested energy from natural environments. The sustainable wireless solution aims at establishing multi-tier heterogeneous green wireless communication networks to integrate different wireless services and utilizing green energy supplies to sustain the network operations and management. The solution consists of three steps, 1) establishing conventional green wireless networks, 2) building multi-tier green wireless networks, and 3) allocating and balancing network resources.
In the first step, we focus on cost-effectively establishing single-tier green wireless networks to satisfy users' basic QoS requirements by designing efficient network planning algorithm. We formulate the minimum green macro cell BS deployment problem as an optimization problem, which aims at placing the minimum number of BSs to fulfill the basic QoS requirements by harvested energy. A preference level is defined as the guidance for efficient algorithm design to solve the minimum green macro cell BSs deployment problem. After that, we propose a heuristic algorithm, called two-phase constrained green BS placement (TCGBP) algorithm, based on Voronoi diagram. The TCGBP algorithm jointly considers the rate adaptation and power allocation to solve the formulated optimization problem. The performance is verified by extensive simulations, which demonstrate that the TCGBP algorithm can achieve the optimal solution with significantly reduced time complexity.
In the second step, we aim at efficiently constructing multi-tier green heterogeneous networks to fulfill high-end QoS requirements of users by placing green small cell BSs. We formulate the green small cell BS deployment and sub-carrier allocation problem as a mixed-integer non-linear programming (MINLP) problem, which targets at deploying the minimum number of green small cell BSs as relay nodes to further improve network capacities and provide high-quality QoS wireless services with harvested energy under the cost constraint. We propose the sub-carrier and traffic over rate (STR) metric to evaluate the contribution of deployed green small cell BSs in both energy and throughput aspects. Based on the metric,
two algorithms are designed, namely joint relay node placement and sub-carrier allocation with top-down/bottom-up (RNP-SA-t/b) algorithms. Extensive simulations demonstrate that the proposed algorithms provide simple yet efficient solutions and offer important guidelines on network planning and resource management in two-tier heterogeneous green wireless networks.
In the last step, we intend to allocate limited network resources to guarantee the balance of charging and discharging processes. Different from network planning based on statistical historical data, the design of resource allocation algorithm generally concerns relatively short-term resources management, and thus it is essential to accurately estimate the instantaneous energy charging and discharging rates of green wireless network devices. Specifically, we investigate the energy trading issues in green wireless networks, and try to maximize the profits of all cells by determining the optimal price and quantity in each energy trading transaction. Finally, we apply a two-stage leader-follower Stackelberg game to formulate the energy trading problem. By using back induction to obtain the optimal price and quantity of traded energy, we propose an optimal algorithm, called optimal profits energy trading (OPET) algorithm. Our analysis and simulation results demonstrate the optimality performance of OPET algorithm.
We believe that our research results in this dissertation can provide insightful guidance in the design of next-generation wireless communication networks with green energy. The algorithms developed in the dissertation offer practical and efficient solutions to build and optimize multi-tier heterogeneous green wireless communication networks. | en |
