Energy-efficient Network Deployment And Resource Management In Heterogeneous Cellular Networks

Recent years, the huge growth in popularity of wireless smart devices and the surge of various mobile applications have brought many challenges for cellular communication system. On the one hand, the massive mobile devices and kinds of application services, such as high-speed downloads, video calls and online games, have led to the increasing demand for Quality of Service (QoS), including huge capacity, high data rate and low delay. On the other hand, in order to satisfy the increasing traffic demand, network operators have deployed more Base Stations (BSs) and supporting facilities, which result in the increase of the associated energy consumption. Considering the enviroment impact and operating costs, network operators pay more and more attentions to the system energy efficiency. Therefore, joint consideration of QoS and energy efficiency, especially optimizing system energy efficiency while guaranteeing the users’QoS, is practical and meaningful.In order to meet the service needs, both LTE and WiMAX standard groups have introduced low power nodes (denoted as micro BSs (MiBSs)) on the basis of traditional macrocell networks to expand system capacity, offload the traffic of macro BSs (MaBSs), enhance indoor coverage and improve the performance of cell-edge users. These MiBSs and MaBSs constitute the heterogeneousness cellular networks. In addition to providing better QoS performance, comparing to traditional macrocell networks, heterogeneousness networks have many advantages and challenges in improving energy efficiency performance, such as high spectrum reuse, low power of MiBSs, short distance communication, random network topology, serious interference, complicated resource mangament and so on. Facing the above advantages and challenges, in this dissertation, we study the network deployment and resource management strategy, which jointly considers BS deployment density, transmission power, sleep mode operation and bandwidth allocation to optimize system energy efficiency under user QoS constraints in heterogeneousness cellular networks. Our main research work and contributions are as follows:1) This dissertation addresses the energy efficient deployment problem in heterogeneous networks, and jointly optimizes BS density and transmission power to minimize the area power consumption under coverage performance constraints.Poisson Point Process (PPP) model is utilized to derive the relations between the average coverage probability (cumulative distribution function of signal to interference plus noise ratio) and deployment strategy (i.e., BS density and BS transmission power). Based on the results, we formulate an optimization problem, which jointly optimizes MaBS density, MaBS transmission power, MiBS density and MiBS transmission power to minimize the area power consumption under coverage performance constraints, and give out the optimal deployment strategy. Through extensive simulations with practical data sets, we find that:compared to homogeneous network deployment, heterogeneous network deployment has absolute advantage in energy efficiency performance; optimizing the transmission power can further reduce system power consumption.2) This dissertation studies the MaBS sleep strategy in heterogeneous cellular networks to minimize the area power consumption under user power consumption constraint while maintaining the original coverage performance.PPP model is used to derive the relations among the downlink average coverage probability, uplink user power consumption and proportion of sleep MaBSs. Based on the results, we propose three schemes based on increasing the transmission power of remaining active MaBSs and additionally introducing MiBSs to maintain the original coverage performance after switching off some MaBSs. Besides, we achieve the maximum proportion of sleep MaBSs to guarantee that the uplink power consumption is in users’tolerance range. After that, we formulate an area power consumption minimization problem under user power consumption constraint while maintaining the original coverage performance, and achieve the optimal proportion of sleep MaBSs, MaBSs’ transmission power and introduced MiBSs’density.3) This dissertation studies the MiBS sleep strategy in heterogeneous cellular networks, proposes a threshold-based MiBS sleep strategy and analyzes the tradeoff relation between system power consumption and user throughput.A threshold-based MiBS sleep strategy is designed. When the load of a MiBS is below the predefined energy saving threshold and can be accommodated by its nearest MaBS, the MiBS will be switched to sleep mode. On the contrary, if the load of a sleep MiBS exceeds the threshold or the nearest MaBS is overload, it will be re-activated. The mode switching problem is modeled as Quasi Birth-and-Death (QBD) process. Using tools from PPP model and markov chain theory, we derive the expressions of MaBS average load, MaBS average power consumption, average number of sleep MiBSs, average load of MiBS, MiBS average power consumption, average data rate of macro user and micro user. Treating the ratio between system average power consumption and system average throuphut as energy efficiency metric, we achieve the relation between energy efficiency and energy saving threshold, then decide the corresponding threshold value.4) This dissertation studies the energy-aware scheduling strategy in multi-hop heterogeneous cellular networks to minimize the system power consumption while guaranteeing the system stability or average delay performance.Applying the cognitive radio technique in cellular networks to enhance system performance, we propose a multi-hop heterogeneous cognitive cellular architecture which consists of MaBSs and relay stations. This architecture can extend the celluar available spectrum resources without specific requirement on users’cognitive capability. Considering both temporal and spatial spectrum features in the proposed architecture, we propose scheduling the transmission of delay-tolerant traffic when the network nodes have more available spectrum bands to reduce the network power consumption. A cross-layer stochastic optimization framework is mathematically formulated, which jointly considers power allocation at physical layer, link scheduling at link layer, and flow routing at network layer to minimize the system power consumption while guaranteeing the system stability. Based on Lyapunov optimization technique and branch-and-bound framework, we develop an energy-efficient scheduling algorithm. Theoretical analysis shows that the proposed algorithm can not only stablize the system, but also achieve the average power consumption deviated no more than O(1/V) from the optimal result where V is a control parameter.