Joint Radio Resource Management In Heterogeneous Wireless Networks

The coverage area of the radio access network usually overlaps with each other in the heterogeneous wireless network (HWN), and this creates a new radio resource dimension, radio access network (RAN), for the wireless user. Combing with the ex-isting radio resource within each RAN, the user in the HWN can obtain the radio re-source in dimensions of power, spectrum, time, RAN, storage, and computation. To fully explore the multiple dimensions of radio resource in the HWN, we should con-struct an efficient matching among user traffic, RAN, and the radio resource within each RAN. As a result of such efficient matching, we could obtain multiple hetero-geneous gains:multi-user diversity gain, multi-RAN diversity gain, multi-RAN parallel transmission gain, and traffic matching gain. Therefore, in this thesis, we will focus on designing the flexible network control architecture and efficient joint radio resource management and traffic splitting algorithm for the HWN control. Since the traffic allo-cation among RANs and the radio resource management within each RAN depend on each other, to resolve their coupling, we decouple the HWN control objective with two sub-objectives:1). the radio resource management within each RAN, and 2) the traffic allocation among RANs. We first concentrate on the first sub-objective, and design the near-optimal power allocation scheme for the interference-limited RANs, and the joint radio resource management scheme for the orthogonal RANs. We then proceed to the second sub-objective with a first attention on the orthogonal HWN constituted by the orthogonal RANs. Exploring the radio resource in the RAN dimension, we design the selective single RAN transmission scheme in the orthogonal HWN. Also, we al-low parallel transmission in the orthogonal HWN, and optimize the uplink transmission power of the wireless user. For the HWN consisted of the orthogonal RANs and the interference-limited RANs, considering the network dynamic, we design an adaptive cross-network cross-layer HWN control algorithm, and provide the theoretical perfor-mance bound. We summarize our main contributions as follows:1. We study the power allocation problem in the interference-limited RANs. Con-sidering the channel fading, our dynamic power allocation algorithm can adapt to the dynamic in the interference-limited RANs. To resolve the NP-hard com-plexity of the power allocation in the interference-limited RANs, we first utilize successive convex approximation to derive its local optimum. Then, we design an efficient method to force the local optimum to approach the global optimum. We also quantify the tradeoff among the network throughput, the traffic delay, and the network computational ability.2. For the orthogonal RANs, we study the joint power and resource block (RB) al-location problem for the frequency division relay, and the joint subframe, power, and RB allocation problem for the time division relay. Considering maximum and average power constraints, our control strategy can adapt to the user traffic dy-namic and the channel fading. Even though the joint subframe, power, and RB allocation problem is a mixed-integer non-linear optimization problem, we could derive its optimal solution by continuous relaxation and dual decomposition.3. We study the selective single RAN transmission scheme in the orthogonal HWN. We explore the multi-RAN diversity gain and multi-user diversity gain by con-structing an efficient matching among user traffic, single RAN, and the radio resource within each RAN. To overcome the control complexity introduced by the resource granularity and the single RAN constraint, we use the lineariza-tion method to form an integer linear optimization problem, and design a low complexity single RAN transmission method in the orthogonal HWN.4. In the HWN consisted of orthogonal RANs, we allow parallel transmission for the wireless user over RANs. By flexibly joint radio resource management in the HWN, we optimize the uplink transmission power of the user equipment to lengthen the battery lifetime. We consider the practical resource granularity con-straint of the orthogonal RANs. To reduce the HWN management complexity, we propose a rate-power efficiency based low complexity HWN RB allocation method.5. For the HWN consisted of the orthogonal RANs and the interference-limited RANs, we design an adaptive cross-network cross-layer (CNCL) HWN control method. Such CNCL control method can adapt to the traffic dynamic and the channel fading. Considering the maximum and average power constraints, we can asymptotically solve the power control problem in the interference-limited RANs, and optimally solve the joint power and RB allocation problem in the orthogonal RANs. Also, our CNCL method can deal with the user traffic het-erogeneity, and can adaptively allocate the user traffic among the RANs. We theoretically give the HWN performance bound, and quantify the tradeoff among network utility, queue delay, and network computational ability.