Resource Management in Cloud Radio Access Networks
This thesis focuses on resource management both in communication and computing sides of the cloud radio access networks (C-RANs). Communication and computing resources are bandwidth, power, baseband unit servers, and virtual machines, which become major resource allocation elements of C-RANs. If they are not properly handled, they create congestion and overload problems in radio access network and core network part of the backbone cellular network. We study two general problems of C-RAN networks, referred to as communication and computing resource allocation problem along with user association, base band unit (BBU) and remote radio heads (RRH) mapping problems in order to improve energy efficiency, sum data rate and to minimize delay performance of C-RAN networks.
In this thesis, we propose, implement, and evaluate several solution strategies, namely posterior probability based user association and power allocation method, double-sided auction based distributed resource allocation method, the energy efficient joint workload scheduling and BBU allocation and iterative resource allocation method to deal with the resource management problems in both orthogonal and non-orthogonal multiple access supported C-RAN networks. In the posterior probability based user association and power allocation method, we apply Bayes theory to solve the multi-cell association problem in the coordinated multi-point supported C-RANs. We also use queueing and auction theory to solve the joint communication and computing resource optimization problem. As the joint optimization problem, we investigate the delay and sum data rate performance of C-RANs. To improve the energy efficiency of C-RANs, we employ Dinkelbach theorem and propose an iterative resource allocation method. Our proposed methods are evaluated via simulations by considering the effect of bandwidth utilization percentage, different scheduling weight, signal-to-interference ratio threshold value and number of users. The results show that the proposed methods can be successfully implemented for 5G C-RANs.
Among the various non-orthogonal multiple access schemes, we consider and implement the sparse code multiple access (SCMA) scheme to jointly optimize the codebook and power allocation in the downlink of the C-RANs, where the utilization of sparse code multiple access in C-RANs to improve energy efficiency has not been investigated in detail in the literature. To solve the NP-hard joint optimization problem, we decompose the original problem into two subproblems: codebook allocation and power allocation. Using the graph theory, we propose the throughput aware sparse code multiple access based codebook selection method, which generates a stable codebook allocation solution within a finite number of steps. For the power allocation solution, we propose the iterative level-based power allocation method, which incorporates different power allocation approaches (e.g., weighted and successive interference cancellation ) into different levels to satisfy the maximum power requirement. Simulation results show that the sum data rate and energy efficiency performance of non-orthogonal multiple access supported C-RANs significantly increases with the number of users when the successive interference cancellation aware geometric water-filling based power allocation is used.