Research On Rate Splitting Multiple Access Technology In Cache-Enabled Cloud Radio Access Networks

The rapid growth in smart mobile devices and the wide use of multimedia applications motivate the evolution of network architecture.Among the designed network architectures,cloud radio access network（C-RAN）has received considerable attention because of its efficient resource management.More specifically,C-RAN migrates the processing function of conventional base stations（BSs）to a centralized resource pool.Then,it replaces the high-cost BSs with simple-structured wireless remote radio heads（RRHs）and employs the coordinated multi-point technique to achieve communications with users.However,the dense deployment of RRHs aggravates the burden of fronthaul link,and thus the fronthaul capacity crunch has become a significant bottleneck of C-RAN to enhance the network capacity.Edging caching can pre-store some popular and reusable contents to ease the fronthaul load.To boost the spectrum efficiency in multi-user systems,the rate splitting（RS）multiple access technology is proposed.RS can partly decode the interference and partly treat the interference as noise when decoding the desired signal.Therefore,to solve the transmit rate saturation,communication security,and the limitation of fronthaul capacity in the content delivery process of cache-enabled C-RANs（C²-RANs）,this dissertation studies RS-based transmit scheme and its resource allocation algorithm.The main contributions are summarized as follows.（1）Resource Allocation Strategy for Rate Splitting-based Multigroup Multicast SystemsThis dissertation proposes an RS-based multigroup multicast content delivery scheme for C²-RANs.RRH scheduling,multicast beamforming,and common rate allocation are jointly investigated to maximize the users’fairness rate and to ensure the fronthaul capacity and RRH transmit power constraints.To solve the formulated discrete non-convex problem,we propose two iterative optimization algorithms and analyze their convergence,optimality,and complexity.The first one is designed by using Woodbury matrix identity,first-order Taylor expansion,and the iterative weighted（?）₂-norm to construct surrogate functions,and then transform the formulated problem to a convex one.The second one is a sub-optimal two-tier alternating iterative algorithm,which uses the quadratic transform（QT）method to construct the surrogate functions.It solves the outer-tier problem with the closed-form expressions and the inner-tier one via the convex framework.Simulation results show that,compared with conventional delivery schemes,the RS-based scheme improves the content delivery rate.Besides,compared with the semi-definite relaxation（SDR）algorithm,the proposed two algorithms can greatly reduce the running time and ensure network performance.（2）Rate Splitting-based Transmit Scheme Design and Performance Optimization for Physical-Layer SecrecyThis dissertation employs the RS technique to propose a novel physical-layer secure transmit scheme for C²-RANs.This secure approach uses the common stream for dual purposes,serving as the desired message for legitimate users and artificial noise（AN）to weaken eavesdroppers’decoding ability.On the premise of ensuring that the common stream can act as AN,we jointly steer RRH scheduling,beamforming,common rate allocation,and common secrecy rate allocation to maximize the secrecy rate fairness among all legitimate users.However,the formulated problem involves the difference between two logarithmic functions and the coupled variables,which is hard to solve.To this end,we use the QT method to construct the surrogate functions for non-convex constraints and then design a two-tier alternating iterative optimization algorithm to solve the recast problem.Extensive simulation results show that the proposed transmit scheme is superior to the AN-aided physical-layer secure transmit scheme.Also,the proposed algorithm achieves 3.15 bps/Hz performance gains over the SDR algorithm and successive convex approximation（SCA）algorithm.（3）Research on Interference Management based on Rate Splitting in D2D Enhanced Heterogeneous NetworksThis dissertation proposes an RS-based interference management scheme for content delivery scenarios in D2D enhanced multi-subcarrier heterogenous C²-RANs.We joint study the cellular user grouping,D2D subcarrier allocation,RRH scheduling,beamforming,and common rate allocation to maximize the sum rate.To solve the formulated combinational optimization problem,we divide it into three subproblems,namely cellular user grouping,D2D subcarrier allocation,and sparse beamforming design and resource allocation.We first prove that two cellular users with low channel similarity should be grouped together and then develop a greedy searching algorithm and a fairness-ensuring accelerated bisection searching algorithm for the first subproblem.We model the second subproblem as a many-to-one matching problem with peer effect.Then,we use Gale-Shapley（GS）algorithm to obtain the initial matching state and then employ swaption operation to tackle the peer effect.The subcarrier allocation algorithm can achieve a two-sided stable matching result.We use the QT method to construct surrogate functions and design a two-tier alternating iterative algorithm to solve the third subproblem.By integrating these algorithms,we propose an overall iterative optimization algorithm to solve the initial problem.Simulation results validate the superiority of our proposed interference management scheme and algorithms.（4）Research on MRRH Cooperative Transmission Scheme in Hierarchical Rate Splitting Multiple Access SystemsThis dissertation proposes a hierarchical RS content delivery scheme and macro RRH（MRRH）-aided cooperative transmit mechanism for C²-RANs.This cooperative scheme aims to provide an alternative method for the general RRHs to request the caching-missing contents,which can ease the fronthaul burden.We jointly optimize the user grouping,RRH decoding order,RRH scheduling,beamforming,and common rate allocation to minimize the content delivery delay.To solve the formulated discrete non-convex problem,we divide it into three subproblems.The first one is to group users to reduce intra-group interference.For that,taking users’channel similarity,requesting content size,and size difference among users into account,we develop a user grouping algorithm with low complexity.The second one is to optimize the RRH decoding order to reduce the broadcast link delay.For that,we propose the decoding order update algorithm to achieve the global optimum.The third one is to jointly design sparse beamforming matrices,bandwidth resource of fronthaul links,and common rate allocation to reduce the content delivery delay.For that,we first extend the QT method to the multiple receiving antennas scenario,and then construct surrogate functions and design a two-tier alternating iterative algorithm.By integrating these three algorithms,we develop an overall algorithm to solve the initial problem.Simulation results show that the proposed cooperative transmit scheme can avoid network breakdown when there is no available fronthaul capacity.Besides,compared with the existing algorithms,our algorithm can reduce content delivery delay.There are a total of 36 figures,8 tables,and 156 references in this dissertation.