The Study Of Transmission Scheme And Resource Allocation For Dense Networks In Wireless Communication

As one of the key technologies in the wake of the 5G era and the next generation wireless communication network,network densification is capable of meeting the explosive proliferation tele-traffic,massive equipment access and high network capacity demands.In order to fully exploit the benefits,network densification relies on the advanced transmission technology,with the aid of the designed allocation strategy of the limited resources to achieve a better network performance.This dissertation focuses on the study of transmission schemes and resource allocation for dense networks in 5G wireless communication,which specifically aims at the energy efficiency,load balancing,resource scarcity and secure communication key issues in dense networks.To this end,a series of the corresponding solutions are designed,with the goal of achieving the maximization of the spectral efficiency,green communication or secure communication.The main works and contributions of this dissertation can be summarized as follows:1)To investigate the energy efficiency of heterogeneous networks,a QoS-based transmission and power allocation scheme is proposed for maximizing the energy efficiency.Explicitly,RZF beamforming is first utilized to alleviate the interference and then the transmit power is optimized for satisfying both QoS and power constraint.To this end,the formulated optimization problem is first transformed into a series of subtractive subproblems.Next,by adopting successive convex approximation method,the subtractive subproblem is further transformed into a tractable geometric programming problem.Numerical results validate the effectiveness of the proposed scheme,and highlight the effects of the number of the antennas on the total energy efficiency.2)Considering the unfairness brought by the large number of antennas at MBS in heterogeneous networks,a fairness-based joint user association and power allocation scheme is put forward.Firstly,the lower bound of an asymptotic ergodic rate closed-form expression is derived over the estimated channel error under the imperfect channel.Next,the fairness-based joint optimization problem is formulated,which maximizes the proportional fairness of the user rate while satisfying the load and power constraints.To solve it,the problem is decoupled into a pair of subproblems,which can be solved by low-complexity methods respectively.Numerical results show the proposed solution outperforms other existing methods in terms of user fairness,and reveals that both the number of APs and of the antennas significantly contribute to the fairness on the spectral efficiency.3)A joint user-centric user association and resource allocation framework is proposed for maximizing the spectral efficiency of UDNs(Ultra Dense Networks)in terms of load balancing and limited resources issues.Explicitly,we consider practical AP traffic-load and RB constraints,and design joint user-centric clustering and resource allocation scheme based on a graph-theoretic framework.To be specific,the problem is decoupled into two independent subproblems,which can be solved by bipartite graph matching and by graph coloring method,respectively.Numerical results quantify the benefits of user-centric clustering architecture,and reveal the proposed solution offers superior performance under a range of practical system settings.4)To further investigate the limited resources issue in UDNs,a novel community-based user-centric clustering framework is proposed,by taking the users' group structure into account.Based upon the CUC clustering model,a joint CUC clustering and resource allocation solution with the goal of maximizing the resource utilization is conceived.In this light,we rely on a unsupervised learning method – Louvain method to decouple the formulated optimization problem into three subproblems: user-centric clustering,CUC clustering and resource allocation.Next,these subproblems are solved sequentially which conceive a Louvain-based joint user-centric clustering and resource allocation solution.Numerical results show that our proposed design offers superior performance over the state-of-the-art in terms of the exploitation of orthogonal RBs.5)In terms of secure communication and energy efficiency in UDNs,a novel secure usercentric clustering architecture is proposed.Then,the joint user association and transmission design for secure UDNs is investigated from a secrecy-energy-efficiency perspective.The optimized problem is formulated for both known and unknown eavesdropper channel state information(CSI)under different jamming strategy,with the aim of maximizing the secrecyenergy-efficiency while satisfying rate QoS as well as secure rate QoS.Finally,numerical results reveal the quantitative benefits of the proposed architecture,and show that the proposed single-function jamming strategy has potential merits in terms of increasing the secrecy energy efficiency by exploiting the eavesdropper's CSI.