| In view of the development bottleneck of the Internet of Thing(Io T)technology and the needs of the Io T node equipment,this thesis constructs the simultaneous wireless information and power transmission(SWIPT)in ultra-dense cloud radio access network(UD-CRAN)with wireless fronthaul to meet the future development needs of Io T.Then,combined with the characteristics of SWIPT technology and ultra-dense networks,such as base station diversity,line-of-sight wave transmission,and the base station on/off,this thesis designs a secure transmission scheme for wireless fronthaul link and access link.Specifically,this thesis equates the wireless fronthaul link to a multi-user multiple-input multiple-output system and then uses a generalized zero-forcing method based on singular value decomposition to design the beamforming matrix to support multi-stream fronthaul transmission of each remote radio head(RRH).Finally,the beamforming matrix is used to achieve power allocation for different RRHs.Aiming at the problem of secure transmission caused by SWIPT in the access links,this thesis regards strong interference as "green interference",using interference as artificial noise and energy signal,while meeting the secure communication of information receivers and energy harvesting of energy receivers.This thesis firstly solves the problem of maximizing the secrecy rate of the SWIPTbased UD-CRAN system,using the beamforming of information and energy transmission and link selection between RRH-users to construct the problem of maximizing the total secrecy rate of the system.In order to solve the constructed mixed-integer non-convex problem,this thesis uses the sparsity of the beamforming matrix to approximate integer variables,Then,using the successive convex approximation(SCA)to handle non-convex constraints in the problem,this thesis proposes the sparsity-based SCA algorithm.In order to balance algorithm performance and complexity,this thesis proposes the Deflation-based SCA algorithm,which uses the Deflation method to determine integer variables and then approximates non-convex constraints using the SCA method.Finally,numerical results are provided to verify the effectiveness of the proposed scheme.In the simulation results,the ultra-dense characteristics of the UD-CRAN network and the impact of RRH user selection on the overall secrecy rate of the system are analyzed.Then,this thesis considers the compromise between the sum secrecy rate and the total power consumption of the system.Based on beamforming design and link selection between RRH and users,RRH on/off is introduced to construct the problem of maximizing secure energy efficiency.In order to solve the constructed mix-integer non-convex fractional problem,this thesis first uses Dinkelbach’s Transform to get an iterative algorithm for fractional programming and obtains the approximate solution of the fractional programming through iteration.In order to solve the mix-integer non-convex problem during the iteration process,the SCA method is used to approximate the problem as a mix-integer second-order cone program(MI-SOCP)problem and the SCA-MISOCP algorithm is designed.In order to further reduce the complexity of the SCA-MISOCP algorithm,a low-complexity alternating optimization algorithm based on semidefinite relaxation(SDR)was proposed.Numerical results are presented in Chapter 4,which demonstrate the superiority of the proposed algorithm compared to existing algorithms.Then,simulation results are analyzed and explained to further validate the feasibility of "green interference". |
