| With the access of a large number of wireless devices,the integrated radar communication(Rad Com)system with shared spectrum has become one of the key technologies to alleviate the shortage and expensiveness of spectrum resources.The similar antenna architecture and signal processing mo de of independent radar communication system provide the possibility of integrated design.The massive multiple-input multiple-output(MIMO)beamforming technology has the advantages of high energy efficiency,high-frequency spectral efficiency and link reliability.It can form a beam with more directional concentration and stronger expected directional signal for the specified downlink users,which is also very important for the future wireless communication system.Therefore,it is of great research significance to combine the Rad Com system with massive MIMO beamforming technology.However,in the radar communication integrated system with all digital beamforming architecture,each antenna is connected with an radio frequency(RF)chain,while the large antenna array needs hundreds of RF chains,which will bring high cost,unbearable hardware complexity and excess ive power consumption.Among them,the digital-to-analog converter(DAC)accounts for the largest power consumption of the RF chain in the downli nk.The power consumption caused by the DAC is exponentially related to its own quantization accuracy.To settle the above problems,low-precision DAC can be configured to reduce the cost of each RF chain or directly reduce the number of RF chains.Hence,this thesis first takes the quantization of low-precision DAC as the constraint,studies the all digital beamformer of the Rad Com system,then considers the hybrid beamforming scheme to reduce the RF chain,extends the low-precision DAC to the hybrid structure,and discusses the trade-off between performance and hardware complexity.The hybrid beamforming scheme is the combination of baseband domain digital beamforming architecture and RF domain analog architecture.The research contents of this thesis are as follows:(1)The fully-digital beamforming design of Rad Com system is expressed as a weighted optimization problem.At the communication end,the criterion is to minimize the mean square error of estimation symbol and transmission data symbol;at the radar end,the design goal is to minimize the residual error between transmission beam and ideal beam,and the weight factor is introduced to control the proportion of the two functions,so as to realize the information exchange of downlink equipment and form detection beam to complete radar perception.The low-precision DAC quantization is used as a constraint to control the modulus of the real part and imaginary part of the transmitted signal element.Aiming at the non-convex problem of multi-objective joint optimization formed by the final objective function,the multi-variable alternating minimization framework is used to separate the variables,so that it can be solved independently iteratively.For the variable emission signal,particle swarm optimization simulated annealing optimization is introduced to obtain the approximate optimal solution in the discrete constraint set.(2)For the sake of further energy conservation and the trade-off between hardware complexity,power consumption and performance,th e low-precision DAC and hybrid beamforming fully connected structure are combined to carry out a joint integrated system design.The hybrid structure of nonlinear mapping in d igital domain and linear matrix in analog domain is considered to improve the per formance.First,it is assumed that the analog beamforming design is only related to the current channel variation,so the analog mapping scheme is independently pre-designed based on singular value decomposition.For the nonlinear optimization problem of dig ital beamforming in the baseband domain,constructive interference is considered at the communication end,that is,the beneficial interference is used to promote the receiving symbol principle to judge the boundary line of the region,thereby increasing the bit error rate of downlink communication.At the radar end,the similarity between the transmitted signal and the rad ar signal is taken as the constraint of the objective function,in which the similarity factor controls the proportion of the two functions.The low-precision DAC quantization is still used as a constraint condition to constrain the modulus value of the qu antized signal after digital beamforming.Since the resulting problem is non-convex,convex relaxation techniques are used to relax the quantization constraints and then solve it.(3)Finally,the simulation experiments are carried out.The numerical structure proves that the adopted method has obvious performa nce gains compared with the traditional beamforming technology in terms of bit error rate and achievable rate,and can form a power-concentrated detection beam for a specified angle. |
