Research On Self-Interference Cancellation Methods For DFRC Systems Based On Symbol-Level Precoding

To ease the problem of wireless spectrum resource congestion,integrated sensing and communication(ISAC)is widely regarded as a low-cost solution that can significantly improve wireless resource utilization.Considering the cost,complexity of hardware and system optimization,dual-functional radar-communication(DFRC)systems can share spectrum and even hardware platforms between two functions to realize ISAC.However,in order to meet the communication and sensing requirements of full duplex in DFRC systems,self-interference(SI)is a major issue that must be addressed.Precoding technology utilizes multiple-input multiple-output(MIMO)gain to achieve efficient SI cancellation(SIC).Furthermore,symbol-level precoding(SLP)utilizes channel state information and data symbols to convert multi-user interference into available power,achieving better performance than traditional linear precoding.Therefore,this paper conducts research on the SIC method based on SLP for the MIMO-DFRC systems,and the specific research content is as follows:(1)Consider a MIMO-DFRC single base station system where a dual-function base station(BS)senses a point target while communicating with multiple downstream users.Aiming at the SI problem in this system,this paper proposes a SIC scheme based on SLP.Specifically,the scheme minimizes the sum of SI at all distances within the range of interest of the BS under the quality of service requirements of communication users and specific radar waveform pattern to suppress the SI leakage power,thereby reducing its impact on radar detection performance.The optimization problem based on the above requirements is highly nonlinear and its dimension increases with the number of antennas and transmitted symbols,so it is difficult to solve.This paper adopts a new two-layer iterative optimization algorithm that solves the above challenges through Penalty Dual Decomposition(PDD)and Block Coordinate Descent(BCD)methods.The simulation results show that the proposed SIC design based on SLP can effectively reduce the SI leakage power,thereby improving the detection performance of the DFRC systems.(2)In order to further address the issue of unstable SI suppression performance of the previous SIC design scheme for targets with different distances,this paper proposes a second fair SIC symbollevel precoding scheme,which minimizes the largest SI at different distances while ensuring communication and radar sensing.Due to the high complexity of the previous algorithms based on PDD and BCD,this paper adopts the accelerated gradient projection method to optimize them and use it to solve the two SIC-SLP design problems proposed in this paper.Simulation results have verified the efficiency of the proposed algorithm,and the second SIC design scheme achieves more stable SIC suppression performance and improves target detection performance.