Research On The Joint Design Method Of Antenna Selection And Symbol Level Precoding For XL-MIMO

Extra-large scale massive multiple-input multiple-output（XL-MIMO）system can take advantage of the progressive orthogonality of the spatial channel to reduce the multi-user interference（MUI）and improve the spectrum efficiency by deploying hundreds of antennas at the base station to serve a small number of users.However,the the physical size of the array and the distance from the base station to the user are usually of the same order of magnitude in XL-MIMO scenarios,which makes the channel show near-field and spatial non-stationary characteristics,resulting in relatively little improvement in transmission performance of the system by some antennas in the array.Therefore,activating these antennas will reduce the energy efficiency of the system.To solve this problem,antenna selection（AS）can be used to activate the subarray that significantly improves the system performance.In addition,symbol level precoding（SLP）can convert MUI that is destructive to performance into beneficial signal,namely constructive interference（CI）,so as to furtherly improve system energy efficiency.Therefore,this thesis organically combines the above two technologies for XL-MIMO system.The specific work is as follows:Firstly,the joint optimization problem of AS and SLP under the constraint of symbol error probability（SEP）at the receiver is constructed to optimize the total energy consumption of XL-MIMO system for the downlink transmission scenario.Among them,the AS will make the l₀ norm term appear in the objective function,which leads to optimization problem as NP-hard.Therefore,this paper transforms the original problem into a convex problem by ₁l norm approximation.However,the approximate processing will bring errors in sparse mode,so an AS correction algorithm based on normalization is proposed to realize the joint design of the number of AS,AS mode and SLP.In addition,the design of SLP will also be affected by l₀ norm,so an SLP step-by-step design algorithm based on AS mode is proposed to furtherly reduce the total energy consumption of the system.Finally,the simulation results verify the effectiveness of the proposed scheme.Secondly,the AS is usually realized through switches,and its maximum switching rate is difficult to reach the symbol level change rate in the actual system.Therefore,there is a problem that the symbol level change rate does not match the switching rate.Aiming at this difficulty,a group sparse optimization problem of joint design of AS and SLP is established.Due to the l_2,0 norm term in its objective function,the optimization problem is NP-hard.In this paper,the l_2,1 norm is used to relax the original problem into a convex problem,and an AS correction algorithm based on normalization and an SLP step-by-step design algorithm based on AS mode are proposed.Finally,the performance of the proposed scheme is verified by simulation results.