| In the multi-user Multiple Input Multiple Output(MIMO)wireless transmission system,the zero-forcing precoding is the simplest precoding method,which can effectively remove inter-symbol interference(ISI)and inter-channel interference(ICI)between users,it has gained a lot of attentions.The goals of the traditional zero-forcing precoder design is to maximize signal-to-noise ratio(SNR)or minimize mean square error(MSE).However,the precoder thus designed may not retain the structural information of source signal.In this paper,to overcome the inadequacy of the precoder based with SNR and MMSE in image transmission,a method of weak forced zero precoder design based on structure similarity index(SSIM)is proposed.The main work of this paper is as follows:1.An algorithm of causal weak zero precoder design with SSIM is proposed under the condition of total power constraint or per-antenna power constraint.Based on the state space representation of the channel transfer function and the weak zero condition,we derive the parameterized expression of the precoder,and convert the problem of the precoder design into the solution of the optimal parameters.Moreover,polynomial fitting is adopted to convert the non-convex problem of designing the precoder with SSIM index into convex problem.Finally,the convex optimization toolbox is used to find the optimal parameters.The experimental results show that the proposed causal zero-forcing precoder can satisfy the power constraints of the base station and have higher PSNR and SSIM value.2.Aiming at the problem that the precoder cannot obtain higher performance due to the limitation of causality,a non-causal zero-forcing precoder based on SSIM index is proposed.Similar to the method of causal precoder design,we derive the parametric representation of the non-causal precoder,and solve the optimal parameters under the condition of total power constraint or per-antenna power constraint.Experimental results show that SSIM-based non-causal weak zero precoding has better performance. |
