Research On Channel Estimation And Signal Detection Massive Algorithm For MIMO Systems Based On Parallel Factor Model

Massive multiple-input multiple-output is one of the key technologies of the fifth-generation mobile communication technology.By increasing the number of antennas,energy efficiency,spectrum efficiency,and communication capacity in the system can be greatly improved.However,massive MIMO technology still faces challenges such as channel estimation,signal detection,and pilot pollution.Although the pilot-based channel estimation technique is easy to implement,such an algorithm requires a large amount of pilot overhead,and also generates pilot contamination problems and reduces channel estimation performance.While conventional signal detection techniques require known channel state information,they are inevitably subject to channel estimation performance.In order to improve the channel estimation accuracy and signal detection performance of massive MIMO technology,this paper makes full use of the multi-dimensional characteristics of the signal,and uses Parallel Factor(PARAFAC)decomposition method to model the signal.The characteristics of the massive MIMO system are used as the constraints in the iterative process of fitting the parallel factor model to study the channel estimation and signal detection of the single-cell massive MIMO system.Its main research contents and innovations are as follows:1.In order to reduce the pilot overhead,the received signal at the base station is constructed as a multi-dimensional tensor.A channel estimation and signal detection scheme dominated by a parallel factorization model is proposed.The proposed scheme can jointly complete channel estimation and signal detection by transmitting the pilot symbols once.The simulation results show that compared with the traditional pilot-assisted channel estimation and linear signal detection methods,the accuracy of channel estimation and the performance of bit error rate can be greatly improved2.Aiming at the problem of slow convergence and high complexity of Alternating Least Squares(ALS)for traditional fitting parallel factor model,this paper proposes a constrained bilinear alternating least squares(CBALS)channelestimation algorithm.The proposed algorithm utilizes the asymptotic orthogonality of massive MIMO channel matrices and incorporates it as a constraint into the iterative process of the fitting algorithm.The simulation results show that compared with the existing pilot-based channel estimation method,the proposed algorithm significantly improves the accuracy of channel estimation;Compared with Bilinear Alternating Least Squares(BALS)with traditional fitting parallel factor model,the proposed method has better estimation performance with lower transmission power,and has faster convergence speed and lower complexity.3.In order to reduce the influence of channel estimation on signal detection accuracy,this paper constructs a three-order tensor model for the base station receiving signal,and proposes a signal detection algorithm for Bilinear Alternating Least Squares Projection(BALSP).The algorithm can complete signal detection with unknown channel state information.In the proposed algorithm,the source matrix sent by the user is used as a loading matrix in the parallel factor model,and the constant model property of the source matrix is introduced as a constraint condition,and the source matrix is reconstructed in the iterative process.On this basis,the unique decomposition conditions of the massive MIMO source constant modulus constrained parallel factor model are derived.The simulation results of the proposed algorithm show that compared with the traditional Bilinear alternating least squares algorithm of the parallel factor model,it not only improves the convergence speed of the algorithm,but also has better detection performance.Compared with the traditional linear signal detection method MMSE,channel estimation is not required,which significantly improves the bit error rate performance.