Research And Implementation Of Signal Detection For Power Line In MIMO System

MIMO technology can realize a higher rate of data transfer in power line communication（PLC）,and optimization of signal detection algorithm can efficiently enhance the reliability and effectiveness of MIMO system.However,the existing signal detection algorithm has the problem of high complexity or insufficient performance.At the same time,the signal detection based on machine learning has been gradually applied to MIMO system and achieved a measure of success,but there are still some deficiencies.In view of the above problems,the main solutions proposed in this thesis are as follows:1.Power line channel modeling is introduced in this thesis,and then some classical detection algorithms in MIMO system are analyzed and compared.Based on parallel interference cancellation（PIC）,which has good performance and relatively small time delay,an improved PIC algorithm is proposed.The improved algorithm reduces the computational complexity by merging the common parts of sub-detection and substituting the estimated results of sub-detection into the subsequent sub-detection in order to improve their performance.Simulation results show that the performance of the improved algorithm is about 0.5d B higher than that of the traditional PIC algorithm when bit error rate is10^4-under the QPSK modulation mode in 2×2 MIMO-PLC system.2.The Det Net signal detection algorithm using neural network with great performance and the improved Sc Net algorithm based on Det Net are introduced.On this basis,the improved PIC algorithm is combined with Sc Net algorithm,and an improved PIC-Sc Net algorithm is proposed.Compared with Sc Net algorithm,the improved PIC-Sc Net algorithm can achieve better performance with fewer network layers and faster convergence speed.At the same time,the bit error rate performance of the improved algorithm can be simulated under a non-ideal channel.The results show that when the normalized mean square error of the estimated channel frequency response is less than10^4-,it is close to the performance curve of the improved algorithm under ideal channel.3.A reliable MIMO-PLC system is designed and then the improved PIC-Sc Net algorithm is applied to the system after fixed-point simulation,which is realized and simulated with field-programmable gate array（FPGA）.Then apply the PIC-Sc Net that is fixed-point simulated to the MIMO-PLC system and realize the PIC-Sc Net with FPGA.Finally,the bit error rate on the received data from the system is analyzed.The results show that the bit error rate is consistent with the bit error rate of the received data that is fixed-point simulated.By that way the correctness of the FPGA design and the feasibility of the improved algorithm are verified.