Research On Channel Modeling And Signal Detection Based On Quantum Algorithm

The rapidly increasing number of users and the amount of communication data put forward higher requirements on the performance and efficiency of the communication system.However,the traditional communication signal processing method has the contradiction of mutual limitation of complexity and performance,which has a certain impact on the requirements of communication system such as high reliability and low delay.Quantum computing is a new computing method based on the theory of quantum mechanics.Due to its superposition principle and parallel computing method,it can show great potential in some specific problems beyond the classical computing method.Therefore,for the development of 5G and future communication systems,the high computing efficiency of quantum computing can bring new ideas to signal processing and mass data processing.Based on quantum machine learning and intelligent communication,channel modeling and signal detection schemes based on quantum computing are proposed in this paper.Firstly,the channel modeling scheme based on quantum generative adversarial network(QGAN)is studied in this paper,and QGAN algorithm is proposed combining the generative adversarial network(GAN)with the quantum Born machine model.Taking the small-scale fading channels as example,the model building,algorithm design,programming implementation and experimental analysis are carried out.The process of cost function gradient descent and the K-L divergence between the generation distribution and the target channel sample are analyzed in the experiment.The innovation of this scheme is to quantize the generative network part of GAN network model.According to the structure of quantum Born machine model,a multi-layer quantum circuit network model is built to simulate random channels.In this scheme,compared with the traditional GAN algorithm,the scheme based on QGAN network model can improve the time complexity from exponential level to polynomial level,which shows the improvement in complexity performance of the channel modeling scheme based on QGAN algorithm.On the basis of the above research,quantum machine learning algorithm is introduced into the field of wireless communication,and a signal detection scheme based on quantum neural network is proposed.Combining the concept of quantum computing and deep learning,the parameterized quantum circuit structure is introduced into the classical deep neural network(DNN),and the structure of quantum deep neural network(QDNN)is constructed.The quantum deep neural network is applied on signal detection scheme.The OFDM system is taken as an example for simulation experiments.For different number of pilot signals,the bit error rate curves of different signal detection schemes are drawn for performance comparison.Experiments show that the signal detection method based on QDNN can solve the channel distortion problem.The performance of QDNN is best when the number of pilots is insufficient,and the performance is equivalent to MMSE scheme when the number of pilots is sufficient.Compared with the signal detection method based on traditional deep learning,the signal detection scheme based on QDNN gives full play to the good parallel computing capability of quantum computing.The bit error rate under this signal detection scheme is lower,which fully verifies the feasibility and superiority of the scheme.