Quantum Computing Based Algorithm Research On Information Processing In Wireless Communication System

Quantum computation has outstanding advantages in parallel computing,uncertainty com-puting and storage performance,making it the strongest candidate to replace the classical Von Neumann acrchitecture computer and to solve Moore's Law failure.In recent years,with the rapid development of wireless communication,Internet and other fields,this results in the ex-plosive growth of the size and dimension of data in wireless communication system.Thus,how to quickly and efficiently store and process data in the system is an urgent problem to be solved.Information processing of wireless communication system based on quantum computa-tion is a crossover research direction of quantum computing and information processing,which brings the new direction and perspective to the development of future wireless communication technologies,and also points out a potential application direction for quantum computation.Moreover,the existing research has already shown huge advantages and broad prospects of this direction.However,as a whole,this field is still in the initial stage of research,and there are many research gaps which need immediate attention.This article studies five sub-problems related to information processing in wireless communication system based on the unique ad-vantages of quantum computation,including three open problems and two multi-user detection related problems.These five sub-problems not only make up for the application limitations of the current research,but also point out an important path to be explored for the application of quantum computing in wireless communication systems.The innovative results of this article are summarized as follows:1.Propose the quantum algorithm for multiple signal classification.First of all,quantum singular value estimation is used to realize the subroutine of the spatial covariance matrix reconstruction,in which the realization process of the left and right singular vector transformation is designed.Then,we propose two quantum algorithms for the space covariance matrix eigen-decomposition,which are the eigen-decomposition based on the density matrix exponentiate and the variational quantum state eigensolver.Finally,the quantum subroutine of the space arrival direction search is implemented,in which the core quantum labeling operation can be further improved and extended to existing quantum artificial intelligence algorithms.Theo-retical analysis and experimental verification show the superiority of our quantum algorithm compared to traditional counterpart.2.Propose the quantum algorithm for spectral regression.This algorithm can be used in the information preprocessing module of wireless communication systems to reduce the di-mensionality of large-scale data.This paper first designs a quantum algorithm to construct the adjacency graph and matrix which is often neglected by many quantum algorithms.Then,based on whether there is the label in the training sample data,we propose two quantum algorithms to obtain the principal generalized eigenvectors,which are the quantum Schmidt orthogonaliza-tion algorithm and the quantum variational generalized eigen-decomposition algorithm.Finally,this paper realizes the transformation matrix based on the quantum singular decomposition and the singular vectors transformation process,which can transform high-dimensional data to low-dimensional data.Compared with the classical spectral regression algorithm,theory and ex-periment prove that our quantum algorithm can achieve a polynomial or even an exponential speedup.3.Design quantum algorithms to realize von Neumann entropy.First,the DQC1 model and quantum block-encoding algorithm are introduced to provide algorithm tools for subsequent use.Then,three quantum algorithms are proposed to calculate the approximate value of von Neumann entropy,which are the approximation methods based on the Taylor series truncation,the Chebyshev polynomial approximation and the variational quantum state eigensolver.Com-pared with classical counterparts,the above three algorithms can at least provide a polynomial speedup.4.Design the quantum algorithm to realize MMSE-based massive MIMO uplink signal de-tection.A quantum algorithm is designed to estimate the transmitted signal.In view of the difficulty in the subsequent traditional operations of the quantum state form of the transmit-ted signal,we have implemented an effective quantum state information extraction technology which can extract the phase and amplitude of the quantum state.Finally,in order to verify the applicability of the quantum algorithm,the distribution of parameters is deeply analyzed,which affect the quantum algorithm in practical applications,we then prove the advantages of our quantum algorithms compared to classical counterpart.5.Implement CDMA signal demodulation based on quantum approximate optimization al-gorithm.The specific form of CDMA signal demodulation based on maximum a posteriori estimation is analyzed in the case of the prior uniform distribution,and it can be transformed into a Hamiltonian model which is suitable for quantum approximation optimization algorithm.In addition,we improve the above form and propose a maximum posteriori based CDMA signal demodulation algorithm with the non-uniform prior distribution.Finally,the above algorithm is verified on the noisy intermediate-scale(NISQ)devices.Compared with the traditional opti-mization program,the quantum algorithm in this paper has higher efficiency and lower resource usage.