Previous Abstract Return to Session B5

**Wideband DOA Estimation Based on Quantum Charged System Search Algorithm**

*Hongyuan Gao, Guojian Zang, Yuwei Ma, Harbin Engineering University, China*

**Location:** Pavilion Ballroom West

Alternate Number 4

DOA (direction of arrival) estimation is an essential research direction in the field of array signals processing. In recent years, with the development of communication technologies, the wideband signals have a more and more important role in our daily lives. It can be applied to global position, missile guidance, aerospace engineering, satellite launching and other fields. Comparing with narrow signals, wideband signals have many obvious advantages. It can carry more information and have stronger anti-jamming ability against space noise. It is also easier for target detection, feature extraction and parameter estimation. Because the changing of the wideband signals frequency can’t be ignored, so we need transform the wideband signals from time domain to frequency domain to deal with it. CSM-ML is a normal method to deal with wideband signals, which is based on the coherent signal-subspace method (CSM) and maximum likelihood (ML) estimation. The CSM is to focus the signal space at the non-overlapping frequency point on the reference frequency points in the frequency domain. Then the DOA estimation can be carried out by using the ML estimation method. Through this method, the CSM-ML function of the wideband signals can be obtained, and the angles of the incident signals are estimated by finding the optimal solutions. However, the CSM-ML function is a multidimensional and nonlinear function. The process of finding the solution need optimize the CSM-ML function which takes the incident direction as variable. It has heavy calculation burden, high complexity, slow convergence and the local optimum solution is got in solving process easily. In order to efficiently resolve above problems about the CSM-ML function, a novel intelligent algorithm named quantum charged system search(QCSS) is proposed in the paper.

The QCSS is established for wideband DOA estimation by combining quantum computing and the charged system search (CSS) algorithm. The CSS was pioneered by A.Kaveh and S.Talatahari in 2010, and it is a novel intelligent optimization which is established for continuous optimization problems utilizing the governing Coulomb law from physics and the governing motion from Newtonian mechanics. In the CSS, each possible solution containing a number of decision variables represents a charged particle containing a number of electric charges, and their relationship describes the resultant force. When a solution has a better performance, the amount of the charge also is larger and it can have a stronger force on other particles. So the amount of the charge will be defined considering the objective function value. There are some rules developed to introduce the CSS. It involves the magnitude of the charge, the initial possible solution, the attract forces between two charged particles, the resultant electrical force, the velocity of each charged particles and so on. However, the wideband DOA estimation based on CSS still exists many problems, such as slow convergence rate, low accuracy. Quantum Computation Intelligence is inspired by quantum computation and quantum mechanism in physics. Quantum Computation is pioneered by Richard Feynman in 1982, and has made great progress now. Quantum charged system search algorithm combines quantum computation and charged system search algorithm. In the QCSS, the charged particles are quantized by designing a new coding method utilizing quantum computation. According to the research of quantum rotation angle and the resultant electrical force, design the updated method of the quantum rotation angle in QCSS. Then the charged particles can be evolved by the quantum rotation gate mechanism.

In computer simulation experiments, the incident signals are linear frequency modulation signals, and the antenna array is uniform array. We carry out four computer simulation experiments with PSO-CSM-ML, QPSO-CSM-ML, AP-CSM-ML and QCSS-CSM-ML. The first two simulation experiments depict the RMSE of two independent signals and coherent signals respectively. From the result of the simulation experiments, the performance of the four algorithms is similar in low SNR environment, but with the SNR increasing, the RMSE of QCSS-CSM-ML is smallest among the four algorithms So the wideband DOA estimation based on QCSS-CSM-ML has better convergent accuracy. The third simulation experiment depicts the RMSE of two wideband signal sources with the angle difference changing. From the result of the simulation experiment, with the signal difference increasing, the RMSE of QCSS-CSM-ML is always smallest. It can be concluded that the robustness of QCSS is the best, and it has the highest resolution ratio. The fourth simulation experiment depicts the fitness changing of three algorithms with the number of iterations increasing. From the result of the simulation experiment, QCSS-CSM-ML has converged about 20 iterations, and its fitness value is the highest. It is obviously that QCSS-CSM-ML has faster convergence speed and higher accuracy.

From the results of four computer simulation experiments, the performance of the QCSS is best. Comparing with other algorithms, the QCSS has faster convergence speed and higher accuracy, and it can avoid getting into local optimum solution and decrease the calculation amount. The wideband DOA estimation method based on QCSS-CSM-ML is superior to the previous wideband DOA estimation methods based on intelligence algorithms.

The QCSS decreases the calculation burden and improves the accuracy, meanwhile it proves the validity of the QCSS and offers a new idea to solve the problems of wideband DOA estimation. It promotes the development of quantum intelligence optimization algorithm, and proves the possibility of success that combines quantum computation and intelligence algorithm.

Previous Abstract Return to Session B5