Font Size: a A A

Research On Direction Finding Algorithms Of Non-circular Signals And Its Dsp Realization

Posted on:2010-03-30Degree:MasterType:Thesis
Country:ChinaCandidate:W RenFull Text:PDF
GTID:2198330332478524Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Spatial spectrum estimation for direction finding (DF) is an important branch of array signals processing, it has a great range of applications in the field of radar, sonar and telecommunication etc. In recent years more and more attention has been paid to utilizing the properties of signals to improve the performance of direction finding. For example, Modern telecommunication system widely uses modulated signal such as BPSK, MASK, AM and PAM, and its statistics are rotationally variant, therefore they are referred as non-circular signals. By employing its non-zero elliptic covariance property in DF, greater number of processed signals can be achieved comparing with the number of array elements, also the performance of direction finding could be enhanced. This thesis studies the non-circular signals direction finding, the main research work is detailed as follows:1. The estimation algorithm of the number of non-circular signals is studied. Firstly this paper studies the number estimator of signals in the context of the white Gaussian noise. The theory derives the choice of panel function, due to the the utilization of the non-circular property. Simulations suggest both algorithms have better performance. Secondly, this paper studies the number estimator of signals in the background of the colored noise. The non-circular property is utilized, based on Kullback information criterion and an eigenvalue calibration method, to provide two modified algorithm, i.e. NC-KIC and NC-CMDL. Simulations validate:NC-KIC produces better performance in low SNR, small number of snapshots; NC-CMDL is easier to realize, and yields better performance in the array extension. At last, this paper introduces an estimator of the number of signals when circular signals and noncircular signals impinge on the antenna array.2. The direction finding algorithm based on MUSIC and ESPRIT algorithms for noncircular signals is studied. This paper firstly introduces the NC-MUSIC (Non-circular MUSIC) algorithm, and the usage of the Euler's equation transforms the calculation of NC-MUSIC to real number domain from complex number domain (Euler NC-MUSIC), its computational load become less.Simulations suggest that NC-MUSIC algorithm can find directions for more signals than the number of antennas, and the performance is improved. Euler NC-MUSIC shares the same performance of the NC-MUSIC algorithm. Then this paper introduces the NC-ESPRIT (non-circular ESPRIT) algorithm of direction finding for non-circular signals. With the utilization of unitary transformation, the calculation of NC-ESPRIT is transformed to real number domain from complex number domain (Unitary NC-ESPRIT), and its computational load also becomes less. Simulations show that NC-ESPRIT algorithm can also find directions for more signals than the number of antennas, and the performance is improved. Also unitary NC-ESPRIT shares the same performance of the NC-ESPRIT algorithm.3. The 2-D dirction finding algorithm for non-circular signals is studied. Based on uniform parallel array, this paper firstly introduces a 2-D ESPRIT direction finding algorithm. Then a 2-D extended rank reduce algorithm for non-circular signals (2D-NC-ERARE) is introduced, which uses the noncircular property of the signals. Simulations prove that the 2D-NC-ERARE algorithm can estimate larger number of signals than the array elements, the azimuth and elevation are paired automatically, and this algorithm also yields better performance.4. The direction finding algorithm when both non-circular and circular signals impinge is studied. Aiming at the problem of both circular and noncircular signals impinge on the antenna array, an original algorithm (Noncircular and circular MUSIC, NCAC-MUSIC) utilizes the non-circular property is introduced firstly. By introducing the propagator method, the modified algorithm (modify NCAC-MUSIC, MNCAC-MUSIC) alleviates the heavy computation of the eigenvalue decomposition. Simulations suggest that this modified algorithm share the same performance with the NCAC-MUSIC algorithm when the complexity of the algorithm is reduced.5. Based on ADSP TS201, the algorithm that handles the low computational load when both non-circular and circular signals impinge has been coded. A series of actions have been taken to optimize the coding, it also demonstrate the process and shows the results.
Keywords/Search Tags:estimation of the direction of arrival, non-circular signals, colored noise, estimation of the number of signals, MUSIC algorithm, ESPRIT algorithm, propagator method
PDF Full Text Request
Related items