� ��� However, estimates of the required HOS often introduce larger variances; Moreover, if the sources happen to have null cumulants of the chosen order, they remain unaccounted for in the algorithm. of subspace-based estimation techniques applied to the problem of the ... since (19) with (20) where denotes the rank of . More specifically, when fourth-order cumulants are considered, a different target-matrix (e.g., a diagonal slice [2] or a contracted quadricovariance. We derive an asymptotic expression for the upper bound on the quality of the estimated signal in terms of the output signal-to-interference ratio. The final two chapters deal with likelihood ratio statistics, maximum likelihood estimation and the effect on inferences of conditioning on ancillary or approximately ancillary statistics. endobj Simulation results are presented and comparisons are 31 0 obj endobj ), y-axis : DOA variance in deg**2. Simulation results are presented and comparisons are made to show that the performance of the fourth-order cumulant-based methods (beamforming, MUSIC) is superior to that of their equivalent autocorrelation-based methods when the additive noise sources are colored Gaussian with unknown correlation matrix. High-resolution direction finding algorithms using higher-order Asymptotic performance results for the output SIR are developed. endobj endobj endobj This is discussed via the notion of fourth-order signal subspace. It is shown however that if the sources' power is allowed to vary with time, unique. <> <> � ��� operators on the signal and noise subspaces. <> High-resolution direction finding algorithms using higher-order cumulants of the array data is addressed. It is well known that for stationary, i.i.d. We present the asymptotic performance analysis of the interpolated root-MUSIC and manifold separation (MS) techniques for direction-of-arrival (DOA) estimation in arbitrary non-uniform sensor arrays. Since the effective array aperture is extended according to the higher-order steering vector expression, a virtual spatial covariance matrix is derived using the forth-order cumulants. x�3R��2�35W(�*T0P�R0T(�Y@���@QC= P A�J��� �14Up�W� Without the ability to look at each antenna simultaneously (which would be the case if one were to use multiple receivers, also known as N-channel DF) more complex operations need to occur at the antenna in order to present the signal to the receiver. x�3R��2�35W(�*T0P�R0T(�Y@���@QC= P A�J��� �Tp�W� Int. /Contents 20 0 R>> One of them is the proximity of signal sources and under the condition of the colored noise. We also introduce two blind algorithms for signal separation using only fourth-order cumulants: one is asymptotically optimal and the other is suboptimal but significantly simpler. /Contents 8 0 R>> However, estimates of the required HOS often introduce larger variances; Moreover, if the sources happen to have null cumulants of the chosen order, they remain unaccounted for in the algorithm. methods (beamforming, MUSIC) is superior to that of their equivalent � ��� Wählen Sie ein Land/eine Region für Ihren Einkauf. The estimation performance, however, can be improved. easily be reformulated using fourth-order cumulant matrices instead of Subsequent chapters deal with asymptotic approximations based on Edgeworth expansion and saddlepoint expansion. Results are shown for both DOAs in terms of the root mean square estimation error (RMSE) vs. µ, for SOS-based, 4th-order-cumulants-based (contracted quadricovariance, ... ESPRIT, MUSIC, minimum variance), therefore, they are essentially based on second-order statistics [2][3]. � ��� In this way, the redundant information contained in the cumulants are removed, and the computational complexity is reduced for the dimension of the virtual covariance matrix is decreased, which is equal to the number of the virtual array elements. Moreover, we derive closed-form expressions of asymptotic distributions of the estimation errors which are optimized by those cost functions. 28 0 obj Blind array processing techniques have been gaining much interest in the last several years. In blind array processing the structure of the array manifold is assumed to be unknown and needs to be estimated from the available data. quadricovariance outperforms the diagonal slice in all respects. 1 0 obj sources the answer is negative due to the inherent unitary matrix ambiguity of the output second order information. /Contents 26 0 R>> <>>>] It is a Wir verwenden Cookies und ähnliche Tools, um Ihr Einkaufserlebnis zu verbessern, um unsere Dienste anzubieten, um zu verstehen, wie die Kunden unsere Dienste nutzen, damit wir Verbesserungen vornehmen können, und um Werbung anzuzeigen. <> that the harmonic decomposition methods (signal and noise subspace) can <> <> <> direction of arrival of plane waves. cumulants of the array data is addressed. scenarios, the proposed methods has significantly better performance endobj © 2008-2020 ResearchGate GmbH. This approach to DF offers some advantages and drawbacks. endstream endstream The paper is not intended as being a review, but other related approaches based on higher-order statistics are briefly reported. <> 26 0 obj /Contents 36 0 R>> Multiple Signal Classification (MUSIC) algorithm is a well-known method in direction of arrival (DOA) estimation for having high resolution capability and therefore it's statistical performance have been investigated for different situations. The COP solution provides a weight vector to the look direction such that it is constrained to the th-order source-signal subspace when the look direction is in alignment with the true DOA; otherwise, it is constrained to the th-order noise subspace. stream x�3R��2�35W(�*T0P�R0T(�Y@���@QC= P A�J��� �1Vp�W� endobj endstream � ��� Our target matrix is based on subtracting the observations' correlation matrix from Hessians of their second characteristic function, evaluated at several "processing points". 16 0 obj stream endobj This dissertation, "New Direction Finding and Beamforming Algorithms for Sensor Arrays With Uncertainties" by Bin, Liao, 廖斌, was obtained from The University of Hong Kong (Pokfulam, Hong Kong) and is being sold pursuant to Creative Commons: Attribution 3.0 Hong Kong License. This is followed by an examination of the effect of making a polynomial transformation of the original variables. This book provides a systematic development of tensor methods in statistics, beginning with the study of multivariate moments and cumulants. � ��� endstream 23 0 obj yield a noniterative estimation algorithm when a uniform linear array is the contracted quadricovariance. In this paper we propose a new target-matrix to substitute HOS as the input to MUSIC. [10] C. Berthelot, E. Benzoni, private communication. 10 0 obj : Single Source at 0 dB, y-axis : DOA variance in deg**2. <> Ideas from subspace and covariance matching methods are combined to <> Entdecken Sie jetzt alle Amazon Prime-Vorteile. Because of their importance in statistical theory, invariant functions of the cumulants are studied in some detail. Lieferung verfolgen oder Bestellung anzeigen, Recycling (einschließlich Entsorgung von Elektro- & Elektronikaltgeräten). x�3R��2�35W(�*T0P�R0T(�Y@���@QC= P A�J��� �1Wp�W� Join ResearchGate to discover and stay up-to-date with the latest research from leading experts in, Access scientific knowledge from anywhere. endstream /Contents 30 0 R>> Geben Sie es weiter, tauschen Sie es ein, © 1998-2020, Amazon.com, Inc. oder Tochtergesellschaften. x�3R��2�35W(�*T0P�R0T(�Y@���@QC= P A�J��� �14Sp�W� subspace-based technique, as it uses a series expansion of projection <> /Contents 12 0 R>> Conventional array processing algorithms such as MUSIC assume that the structural information of the array manifold is known. elements in time and space, and µ is a parameter for controlling the noise power. endobj 36 0 obj A novel forth-order multiple signal classification (MUSIC) algorithm was proposed based on the virtual array technique. Conference Record. <> This implies that in such In its classical form, the multiple signal classification (MUSIC) algorithm applies eigen-decomposition to the estimated correlation matrix of the noisy sensors' outputs in order to estimate the directions of arrival (DOAs) of several sources. <> <>/Group <> Realizing the difficulty in identifying stationarity in the received sensor signals, the estimate of the noise-eliminated local th-order cumulant matrix is marginalized over various possible stationary segmentations, for a more robust DOA estimation. than usually believed. Performance analysis of root-MUSIC-based direction-of-arrival estimation for arbitrary non-uniform a... A subspace method for direction of arrival estimation of uncorrelated emitter signals, Conference: Acoustics, Speech, and Signal Processing, 1992. However, when the noise is spatially correlated (and / or has different variances at different sensors), the correlation-based MUSIC cannot produce consistent DOA estimates, unless the noise covariance is known in advance. endobj Statistics and Applied Probability, Chapman and Hall, 1987. high-order statistics", IEEE SP, 39(9), pp. More than 200 Exercises are provided to illustrate the uses of tensor methodology. In this paper, we focus on the use of local second and fourth order cumulants ( , 2), and the proposed algorithms when outperformed the KR subspace-based algorithms and also the 4-MUSIC for globally non-stationary, non-Gaussian synthetic data and also for speech/audio in various adverse environments. stream endobj 1977 11th Asilomar Conference on. 7 0 obj endstream x�3R��2�35W(�*T0P�R0T(�Y@���@QC= P A�J��� �1�Pp�W� Higher-order cumulants are usually advocated for in presence of additive Gaussian noise, but other potential advantages often are overlooked : non-Gaussian sensor noise cancellation, "more sources than sensors" ability, blind identification.