IEEE Transactions on Signal Processing

Dựa trên phương pháp không gian nhiễu tối thiểu (MNS) được giới thiệu trước đây trong bối cảnh nhận dạng kênh mù, không gian nhiễu tối thiểu tổng quát (GMNS) được đề xuất trong bài báo này cho xử lý mảng, mở rộng MNS liên quan đến việc chỉ có một số lượng cố định các đơn vị tính toán song song. Các thuật toán theo lô và thích ứng khác nhau sau đó được giới thiệu để tính toán nhanh và song song các không gian tín hiệu (chính) và nhiễu (phụ). Độ phức tạp tính toán của GMNS và độ chính xác ước lượng liên quan của nó được điều tra thông qua các thí nghiệm mô phỏng và một thí nghiệm thực tế trong thiên văn vô tuyến. Kết quả cho thấy GMNS đại diện cho một sự trao đổi tuyệt vời giữa lợi ích tính toán và độ chính xác ước lượng không gian, so với một số phương pháp không gian tiêu chuẩn.

The transmitter/receiver diversities in p-input-q-output multiple-input multiple-output (MIMO) channels will play a key role in future high-rate wireless communication. The major challenge in MIMO signal recovery is the mitigation of the inevitable intersymbol interference (ISI) and interchannel interference (ICI) in multipath MIMO channels. The Bezout system theory provides a simple solution for ISI/ICI cancellation and, thus, can serve as a powerful mathematical foundation for MIMO systems. If pq and the MIMO system H(D) is left coprime, the system is perfectly recoverable via a Bezout precoder, assuming the channel information is available at the transmitter. For a robust channel design, a quantitative analysis of signal-to-noise ratio (SNR) of the optimal Bezout equalizer/precoder can be derived from the Bezout null space. Compared with the Alamouti(see IEEE J. Select. Areas Commun., vol.16, p.1451-58, Oct. 1998) space-time coding, the Bezout precoder is shown to be more appealing in slowly time-varying channels, where the feedback of the channel information induces minor overhead. The Bezout system can work well jointly with the space-time block coding (STBC). The STBC can be adopted to artificially construct a larger q'/spl times/p' (q'>p') virtual transfer function without the channel information. A necessary and sufficient condition is provided for perfect recoverability (PR) for the STBC-induced virtual MIMO systems. Based on the virtual transfer function, both channel-independent and channel-dependent precoding strategies are feasible. Via a singular value decomposition (SVD) analysis, the Bezout precoder can be shown to outperform the orthogonal frequency division multiplexing (OFDM) precoder in bit-error-rate (BER), transmission rate, and receiver implementation. The interplay of the two design parameters (N and /spl rho/) is analyzed to provide a simple guideline for an optimal configuration. The combined STBC and Bezout equalization techniques offer a broader spectrum of transceiver system configuration to achieve optimal design tradeoff among BER, transmission rate, and implementation complexity.

In recent years, sparse subspace tracking has attracted increasing attention in the signal processing community. In this paper, we propose a new provable effective method called OPIT (which stands for Online Power Iteration via Thresholding) for tracking the sparse principal subspace of data streams over time. Particularly, OPIT introduces a new adaptive variant of power iteration with space and computational complexity linear to the data dimension. In addition, a new column-based thresholding operator is developed to regularize the subspace sparsity. Utilizing both advantages of power iteration and thresholding operation, OPIT is capable of tracking the underlying subspace in both the classical regime and high dimensional regime. We also present a theoretical result on its convergence to verify its consistency in high dimensions. Several experiments are carried out on both synthetic and real data to demonstrate the performance of OPIT.

Canonical Polyadic (CP) decomposition is a powerful multilinear algebra tool for analyzing multiway (a.k.a. tensor) data and has been used for various signal processing and machine learning applications. When the underlying tensor is derived from data streams, adaptive CP decomposition is required. In this paper, we propose a novel method called robust adaptive CP decomposition (RACP) for dealing with high-order incomplete streaming tensors that are corrupted by outliers. At each time instant, RACP first performs online outlier rejection to accurately detect and remove sparse outliers, and then performs tensor factor tracking to efficiently update the tensor basis. A unified convergence analysis of RACP is also established in that the sequence of generated solutions converges asymptotically to a stationary point of the objective function. Extensive experiments were conducted on both synthetic and real data to demonstrate the effectiveness of RACP in comparison with state-of-the-art adaptive CP algorithms.

In this article, we propose a novel algorithm, namely PETRELS-ADMM, to deal with subspace tracking in the presence of outliers and missing data. The proposed approach consists of two main stages: outlier rejection and subspace estimation. In the first stage, alternating direction method of multipliers (ADMM) is effectively exploited to detect outliers affecting the observed data. In the second stage, we propose an improved version of the parallel estimation and tracking by recursive least squares (PETRELS) algorithm to update the underlying subspace in the missing data context. We then present a theoretical convergence analysis of PETRELS-ADMM which shows that it generates a sequence of subspace solutions converging to the optimum of its batch counterpart. The effectiveness of the proposed algorithm, as compared to state-of-the-art algorithms, is illustrated on both simulated and real data.

In estimation, the misspecified Cramer–Rao bound (MCRB), which is an extension of the well-known Cramer–Rao bound (CRB) when the underlying system model is misspecified, has recently attracted much attention. In this paper, we introduce a new interpretation of the MCRB, called the generalized MCRB (GMCRB), via the Moore–Penrose inverse operator. This bound is useful for singular problems and particularly blind channel estimation problems in which the Hessian matrix is noninvertible. Two closed-form expressions of the GMCRB are derived for unbiased blind estimators when the channel order is misspecified. The first bound deals with deterministic models where both the channel and unknown symbols are deterministic. The second one is devoted to stochastic models where we assume that transmitted symbols are unknown random variables i.i.d. drawn from a Gaussian distribution. Two case studies of channel order misspecification are investigated to demonstrate the effectiveness of the proposed GMCRBs over the classical CRBs. When the channel order is known or accurately estimated, both generalized bounds reduce to the classical bounds. Besides, the stochastic GMCRB is lower than the deterministic one, especially at high SNR.

Bài báo này xem xét việc phân tách mù các nguồn không ổn định trong trường hợp không xác định, khi số nguồn nhiều hơn số cảm biến. Một khung tổng quát cho vấn đề này là làm việc trên các nguồn mà có tính phân tán trong một miền biểu diễn tín hiệu nào đó. Gần đây, hai phương pháp đã được đề xuất liên quan đến miền thời gian-tần số (TF). Phương pháp đầu tiên sử dụng các phân phối thời gian-tần số bậc hai (TFDs) và một phương pháp phân cụm, và phương pháp thứ hai sử dụng một TFD tuyến tính. Cả hai phương pháp này đều giả định rằng các nguồn là tách biệt trong miền TF; tức là, có tối đa một nguồn có mặt tại một điểm trong miền TF. Trong bài báo này, chúng tôi nới lỏng giả định này bằng cách cho phép các nguồn có thể không hoàn toàn tách biệt trong miền TF đến một mức độ nhất định. Cụ thể, số lượng nguồn có mặt tại một điểm nhỏ hơn số lượng cảm biến. Việc phân tách vẫn có thể đạt được nhờ vào việc chiếu subspace cho phép chúng tôi xác định các nguồn có mặt và ước lượng các giá trị TFD tương ứng của chúng. Đặc biệt, chúng tôi đề xuất hai thuật toán dựa trên subspace cho các nguồn không tách biệt trong TF: một sử dụng TFD bậc hai và một sử dụng TFD tuyến tính. Một đóng góp khác của bài báo này là quy trình ước lượng mới cho ma trận trộn. Cuối cùng, hiệu suất số học của các phương pháp đã đề xuất được cung cấp, làm nổi bật sự gia tăng hiệu suất của chúng so với các phương pháp hiện có.

Multiple-antenna channel coding for orthogonal frequency-division multiplexing (OFDM) transmission over dispersive channels is reconsidered because with frequency interleaving, the effective channel characteristic across subcarriers is rather fast fading. The channel does not comply with the quasistatic model widely assumed for space-time trellis codes (STCs). For that reason, we first study the ideal fast-fading multiple transmit and receive antenna channel and then compare the performance of STCs with that of bit-interleaved coded modulation in fast fading. Mutual information of the ergodic channel is evaluated for numerous modulation scenarios, and capacity comparisons generate guidelines on how to jointly adjust coding rate and modulation cardinality. Bit-based coding offers large flexibility in rate adaptation, and simulation results show that it outperforms STCs in ideal fast fading and, finally, in a realistic OFDM application as well.

In this paper, we examine multicarrier transmission over time-varying channels. We first develop a model for such a transmission scheme and focus particularly on multiple-input multiple output (MIMO) orthogonal frequency division multiplexing (OFDM). Using this method, we analyze the impact of time variation within a transmission block (time variation could arise both from Doppler spread of the channel and from synchronization errors). To mitigate the effects of such time variations, we propose a time-domain approach. We design ICI-mitigating block linear filters, and we examine how they are modified in the context of space-time block-coded transmissions. Our approach reduces to the familiar single-tap frequency-domain equalizer when the channel is block time invariant. Channel estimation in rapidly time-varying scenarios becomes critical, and we propose a scheme for estimating channel parameters varying within a transmission block. Along with the channel estimation scheme, we also examine the issue of pilot tone placement and show that in time-varying channels, it may be better to group pilot tones together into clumps that are equispaced onto the FFT grid; this placement technique is in contrast to the common wisdom for time-invariant channels. Finally, we provide numerical results illustrating the performance of these schemes, both for uncoded and space-time block-coded systems.