In wireless communications, massive multiple-input multiple-output (MIMO) systems in which the base station (BS) is equipped with a large number of antennas can provide significant spectral and energy efficiency by means of simple signal processing. Therefore, massive MIMO is an attractive technology for next generation wireless communication systems and for green communications. There are two operation modes: time-division duplex (TDD) and frequency-division duplex (FDD) for massive MIMO systems, and the debate on the two modes is still discussed in the literature. The originator of massive MIMO systems, Marzetta, has presented the basic frame structures for both the TDD and the FDD modes. However, the basic frame structures do not carefully take the hardware impairments such as phase noise (PN) into account. In this paper, by taking the PN as the key hardware impairment into account, we recommend new and efficient frame structures for both the TDD and the FDD modes. The corresponding performance comparison between the two modes is also provided by using the data transmission efficiency (DTE) as the performance measurement. The comparison may help determine which mode should be used for massive MIMO systems.

This paper proposes a new symbol-based near maximum likelihood (ML) detection scheme for multiple-input multiple-output (MIMO) systems. The proposed scheme can provide not only near ML performance with low complexity but also the capability of unequal error protection (UEP), i.e., data from different antennas can have different reliability. Compared to the conventional symbol-based near ML detection scheme, simulation results demonstrate that the proposed scheme can provide lower complexity under the same bit error rate (BER). Moreover, different from the common UEP schemes, which provide the UEP by the transmitter side, the proposed scheme provides the UEP by the receiver side via the property of symbol-based detection. Accordingly, the proposed scheme may create a new degree of freedom for the UEP design.

Diagonal loading (DL) is one of the widely used techniques against the errors due to steering vector mismatch and finite sample effect. Recently, a variable loading (VL) has shown its advantages over the DL due to using different loading for each eigenvalue of the correlation matrix of the received array data rather than a fixed loading for all of the eigenvalues as the DL. In this paper, a novel robust beamforming method is presented. As compared to the DL and the VL, the weight vector of the proposed method is with a more general form of diagonal loaded correlation matrix. Simulation results show that the proposed method is more robust than the DL and the VL against the errors as mentioned above.

Many machine learning methods in Artificial intelligence literature, such as Neural Networks, Genetic Algorithms, SVM, and Case-Based Reasoning, have been applied to forecast the financial market with high irregularity and uncertatinty. Among them, S

When a user finds the answer using question answering system, sometimes the user may not be satisfied with answers. This is because it is difficult to deliver the intention of a user question to the system due to the lack of expression ability and polysemy and etc. So, the system must provide additional ways to search the correct answers. In this paper, we propose a new method to recommend questions that can satisfy the purpose of user question with high probability. We define the supplementary similarity to calculate the similarity between sentences which have a few common terms. And we define the fuzzy relation product operator to find the questions that are recommended when user is not satisfied with the answers. User can select the question among the list of recommended questions and has more chance to find the answers that the user can be satisfied with. We experiment and show that the performance of the system gets high as the recommendation is repeated.

A parallel-connected structure based on recursive nonsymmetric half-plane (NSHP) digital allpass filters is presented for designing two-dimensional (2-D) recursive doubly complementary (DC) filters. First, the theory of 2-D recursive digital allpass filters (DAFs) with NSHP support region for filter coefficients is developed. Then, the design problem is appropriately formulated to result in a simple optimization problem that minimizes the phase error in the least-squares (L2) sense with a closed-form solution. The novelty of the presented 2-D NSHP DAFs structure is that it possesses the advantage of better performance in designing DC filters over existing 2-D structures based on quarter plane (QP) DAFs and tremendously saves the computational complexity for designing sampling rate converters. Finally, a design example is provided for conducting illustration and comparison.

This paper deals with an eigenspace-based interference canceller under steering angle mismatch, correlated signals, and nonwhite noise. We construct a cost function consisting of two terms to be minimized. One represents the squared norm of the projection of the constraint vectors onto the noise subspace. The other one is related to the squared norm of the phase error vector due to steering angle error. An iterative algorithm is presented to solve the minimization problem. Simulations show the effectiveness of the proposed method.

Time series data, which are a series of one dimensional real numbers, have been studied in various database applications. We extend the traditional similarity search methods on time series data to support a multidimensional data sequence, such as a video stream. We investigate the problem of retrieving similar multidimensional data sequences from a large database. To prune irrelevant sequences in a database, we introduce correct and efficient similarity functions. Both data sequences and query sequences are partitioned into subsequences, and each of them is represented by a Minimum Bounding Rectangle (MBR). The query processing is based upon these MBRs, instead of scanning data elements of entire sequences. Our method is designed: (1) to select candidate sequences in a database, and (2) to find the subsequences of a selected sequence, each of which falls under the given threshold. The latter is of special importance in the case of retrieving subsequences from large and complex sequences such as video. By using it, we do not need to browse the whole of the selected video stream, but just browse the sub-streams to find a scene we want. We have performed an extensive experiment on synthetic, as well as real data sequences (a collection of TV news, dramas, and documentary videos) to evaluate our proposed method. The experiment demonstrates that 73-94 percent of irrelevant sequences are pruned using the proposed method, resulting in 16-28 times faster response time compared with that of the sequential search.

The database query optimizer requires the estimation of the query selectivity to find the most efficient access plan. For queries referencing multiple attributes from the same relation, we need a multi-dimensional selectivity estimation technique when the attributes are dependent each other because the selectivity is determined by the joint data distribution of the attributes. Additionally, for multimedia databases, there are intrinsic requirements for the multi-dimensional selectivity estimation because feature vectors are stored in multi-dimensional indexing trees. In the 1-dimensional case, a histogram is practically the most preferable. In the multi-dimensional case, however, a histogram is not adequate because of high storage overhead and high error rates.

A robust technique for adaptive array beamforming in the presence of steering vector error is proposed. It is based on the fact that an optimized steered-beam adaptive beamformer possesses the maximum output SNR when the steering vector coincides with the signal phase vector. The steering vector is adjusted iteratively to meet the actual signal phase vector. The resulting beamformer provides optimum performance in the sense of satisfying the single-look direction constraint without any tradeoff between the array performance and robustness. Thus the difficulties of using conventional algorithms can be overcome. Results of several computer simulations are given for illustration.>

Calibration-free bearing estimation algorithms for linearly periodic arrays with the presence of sensor positioning errors are investigated. The Toeplitz approximation method (TAM) is used to cope with the two-dimensional (2-D) sensor positioning errors which are assumed to be independent identically distributed and Gaussian in each dimension. After a Toeplitz covariance matrix is reconstructed by the TAM, the eigenstructure-based bearing estimation algorithms, such as MUSIC, can be used as though the array were linear and equally spaced without using any calibration process. To improve the effectiveness of the TAM, a modification for TAM is described. Several examples are provided to illustrate the effectiveness of the proposed methods.>

Estimating the two-dimensional (2-D) angle of arrival for radiating sources in a coherent environment is studied. The concept of spatial smoothing is first extended to a rectangular planar array, and a 2-D search function is formed to estimate the source directions. To avoid performing a 2-D search, an approach based on two one-dimensional (1-D) searches is also discussed. This approach uses rows and columns of the rectangular array to perform 1-D searches. To match the data obtained, a 2-D verification is then performed. Computer simulation results for both approaches based on the MUSIC method are presented.>

We present sectioning approaches for the implementation of two-dimensional recursive symmetric half-plane fiters. Two sectioning algorithms are described and compared with the direct implementation as regards computational requirements. An example of software implementation on a PRIME 750 is given.