Search

Learning a good graph similarity matrix in data clustering is very crucial. The goal of clustering is to construct a good graph similarity matrix such that the similarity of points between the same classes is largest, and the similarity of points between different classes is smallest. In this paper, a more efficient subspace segmentation approach to learn a similarity matrix with large spectral gap is proposed. In our model, a robust self-representation coefficient matrix is learned by utilizing the Schatten- ${p}$ norm instead of the conventional rank function. Besides, the fast block-diagonal structure of the coefficient representation matrix is enhanced by learning and optimizing the co-association matrix with the soft label of clustering results simultaneously in a unified framework. The affinity graphs constructed in this paper can clearly reveal the intrinsic structures of the data sets. Extensive experiments on the real data sets demonstrate that our proposed method can perform better than the state-of-the-art methods.

The short-term prediction of EDH time series plays an exceedingly important role in several fields such as communications, navigation and so on. In this paper, an application of autoregressive integrated moving average (ARIMA) for short-term forecasting of EDH time series is presented. In order to obtain the EDH, a body of sensors such as air temperature, relative humidity and pressure sensors were installed at different height based on the tower platform. EDH was calculated according to Debye theory and a log-squares curve fit. The comparison showed that the predicted EDH values were in good agreement with the measured values. It also indicates that ARIMA provides promising results for short-term prediction of EDH in the experiment.

Traditional low-rank representation (LRR)and variant algorithms based on LRR have been proved as effective methods for subspace clustering tasks in machine learning. However, the majority of the existing LRR-related algorithms relax the rank minimization problem by the trace norm minimization. It is rarely accurate to get the optimal solution guaranteed by the definition. Meanwhile, these methods can not well reveal the internal structure of the classification. In this paper, we propose an efficient and robust subspace clustering method to overcome these drawbacks. A robust self-representation coefficient matrix is learned by utilizing the Schatten- $p$ norm instead of the conventional rank function. Besides, the strongest block-diagonal structure of the coefficient representation matrix is enhanced by learning and optimizing the co-association matrix with the soft label of clustering results simultaneously in an unified framework. The affinity graphs constructed in this paper can clearly reveal the intrinsic structures of the data sets. Extensive experiments on the real data sets demonstrate that our novel proposed method can perform more effective than the state-of-the-art methods.

The nonlinear effect of satellite's high power amplifiers causes the navigation signal distortion of limited bandwidth, which would affect signals in the power spectrum characteristic. This paper gave a signal model based on delay-multiplexing, improved fitting of nonlinear analyzing model (SAM model) on navigation signal. A new method of evaluating nonlinear distortion was presented, which was based on the signal in-band consisted of the effective signal and the jamming signal, and used the maximal effective output power as the evaluating characteristic. The influence of nonlinear distortion was simulated and analyzed on the fitting model in this paper. Further results showed that the delay-multiplexing be better than the direct-multiplexing in improving the effect of nonlinear distortion.

In this paper, we propose a precorrelation direction lock loop (DiLL) scheme and a new modification factor estimation method for the global navigation satellite system (GNSS) receivers. The proposed scheme executes spatial correlation before code demodulator and does not need to do code demodulation for every antenna element. It can decrease the number of code demodulator to 3 regardless of the array element number while the postcorrelation DiLL scheme requires as many as array elements. Moreover, performance can be improved for immunity of undesired signals, multipath and interference, for example, from other different direction of arrivals compared with the postcorrelation DiLL scheme. The new modification factor estimation method is based on the prompt code branch and estimated angle outputs, which are suitable for reality environments. Theory analysis and simulation results show that two schemes are equivalent when the code tracking loop is working on the ideal tracking state under interference-free environment. The proposed scheme can also be used for DS/CDMA systems.

Signal waveform is an important factor to the navigation system such as tracking precision, anti-jamming and anti-multipath. Prolate Spheroidal Wave Functions (PSWF) are widely used in Ultra Wide Band (UWB) communication area. They are complete orthogonal set in the time domain that maximizes energy in the specified bandwidth. This paper presents PSWF bases method and orthogonal decomposition of waveforms. By optimizing coefficients, we can obtain optimum waveforms to minimize target functions with constrains. It presents a new method for waveform design of the navigation signals.