Search

The joint spatial division and multiplexing (JSDM) technology was proposed for the FDD system obtains channel state information difficulty, which can reduces the downlink pilot training and CSIT uplink feedback cost effectively. As the primary part of JSDM, User grouping algorithms has a huge impact on system sum rate and system implementation complexity. Traditionally, such as K-means, K-medoids user grouping algorithms has the defects of selecting the initial grouping center randomly and the number of groups was preset. In this paper, a user grouping algorithm based on the density idea to obtain the number of groups and the initial grouping center was proposed, which divides the users in the cell into cores border and isolated users at first, core users can be connected by density to form a number of core user groups, then take the average of the core user groups as the initial grouping center, and finally, the border and isolated user was grouped by K-means algorithm. And, we proposed an low complexity user scheduling algorithm. The simulation result shows that the proposed algorithm is faster and better than the traditional user grouping algorithm, and the improved JSDM system sum rate based on the algorithm are further improved.

I118 this paper, we propose a crying recognition model based on Deep Neural Network Gaussian Mixture Model Hidden Markov Model (DNN-GMM-HMM), using the acoustical model established by GMM-HMM to decode the reduced dimensionality features, and with the help of DSP Technology, the meaning of baby crying can be accurately analyzed. To attain the goal of intelligent monitoring of babies, this paper mainly adopts STM32, DSP technology and Bluetooth 4.0BLE technology to realize the function of baby crying identification, heart rate monitoring, and security alerts.

Massive multiple input multiple output (MIMO) cellular cells with small radius can satisfy the communication requirements where communication traffic is high. Since the radius of the cell is small, the large-scale fading interference of the second-circle cells outside the central cell still exists. However, it is not realistic that the traditional pilot power control method ignores the interference from the second-circle cells outside the central cell. In fact there is still a large pilot contamination problem. To solve the problem of pilot contamination in the traditional pilot power control method, this paper combines the uplink pilot power control and the orthogonal cell identification. All cells are classified by pilot power control, the cells of each class are grouped again by assigning orthogonal cell identification and the corresponding identifications are added to the head of pilot sequence which is transmitted by the mobile terminals of the local cell. Making sure that the BS identifies the pilots from different groups of cells. The experimental results show that the method of this paper is not only simple, but also can reduce the pilot contamination in multi-cell system effectively.

The pilot contamination is caused by non-orthogonal pilot sequences reuse in uplink, which affects the performance of Massive MIMO systems seriously, so it is necessary to mitigate pilot contamination. In this paper, we propose a pilot contamination precoding scheme to mitigate multi-cell pilot contamination. In the uplink, we propose a cell-defined training scheme, where the same pilot sequence is used in the same cell, and different cells use orthogonal pilot sequences, which eliminates inter-cell interference and introduces intra-cell interference artificially. In the downlink, we adopt Truncated Polynomial Expansion (TPE) precoding to reduce intra-cell interference, since the truncated polynomial of TPE precoding can replace the matrix inversion of Regularized Zero-Forcing (RZF) precoding, which reduces the complexity of RZF precoding and approximates the performance of RZF precoding by suitable truncation orders. Simulation results show the effectiveness of the proposed scheme.

Owing to OFDM Air Interface in the whole system bandwidth only support one fixed waveform parameter, the sidelobe of rectangular pulse frequency response is bigger and the attenuation is slower, which cause that the OFDM system has a sensitive to partial differential frequency rate, high frequency spectrum leakage and large out-of-band (OOB) interference. Therefore, the conventional OFDM can not adapt to 5G in rich business scenarios. Thus, the Filtered-OFDM (F-OFDM) is proposed as a candidate technique for 5G high-data rate wireless communication system. In order to achieve better spectrum leakage performance compared with Hamming window in F-OFDM system, this paper proposes a kind filter that uses Nuttall's Blackman-Harris window in the F-OFDM system, and analyzes the performance of different window functions in the F-OFDM. Simulation results show that the design of F-OFDM is easy to be implemented by Nuttall's Blackman-Harris window, which has a very low OOB emission with the same bit error rate (BER) performance compared with the F-OFDM that is based on Hamming window and the conventional OFDM.

Fuzzy control rules and parameters of traditional fuzzy controller are fixed, and fuzzy control system contains the steady state error. In view of these shortcomings, a new kind of fuzzy controller which has integral link is designed. The values of fuzzy control rule factor, quantifying factors, scaling factor and integrating factor based on ITAE optimal index are optimized by simplex method. In MATLAB Simulink environment, simulation model of fuzzy control system is established and optimization algorithm is programmed. The simulation results indicate that simplex method can still find rational values and fuzzy control system has good dynamic characteristic and static accuracy, even if there are great differences in the characteristics of the time-varying controlled object.

In massive multiple-input multiple-output (MIMO) system, scaling up the antennas of base station (BS) has a clear benefit on sum rate and energy efficiency, but the signal processing complexity can be very high and many algorithms cannot be implemented in practice for high hardware cost. Approximative Matrix Inverse Computations (AMIC) algorithm is a kind of low-complexity precoding for large multiuser MIMO systems, but the Bite Error Rate (BER) performance is shown to be not better than the classical MMSE precoding. To improve the BER performance of AMIC algorithm, in this paper, we use norm minimization algorithm to change the coefficient of the precoding matrix to improve the BER performance of AMIC algorithm. It can verify that the proposed algorithm can achieve better BER performance than the AMIC algorithm by using only a limited number of Neumann series iterations, and keep lower complexity. The proposed scheme is a compromise solution between complexity and BER performance.

In order to resolves the problem of occlusion in process of target tracking, an improved anti-occlusion tracking algorithm was proposed in this paper based on compressive particle filtering (CPF). Compressive sensing theory was introduced into particle filter (PF) framework to ensure the instantaneity of tracking. We apply the histogram with spatial information and sub-part matching ideas in the compressive particle filtering algorithm to enhance the robustness of tracking when the target was blocked by barriers. In this approach, we adopt different strategies to tracking target when the target was occluded or not. When the target was occluded, tracking it by compressive particle filtering algorithm based on sub-part matching and updating the target templates to fits the change of target appearance, otherwise, tracking it by the general compressive particle filtering algorithm. This approach bring about better robustness and tracking speed compared with the particle filtering algorithm and compressive tracking algorithm.

Orthogonal Matching Pursuit (OMP) is the reconstruction algorithm commonly used in compressed sensing theory, which has low complexity and easy implementation. Multi-Candidate Orthogonal Matching Pursuit (MOMP) improve performance and reduce the computational complexity, through selecting multi candidates adding to the optimal atom set at each iteration. Based on MOMP algorithm, this study presented the atoms matching criterion based on Dice coefficient, used the function of its important component vector quickly locate residual signal main component, alternatives to traditional rule of inner product similarity measure method, named DMOMP algorithm. The results of simulation and analysis show that the method can effectively improve the success rate of MOMP algorithm, reconstruction error, etc.

An algorithm to blindly detect frequency hopping (FH) signals corrupted by white Gaussian noise (WGN) is presented. Any parameters of FH are assumed to be unknown, and the noise power can be either known or unknown. The algorithm is based on the difference in cyclostationarity (CS) between FH signal and WGN. The module of estimation of cyclic autocorrelation is picked up with available received signal as the test statistic. A WGN of which power equals to the average power of received signal is constructed at the received end, and then the maximum module of estimation of cyclic autocorrelation is computed with this WGN as the detection threshold. Simulation results have proved that the proposed algorithm is adequate to the environments in which signal-to-noise ratio (SNR) is higher than -3dB.

In order to eliminate the musical noise remained in the results enhanced by short-time spectral attenuation techniques, this paper proposes a novel a priori SNR estimator. The proposed estimator is built on Burg-based power spectral estimation, which takes into account the properties in estimator of smoothness, accuracy and resolution, with the emphasis on their relationships and influences on de-noising. The experimental results demonstrate that: if the proposed estimator is incorporated into the noise suppressor using minimum mean-square error (MMSE) method or log-spectral minimum mean-square error (LSA-MMSE) method, better results will be achieved with the musical noise greatly eliminated and distortion reduced further.

In speech enhancement based on wavelet package, a new concept of time-frequency thresholding and the corresponding algorithm are presented. Based on the masking properties of human auditory system to wavelet denoising, the threshold is modulated to vary with the time-frequency characteristic of the speech's pre-estimate, but not merely the time-variant information as has been adopted in former researches. And via the proposed thresholding, the wavelet package coefficients can be adjusted adaptively and a better denoised speech can be achieved. Experimental results show that compared with the former algorithms the proposed method and algorithm has better performance with less distortion and residual noise.

Threshold choosing is critical in wavelet speech enhancement. In the paper, a novel wavelet packet speech enhancement algorithm is presented based on time-frequency (TF) threshold. Different from the conventional methods in threshold choosing, e.g. invariant threshold and time-variant threshold, the proposed threshold is modulated according to speech TF details other than rough envelops adopted in the recent algorithms based on eager energy operator (TEO) and adaptive noise estimation (ANE). In the new algorithm, the speech TF information is obtained from the frequency-based pre-estimate, and the threshold is modulated with TF characteristic of the pre-estimate. Then via thresholding the wavelet packet coefficients, the contaminated speech can be denoised adaptively. Compared with the former wavelet based algorithms, the proposed algorithm offers more pleasant enhanced speech with less distortion and residual noise in additive Gaussian noise case. Experimental results show its better performance in subjective test, input and output SNR test and modified bark spectral distortion measurement (MBSD).