最新刊期

A strong fault tolerance algorithm was proposed for LDPC (Low-Density Parity-Check) code reconstruction without a candidate set in a non-cooperative context. The algorithm uses Gauss Jordan elimination to find related columns with low Hamming weight and obtains the parity-check vector of LDPC codes according to the constraint relationship contained in the related column. Then the error code corresponding to the "1" position in the related column is removed. If the iteration according to Gauss Jordan elimination and the removing of error code can not get more parity-check vectors, then it makes these resulting parity-check vectors to be sparse and uses the sparse vectors to error correction. Finally, it reconstructs LDPC parity-check matrix effectively by making comprehensive use of parity-check vector solution, error code elimination, parity-check matrix sparsity and LDPC codes decoding to iterate. Simulation results show that, for the (576, 288) LDPC codes in IEEE 802. 16e standard, the proposed algorithm can provide good performance even when the bit error rate is 0. 0022.   

Using multimodal heterogeneous sensors to form body sensing networks (BSNs) is one of the most important ways to continuously sensing users' daily activities, but high energy consumption is the main reason for restricting its development. This paper presents a multimode heterogeneous collaborative sensing method for Parkinson's disease to reduce the energy consumption in sensing the daily activities by BSNs. The proposed method divides activity recognition into two sub-tasks which contain activity detection and status monitoring. And it uses one multi-classifier to model activity detection task and several binary classifiers to model status monitoring tasks, which are based on the chosen optimal sensor sets. Experimental results on two public dataset show that comparing with the conventional method whose sensors run all the time, the energy consumption on data transportation and model computation is reduced by 40% in MHEALTH and 15% in PAMPA2 approximately without losing activity-sensing accuracy. Thus it can help extend the lifetime of BSNs to sense users' daily activities long-termly and continuously.   

Nowdays we usually predict the static value of QoS (Quality of Service) rather than the confidence interval of the QoS in researches toward the prediction of Web services QoS. With the help of non-parametric statistical Bootstrap technique, we propose an approach to estimate and predict the confidence interval of Web services QoS; and then we use the historical QoS data of Web users which are similar to current Web users to predict the confidence interval of QoS values of the current Web users. Furthermore, we estimate the QoS confidence interval of each user invokes each Web service in WSDream dataset1. According to the experiment, we find out that the confidence interval follows a heavy tailed distribution. By randomly choosing 60% to 90% of users and services from WSDream dataset1 as our training dataset and predicting the QoS value of the other 10% to 40% users and services, we find that the average coverage rate is over 70% between the predicted QoS confidence interval and the estimated QoS confidence interval and the maximum average rate is as high as 76%. It is much better to meet personal requirement if we provide an estimated or predicted QoS confidence interval in the service selection or service recommendation.   

This paper presents a discrete bat algorithm to solve the vehicle routing problem with time window (VRPTW). The proposed algorithm defines position, velocity, updated operation of the position, updated operation of the velocity and updated operation of the frequency, and uses a method which combines the penal function with vectorial comparison to deal with constrained conditions. The proposed algorithm adopts random inserted strategy, inserted research strategy for the vehicle with minimum customers, ordinary inserted research strategy, exchanged research strategy and 2-Opt strategy with time window to expand the search space and enhance the convergent rate. Experimental results show that, the proposed algorithm has a stronger optimization capability, higher robustness and less time consumption, key parameter values and strategies used in this paper can improve performances of the proposed algorithm, and according to the hypotheses testing, there exsits a significant difference between the proposed algorithm and comparative algorithms.   

To improve the ability of the classic transform domain fusion methods filter noise, this paper proposed an infrared and visible light image fusion algorithm based on variation multi-scale decomposition method. Firstly, the original infrared image and visible light image were decomposed into structure components and texture components by variation multi-scale decomposition. The guided filtering method was used in the texture components fusion. In the structure components fusion rule, three coefficients including phase consistency, clarity and brightness information were used to measure the weight. Finally, the performance of the result image is evaluated from objective numerical and subjective observation. When compared with the fusion method based on discrete wavelet transform (DWT), non-subsampled contourlet transform(NSCT), sparse representation(SR) and shearlet transform(ST), the proposed fusion method has higher definition and detail information.   

In order to construct linear nearest neighbor(LNN) quantum circuit and reduce its total quantum cost, a matrix-based synthesis and optimization method is proposed. The linear reversible circuit is represented by matrix, and the CNOT(Controlled NOT Gate) analysis based on the matrix is put forward. The best strategy of matrix partition is given, which ensures the number of CNOT gate used in the circuit synthesis is optimal. The matrix representation of swap gate and the NN(Nearest Neighbor) rules are proposed to realize the LNN circuits. The equivalence of two insertion methods of swap gates is proven. Deletion rules of swap gates which are used to make gates adjacent to NN in different cases are proposed, and they can reduce the quantum cost. Experimental results on typical benchmark circuits and comparison against previous algorithms for LNN quantum circuit optimization, the average optimization rate in quantum cost is 34. 31%.   

A low-latency parallel WOLA (Weighted Overlap-add) DFT filter bank design method and its implementation on FPGA are presented. System objective function combined with group delay, asymmetric synthesis window design and iterative algorithm are adopted to reduce the overall system delay during the optimization of DFT filter banks. Calculation delay of FPGA implementation is controlled through multichannel parallel multiplication, multistage pipeline addition chain in key modules of DFT filter banks. The whole design is implemented on a Xilinx FPGA chip of Zynq7020. PESQ test shows that the design can achieve good speech quality. Compared with the serial WOLA structure, the delay of parallel WOLA can be reduced by 1. 192 ms at 16 kHz speech sampling rate, with the group delay reduced by 12% and the calculation delay reduced by 29. 2%.   

To meet the application requirements in high-level security scenarios (i. e., military, national security and banks), and further enhance the security for user authentication protocol in wireless sensor network (WSN), the biometric-based three-factor user authentication protocol (BTh-UAP) is proposed. For defending against the node compromise attack, the simulated attack, the man-in-the-middle attack and the privileged-insider attack in Althobaiti protocol, the smart card and password are taken as its basic secure factors, and the biometric identification that is operated by the biometric identification information generation and reply function is introduced as additional secure factor. In key management, a unique shared key for each node combined with gateway node is delivered to guarantee the independence and security in authentication phase. The shared key between user and gateway node is autonomously chosen to improve the security of the common communication channel. Furthermore, in the circumstance for non-participation of node, the updating scheme for password and biometric identification is designed to achieve the freshness. The results demonstrate that BTh-UAP not only overcomes Althobaiti's security flaws, but also its requirements for computing capability are less than the public-key encryption via using the Dolev-Yao threat model analysis and AVISPA's OFMC simulation. The tradeoff between security and computing costs indicates that BTh-UAP can be applied in high-level security scenarios for resource-constrained wireless sensor network.   

With the arrival of big data era, a large number of RDF (Resource Description Framework) data is flooding the entire Web of Data. Since the indexes of these datasets cannot be fully loaded in main memory when the RDF engines manage these huge datasets, these systems need to perform slow disk accesses to solve SPARQL queries. In this paper, a method named HDVM is proposed to reduce the number of linked data repeated times by extracting the latent triplet relation matrix from the linked dataset, and storing them in the form of subject vector, predicate vector and object matrix, which allows SPARQL queries to be full-in-memory performed without decompression. The experimental results show that the HDVM (Header Dictionary Vector Matrix) model proposed in this paper can improve the compression rate by 3%~20% compared with HDT (Header-Dictionary Triples), and the query time on billion-level-size dataset reaches average 400 milliseconds.   

The technology of accelerated degradation testing has become an efficient approach to evaluating the reliability of the degrading product. However, the method of analyzing accelerated degradation data, which excessively dependents on subjective experience, results in the inaccuracy of the reliability evaluation. In the paper, a more objective method based on acceleration factor constant principle is proposed. First, the changing rules of the parameters of degradation models are deduced according to the acceleration factor constant principle. Next, the effectiveness of the degradation data under each accelerated stress is identified through the parameter equation independent of accelerated stress. The key is that a t statistic is constructed to verify whether the parameter estimates satisfy the parameter equation. Then, the acceleration models of the parameters dependent on accelerated stress are constructed. Last, the effective accelerated degradation data is utilized to estimate parameters, so the reliability under the normal stress level can be extraplolated. The proposed method is demonstrated by taking the inverse Gaussian process as an example. Both the simulation test and case application indicate that the study of the paper provides a more objective and reasonable technical approach to reliability evaluation based on accelerated degradation data.   

With the development of electrophysiological technology, the spike signals that electrodes record contain multi-neuron overlapped spikes. This paper presents a classification method based on a compressed sensing algorithm and a maximum a posteriori (MAP) estimate to sort the overlapped spikes. The compressed sensing algorithm is used to obtain sparse signals, and the maximum a posteriori estimate is used to search an optimal value in the sparse signals. In experiments, we use one group of simulation data and two groups of measured data to verify the method. The experimental results show that when the spike waveform shapes in the data are similar, the proposed method has fewer sorting errors compared with the existing algorithms, k-means clustering and CBP (Continuous Basis Pursuit).   

In quasi-synchronous code-division multiple-access (QS-CDMA) system, Gaussian integer sequences with zero correlation zone (ZCZ) used as address sequences can not only suppress the multiple access interference (MAI) and the multipath interference (MPI), but also possess higher spectrum efficiency and transmission bit rate. However, the construction of the sequences is limited at present. In order to solve the problem, this paper presents a method of constructing Gaussian integer sequence sets with ZCZ and perfect Gaussian integer sequences by filtering operation. Based on perfect sequences and periodic sequence sets with ZCZ, the optimal or almost optimal Gaussian integer ZCZ sequence sets can be obtained. Meanwhile, based on perfect sequences, a class of perfect Gaussian integer sequences with odd or even period is constructed. The achieved results of this paper provide more address selection space for high-speed QS-CDMA system.   

Crime scene investigation (CSI) image retrieval is an important means to obtain material evidence for case solving. This paper describes the CSI image datasets, which are classified into different categories according to the content of the data, such as shoe marks, finger prints, tattoo, etc. This paper provides a survey on state-of-the-art techniques in CSI image retrieval focusing on low-level feature extraction and high-level semantic learning. Low-level CSI image features mainly include color feature, texture feature, boundary descriptor, etc. And, three categories of high-level semantic extraction techniques for CSI images are identified including using semantic template and database ontology, machine learning techniques and introducing relevance feedback. In addition, based on practical requirements from the police on using CSI images to find evidence clues, a few research directions are suggested such as introducing prior knowledge of the police to enhance retrieval efficiency.