- EI核心期刊
A secure and energy-efficient range query algorithm in two-tiered wireless sensor networks was proposed. In data storing phase, each sensor node encrypts its sensory data, generates order-preserving encoding by employing order-preserving function and then transmits the encoded and encrypted data to its corresponding storage node. In query processing phase, the Sink converts query range into the lower bound and upper bound of polynomials through Cramer's rule and then sends the query request to the storage node. According to the order-preserving encoding and polynomial information, the storage node implements comparison without knowing the actual values of the collected data and queried range and thus obtains the final query result which is then returned back to Sink. Next, The Sink acquires the final query result and verifies the validity and integrity of sensory data. Analysis and experiment show that the algorithm can ensure the privacy of the sensory data, the query result and the query range, and has an advantage over the existing methods in the energy consumption.
An algorithm of multiple frequency offsets estimation for distributed multiple input multiple output orthogonal frequency division multiplexing (MIMO-OFDM) systems was discussed. Accounting for the difficulty and lack of blind estimation methods, the scheme uses the deconvolution blind source separation (BSS) to distract the effect of multipath and ambiguity and obtain the frequency offset signals by considering that the characteristics of the multipath signal model is similar to the signal in BSS. The multidimensional frequency estimation problem is thus turned into multiple single dimension problems, and then the frequency offset value is obtained by a single frequency offset estimation algorithm. Simulations indicate that the proposed algorithm has better performance in estimating the frequency offsets than pilots-aided ones. Under signal noise ratio (SNR) of 5 dB, the mean square errors of the carrier frequency offset (CFO) reaches to 10-6.
Aiming at the underdetermined blind source separation (UBSS) of frequency hopping, a UBSS algorithm based on filtering was proposed to solve the problems of high computation and low estimated accuracy of signal caused by distortion method in existing time-frequency domain. Firstly, according to the estimated carrier frequencies, some kinds of band pass filters were designed. And the observed signals were filtered by using these band pass filters. After that, we can get the filter signals just composed of one source signal. Thus, the original UBSS problem was transferred into several UBSS problems in which the sparsity in time domain is satisfied. Finally, according to the filtered signals, the source signal was estimated by using the time domain UBSS algorithm. Simulation show that proposed algorithm can get more accurate estimated signals and have fewer processing data and lower computation complexity. The signal to interference ratio o of proposed algorithm is 4dB bigger than that of time-frequency domain method.
The sensitivity equation of the bimorph type piezoelectric acceleration sensor was derived from the piezoelectric constitutive equation and the bending formula of cantilever, the finite element method was also used to build the model of the sensor. A method of improving the performance of the acceleration sensor by fabricating hollow structure on surfaces of the metal beam part was analyzed by using finite element method. A vector hydrophone by using bimorph type acceleration sensors was designed, modeled and tested.
Aiming at analysis on correlations between failures and various indices characterizing network states in failure prediction, an approach for rule discovery based on fuzzy neural network (ARD_FNN) was proposed. It discoveries the correlations in network management system using the abilities to learn and infer by fuzzy logic in FNN. Simulations show that the effective initialization of parameters indicates the convergence directions that accelerate the convergence rates; thus the approach can predict the occurrence of failures accurately and such performance indicators as precision, true positive rate and false positive rate are optimized.
The map reduce program operated on Hadoop distributed file system (HDFS) has a high-energy-cost problem caused by existence of small files. In order to solve this problem, the article established a new energy model of Hadoop node cluster to analyze data then proved that there exists the optimal file size on Hadoop which can reduce the energy cost of program operation to the lowest level, and based on the above data and the margin analysis theory, a judging strategy was put forward, which can find the optimal file size from the angle of energy cost and visit cost. This strategy can merge the small files on HDFS to the optimal file size according to the cost efficiency, so to get the best benefit. The existence of optimal sized data block was proved by examination, and the reasonability and validity of identifying the data block size by the combination of cost and efficiency under the margin analysis theory are proved as well by examination.
Development of networks is restricted by energy consumption. Shutting down part links is an effective energy-saving method for Internet protocol over wavelength division multiplexing (IP over WDM) networks during the period of low traffic requests. Optimization of link-off period is of great influence on the blocking probability and energy-saving effection for IP over WDM networks. According to one-day change of IP traffic requests, an energy efficiency strategy of link time-interval control(SLTIC)was proposed to change the connection state of network topology during off-peak hours. An integer linear programming (ILP) optimization model was derived to optimize the network energy consumption. And a heuristic algorithm was designed to solve the model. Simulation shows that, compared with the original algorithm without link shut, the proposed strategy can decrease the energy consumption effectively. Meanwhile, compared with switch on/off multiple links (SOFM), the proposed SLTIC can achieve a lower request blocking probability.
Cache pollution attack is one of main security threats in named data networking. Focusing on low-rate cache pollution attack, the article proposed a collaborative detection mechanism. This mechanism calculates the request variability of every node in transmission path, adopts the request variability correlation between arbitrary middle node and network edge node as detection indicator. Simulations show that the mechanism can improve the detection rate, ensure the detection reliability. With appropriate collaborative nodes, it can detect locality-disruption more quickly. Compared with other typical detection methods, it is an effectively detection way for low-rate cache pollution attack.
A web service retrieval method based on semantic and syntactic was provided. Firstly, semantic similarity was employed to establish relationship among web services to raise retrieval efficiency; then, the retrieval precision was developed, this article applied propbank to model web service description and presented a clever retrieval algorithm. Experiments show that the new method offers higher efficiency and precision.
The passive location algorithm based on regularized constrained total least squares algorithm (RCTLS) using bearing-only measurements was presented. By this algorithm the non-linear measurement equations are firstly transformed into linear equations, and linear mapping from the measurement noise to the coefficients error is given in accordance with the first order term of Taylor expansion for the coefficients of linear equations about bearing measurements. A quasi-closed-form solution to location was obtained by taking partial derivative of the objective function formed on the basis of RCTLS and ignoring high order terms of the bearing measures. The regularization parameter is determined by analyzing the bias and MSE of RCTLS. Simulations prove that the proposed algorithm achieves better location accuracy than the previous algorithms in the case of fewer observations and higher measurement noise.
A field programmable gate array(FPGA) engine for regular expressions matching based on postfix automata(PFA) was designed. PFA can obtain smaller size and then reduce the space effectively. The limited FPGA resources can be used to implement more regular expressions simultaneously. Simulation by Testbench implies that the number and speed of the supported regular expressions is deeply improved. Compared with Thompson non-deterministic finite automata (NFA) method, it needs less hardware resources and the number of regular expressions that can be implemented concurrently becomes more, so the efficiency of implement has been increased.
For the characteristics such as data classes' variety, transmitting long distance, strong disturbed and strict real-time of the ground-based high resolution photoelectric imaging telescope communication, a communication software system based on field programmable gate arraywas built on the existed hardware platform. According to the telescope, the software system must have functionsas real-time signal processing based on fiber transmission, code and decode of data, synchronous memory, data fusion and transformation. Experiment indicates that the system can transfer 256 channels of RS422 and 256 transistor transistor logicsignals through fiber in one time, longest delay is 35μs; the system also can generate 7 synchronous signals, longest delay is less than 10μs; during the spatial and time information align, time delay is 10μs, spatial delay is 114μs. The whole communication software system has right logic, safe transmission, high real-time characteristic, fits the large ground-based high resolution photoelectric-imaging telescope well.
An algorithm for recommender system based on spectral co-clustering was proposed to defend shilling attacks. The proposed algorithm maintains spectral clustering and co-clustering priors and allows a mixed membership in user and item clusters. The rating deviations were used for mean agreement based on the co-clustering results to recommend for users. Experimental results demonstrated that under the same shilling attack dimensions, our algorithm could decrease the shilling attack affects to recommender systems apparently.
There exist self-interference and mutual interference in multiuser full-duplex system. The interference alignment technique (IA) is applied to eliminate both self-interference and mutual interference in multiuser full-duplex communication. According to the characteristics of the multiuser full-duplex system, it is treated as a one-way channel model and the channel matrix of the corresponding relationship between the equivalent model and the original model is given. Based on this thoughtfulness, the minimum interference leakage alignment (Min-IL) and the maximum signal to interference plus noise ratio algorithm (Max-SINR) were presented to eliminate both self-interference and mutual interference in multiuser full-duplex communications. Simulation proves that the proposed algorithms are effective. Comparing with other schemes, both the system capacity and the average energy efficiency are improved obviously by the proposed schemes.
A dynamic vitual honeynet system using Openflow is proposed to solve the pain point of traffic control in the current honeynet system. This system is established on software defined network architecture with Openflow switch and controller and replaces the existing semi-software forwarding solution which is based on redirection gateway. A virtual honeypot system is designed to virtualize any host with any services and to adjust the honeypot structure dynamically. A new concept of overlay virtual honeynet which defines running multiple virtual honeynets in one honeynet physical entity is advocated. The experimental results in deployed designed honeynet system verify the low forwarding delay of forwarding scheme based on Openflow, dynamic property of honeynet system, and effectiveness of overlay virtual honeynet system.
The classic bit error rate (BER) performance analysis comes from the time domain and the existing research aimed at the condition of rectangular pseudo-random code waveform (PCW). A method was proposed to analyze the BER performance of the spread-spectrum system by the overlap between the spectrum of the spread-spectrum signals and the interference. The concept of spectrum overlap degree (SOD) was given. It can make theory analysis and intuitionistic physical conception consistent so that the BER in the frequency domain can be deduced. Besides, the mathematical expression of BER and SOD was given and BER was proved to be proportional to SOD. Analysis and derivation about SOD gives a criterion to design spread spectrum signal to avoid interference. Furthermore, The given BER is suitable for arbitrary PCW. The validity about SOD was verified through simulation.