- EI核心期刊
Molecular communication can be realized on the nano-scale and can be applied to body area nano-networks because of its advantages of biocompatible. The status of current research and development of diffusion-based molecular communication was summarized and analyzed. Theory and method of body area nano-network, which is based on diffusion-based molecular communication, was introduced. This article focuses on the recent researching and development advents of diffusion-based molecular communication in body area nano-networks. The particulate drug delivery technology was introduced as an exampling application of diffusion-based molecular communication in body area nano-networks.
Due to disadvantages of Goldstein's branch cut algorithm, a new phase unwrapping method for interferometric synthetic aperture radar (INSAR) was proposed according to distribution characteristics of residue. Firstly, the neighbor dipoles preprocessing algorithm was performed to establish independent and short branch cuts. Secondly, the adaptive genetic simulated annealing algorithm was performed to calculate optimized combination of remaining residues. Ultimately, the branch cuts with short length was generated in a short time, and large "isolated island" phenomenon was avoided. Experiment shows that the proposed method has advantages in precision and operation time compared with some typical phase unwrapping algorithm.
Multimedia sensor network is an advanced form of sensor networks. It has multimedia information perception, processing and transmission capacity. For the problem of great redundancy and correlation in the collection data, a node selection approach for data compression was designed. By using the correlation model based on three-dimensional perception, it describes the relevant characteristics between the observed images data. Then the camera node selection algorithms was proposed based on correlation coefficients to reduce the spatial redundancy of sensing data. A set of comparisons are performed to evaluate the effective of our algorithms.
A 2-D direction of arrival estimation method for L-shaped array with low computation complexity and high accuracy was presented. Firstly, the subarray cross-correlation matrix of L-shaped array and the conjugate exchange properties of the uniform linear array steering vectors is given, the proposed method extends the array aperture and constructs a new array received data matrix as well. Then the estimations of 2-D angles is obtained with the estimating of signal parameters via rotation invariance techniques algorithm. Finally, the subspace is estimated via applying the Nystöm method to the autocorrelation-like matrix, which is reconstructed by Toeplitzation-the diagonal elements of the subarray cross-correlation matrix without loss of the array aperture. Therefore, the correct angle pairing is achieved with a few angle search. The proposed method does not require massive angle search, and is high in the angle estimation accuracy, low in computation complexity and few in the necessary number of snapshots. Simulation is presented to validate the correction and effectiveness of the proposed method.
Addressing the information overloading problem, the collaborative filtering is an effective technique, and extensively applied in recommender systems. It make predictions by finding users with similar taste or items that have been similarly chosen. However, as the number of users or items grows rapidly, the traditional collaborative filtering approach is suffering from the data sparsity problem. The sparse user-item associations can generate inaccurate neighborhood for each user or item. A distributed hybrid collaborative filtering method was proposed based on Map Reduce, aiming at improving the recommendation quality. This method utilizes user features and ratings to construct item preference vectors. Then, it clusters items using fuzzy K-Means algorithm, and respectively chooses similar items from each clustering, finally it combines all predictions from each clustering and makes recommendation. Experiments show that the distributed hybrid collaborative filtering method can help reduce the sparsity problem, and improve the recommendation accuracy.
Dissipative soliton is one of the research hotspot in the field of nonlinear optical fiber optics because of its high single pulse energy. The characteristics of dissipative soliton in passively mode-locked Yb-doped fiber laser was studied. First, each device in the cavity is modeled. Then, the evolution of parabolic shape dissipative soliton with strictly linear chirp in the cavity is studied by numerical simulations. Finally, the influences of dispersion and gain saturation energy of Yb-doped fiber on pulse duration and peak power of dissipative soliton is analyzed. Simulations indicate that the pulse duration of the dissipative soliton becomes wider with increase of the dispersion, and the peak power decreases with increase of the dispersion. The effect of gain saturation energy on the pulse duration is non monotonic. When gain saturation energy is 70 pJ, the pulse duration reaches the minimum value. And the peak power of dissipative soliton increases linearly with the increase of gain saturation energy.
A mobile relay selection scheme used in the cellular network was designed, it divides the users into one-hop users and two-hop ones according to the path loss. The one-hop users receive signal from the Base Station (BS) directly. However, the two-hop users could choose the idle one-hop users as Relay Nodes (RN) with the global optimal throughput prediction relaying (GOTPR). The proposed algorithm puts the impact of bandwidth and channel quality onto the rate of communication into account adequately, adjusts the bandwidth of two-hop users based on the equality of throughput of access link and backhaul link, and acquires the optimal matching matrix with the maximum throughput by the Hungarian algorithm. Simulations presents that the proposed method could improve the cell edge users throughput and spectral efficiency significantly.
With its security based on quantum no-cloning and Heisenberg uncertainty relation, the quantum cryptography is a newly developed key distribution protocol. The recent research on continuous variable quantum cryptography assumes that the choice of basis is balanced. However, the performance of continuous variable quantum cryptography is an interesting but open question. By introducing the asymmetry in the shared bipartite two-mode entanglement state, the secret key rate was improved with unbalanced basis. Furthermore, numerical simulations with typical continuous variable quantum key distribution experiment show that the secure distance can be prolonged by 20km for a given secure key rate.
Permissions are extracted as features via static analysis. The information gain (IG) algorithm is applied to select significant features. The Naïve Bayesian (NB) classifier is created which is improved through Laplace calibration and natural logarithm of multiplier. The results with 10-fold cross validation indicate that the improved NB classifier achieves higher accuracy and precision, and the selected features by IG algorithm improve the detection efficiency in ensuring the accuracy of the case. Comparing k-nearest neighbor (KNN) and k-Means classifier, NB classifier has good performance on validity, accuracy and efficiency.
To achieve accurate and timely pre-authorization of access from a subject to an object in access control, a method on knowledge discovery of authorization rules was proposed based on rough set theory to reduce knowledge attributes and to extract decision rules. According to the main characteristics of mappings between trust levels and access rights in trust based access control and combined with access authorization rules, this method makes successful use of the interactive entity authorization information as the data decision information table for the discovery of knowledge. Analysis shows that both the knowledge discovery method and the rough set based authorization rules are effective.
According to the nonlinear and multi-dimensional dynamic characteristics of network traffic, combined with the ability of multi-scale wavelet analysis, a comprehensive forecasting model based on Morlet-support vector regression (Morlet-SVR) and auto regressive integrated moving average (ARIMA) was proposed, in which Morlet-SVR and ARIMA are employed to forecast the approximate signal and the multi-scale detail signals respectively by use of Mallet wavelet decomposition and single reconstruction. The final prediction result is obtained by linear superposition of the layers. Simulations give out comparisons with radial basis function-support vector regression and ARIMA model respectively, the proposed model shows higher prediction accuracy by comparison with three error evaluation measurements.
Working in the 2.4 GHz wireless industrial scientific medical (ISM) band, a buried dielectric patch Yagi antenna was designed. Antenna metal sheet and medium plate is placed alternately and radiate perpendicular to the antenna surface, belonged to the end-fire antenna. The antenna is improved by embedded metamaterial structure. By using the phase compensation characteristics of metamaterial,the section size shrinks to 69% of the original antenna structure, at the same time, the matching and radiation performance is better than that of original antenna. A full-wave finite element simulation method was adopted to design and analyze the antenna, and the improved antenna return loss is decreased by 3 dB in 2.4 GHz frequency, and then E and H-plane radiation field intensity is got 1 dBi, which achieves the small size and flexibility demand of the antenna. Finally, the time domain finite integration method verifies radiation performance.
This article combines probability-constrained optimization technique with constant modulus algorithm to study a new robust adaptive beamforming algorithm for constant modulus signals. Aiming at the two kinds of cases that steering vector mismatch follows the Gauss distribution and unknown distribution, the optimization cost functions are respectively built, and these optimization problems are changed to the second order cone (SOC) programming ones solved with interior point methods. Then, the convex analyses of the optimization problems are carried out as well. In the simulation, the proposed algorithm compares with related robust algorithms to demonstrate its superiority and effectiveness.
In order to relax the embattle conditions and solve the phase ambiguity of existing algorithm, a signal parameter estimation algorithm based on the polarized uniform circular polarized array was studied using the estimating signal parameters via rotational invariance techniques (ESPRIT) algorithm. The phase differences between two adjacent array elements are obtained based on the performance of steering vector using the subspace theory, the phase ambiguity is resolved. Using the relationship of array manifold vector between z-axis electric dipoles and magnetic dipoles, the least squares solution to the signal polarization and direction of arrival estimations were given with matrix operations without spectral peak searching and parameter matching. Simulations verify the effectiveness of the algorithm.
Due to sparse nature of the nature of image, the sparse signal representation theory can be well used in image processing, and with sparse representation theory of continuous improvement, it is also widely used in image de-noising rehabilitation and integration process. The sparse representation of image fusion theory was used to determine the weighting factor fusion rules sparse coefficients, and to solve the optimal weighting coefficients of genetic algorithm to achieve image fusion panchromatic, multispectral images, contourlet transform, principal component analysis (PCA) algorithm and the high-pass filter image fusion algorithm. Also it improves the image clarity spectral fidelity compared to other algorithms.
In order to improve the average user utility of energy harvesting in energy broadcasting system with massive antennas, the maximum ratio transmission precoder was employed. Considering the single energy source scenario, the optimal power allocation scheme is first derived with aid of advantage of the characteristics of massive antennas. To exploit the inter-cell interference, a cooperative energy transfer scheme was proposed in multiple energy sources scenario, and the optimal power allocation for cooperative energy transfer was deduced correspondingly. Simulations indicate that the performance with the optimal power allocation is better than that with equal power allocation in single energy source scenario, and the cooperative energy transfer outperforms the non-cooperative energy transfer in multiple energy sources scenario.
The traditional radar signal sorting methods are concentrated in feature difference in time, space and frequency domains for signal separation and detection. The random overlapping signals of time,space and spectrum in the complicated electromagnetic environment make the existing methods difficult to distinguish. For these problems, a new radar signal sorting algorithm based on the compressed sensing was proposed. The radar signals for sorting can be represented sparsely in dictionary constituted by signal samples. The algorithm can obtain all the information with a small amount of observational data, the radar signals are sorted effectively. Simulation indicates that the signals can be rapidly sorted using this algorithm and the desired results are obtained in low signal noise ratio.
An efficient resource allocation mechanism was proposed for information diffusion in wireless autonomic networks. First, the dynamic arrival and leave process of the information data is characterized by a transmission queue. To cope with the dynamic transmission queues and wireless channels, the information diffusion problem is formulated as a multi-user Markov decision process (MMDP) which can be decomposed by using the principle of dual method. To reduce the computation complexity, a model-based online learning method was proposed based on which the user can make a foresighted decision by iterating only once when the network state changes.
The interference alignment is a transmission technology which aligns interference signal from other transmitters to the same signal subspace for reducing the interference to the desired signal. In recent years this technology has been widespread concerned. An interference alignment precoding optimization algorithm based on power allocation was proposed for the K user multiple input multiple output (MIMO) interference channel system. First, the signal is transmitted according to the best matching of the precoding matrix with the best feature sub channel based on the chord distance matrix by the singular value decomposition of the channel. Then the power allocation is assigned according to the channel matrix strength and also the precoding matrix is optimized. This method can ensure the signal strength maximally. Furthermore, the proposed algorithms calculate the precoding matrix and the receiver inhibition matrix without iteration. Simulation shows that this method not only reduces the complexity of the system, but also improves the bit error rate (BER) performance.
By using the change of variables, the semidefinite programming (SDP) problem for solving the senor network localization (SNL) in 3D was reformulated to be a nonlinear programming (NLP) problems. Feasible direction algorithm was proposed to solve the problem. The number of columns of the variables in the NLP is chosen to be equal 3, so as to avoid the higher dimensional solutions. Computational efficiency is improved by exploiting the sparsity of graph, in which the degree of each sensor node are restricted to a small positive integer. Experiments show that the proposed is efficient and robust, and the speed is faster than the existing interior-point algorithms for the SDP.
Aiming to the difficulty in finding the resonant frequency band in condition monitoring of rolling element bearings, a new optimization method and objective function was proposed. The resonant frequency band can be located through this proposed method. Firstly, the parameters of Gabor filter are optimized through the two-step grid search method, in which the envelope sparseness is as objective function. The vibration signal was filtered through the optimal filter and the envelop signal was calculated. The envelop autocorrelation spectrum was adopted to restrain noise and highlight operation condition information. The effectiveness and advantages of the proposed method were proved through the simulation signal and experimental signals. It is shown that the bearing operation condition can be recognized accurately by the proposed method.
In the existing literature of performance modeling and analysis of centralized multi-hop networks, the models almost did not deal with the impacts of protocol parameters, resulting in lack of theoretical support for the protocol design in practical scenarios. Regarding this problem, the schedule delay and effective throughput models which fully consider the network parameters, service parameters as well as protocol parameters were proposed. The models and numerical simulation show the impacts of parameters such as the number of nodes, network bandwidth, the number and average hop-count of service flows, proportion of control slots and bit number within one slot. The optimization strategy of proportion of control slots and bit number within one slot under various network scenarios and service requirements were further obtained.
The theoretical limitation of hacker attack ability and honker defense ability was given. If a hacker wants to successfully beat honker k times, he should have so good skills that he can achieve the purpose with probability arbitrarily close to 1 in the k/C times' offensive; If a hacker achieve S times' real success after n times' attack, there must be S≤nC. For the honker, if he wants to really successfully defense the hacker for R times, he must have so good skills that he can achieve the purpose with probability arbitrarily close to 1 in the k/C times' defensive; If a honker achieve N times' real success after N times' defensive, there must be R≤ND. C and D respectively represents the channel capacity of offensive channel and defensive channel. If C<D, the hacker Honker will be losed; If C>D, the honker will be losed; if C=D, it means that they will get considerable strength.