- EI核心期刊
Fault management is one of the most important parts of network management. It is a challenge problem to quickly and accurately locate the faults of network. Bayesian networks model is a prominent way to solve that problem, but it is limited when the state of the nodes changes over time. Present algorithms based on Bayesian networks model may solve the problem with higher accuracy, but the algorithms are very compex and not proper for large scale of network. A new efficient inference algorithm that can diagnosis multifault in dynamic Bayesian networks is proposed. Then, by analysis of its complexity, it is proven that the approximation algorithm has a much lower time complexity than the lower bound of exact algotithm in the dynamic Bayesian networks. Finally, the experiments show that the accuracy of the new algorithm is slightly lower than the exact inference algorithm, but its efficiency is much higher than the exact inference algorithm. This new algorithm can be applied to the large communication networks.
In order to reduce channel change time of Internet protocol television (IPTV) in Internet protocol（IP） multicast environment, a novel scheme of IPTV rapid channel change in IP multicast environment is proposed. This scheme utilizes terminal's upload bandwidth to supply rapid media stream rate to other terminals, and fulfills playout buffer in channel change termianals for a short time period, so as to reduce the channel change time. According to theoretical analysis and simulation results, under a highly channel change rate, only no more than 20% additional bandwidth is needed to reduce the channel change time to 15%-45% of original multicast channel change time.
According to the deficiency of the mean price based pricing method in combinatorial double auction for grid resource allocation, a unit price based pricing algorithm is presented. Based on combinatorial double auction and grid trust theories, a weighted average algorithm based grid resource allocation and pricing strategy is proposed. This algorithm identifies malicious nodes through the weighted average of the trust degree thresholds, and assesses the comprehensive competitiveness of the normal nodes through the weighted average of normalized bid prices and trust degrees. Simulations show the algorithm represents high trade rate and can prevent malicious nodes from entering into the market. The trade utility can give nodes incentives to improve their competitiveness.
To solve the problem of channel resource wasted by exposed terminals in mobile Ad hoc network, exposed terminals is researched. It is pointed out that the key problem is local network topology. Then a concurrent transmission algorithm based on local topology is proposed. In this algorithm, exposed terminals determines whether conflict happens in concurrent transmission through local network topology and complete concurret transmission by synchronization algorithm. Combined with IEEE 802.11 DCF protocol and cross layer mechanisms, a new protocol local topology based distributed coordination function(LTB_DCF) is developed. Simulation results show that compared with DCF, LTB_-DCF protocol improves the network throughput greatly with moderately low overhead.
For the finite buffer resources in networks-on-chip, a buffer allocation algorithm is proposed. The proposed algorithm first estimates the load on each input channel in different routers across the chip by the performance analytical model. Then, it uses genetic algorithm to allocate the resources according to the distribution of the load on all the channels. Simulation results indicate that, under uniform traffic, this algorithm can get less average packet latency than those by uniform allocation and greedy allocation strategy. At the same time, about 33.3% buffer can be saved.
Linear feature transformation was investigated to improve the classification accuracy of support vector machine (SVM) by preprocessing, and a hybrid method combining the modified particle swarm optimization (PSO) with SVM was presented. In the method, features top-ranked were preselected by linear weighted combination of t-statistic extended, Fisher's discriminant ratio and random forests feature importance scores, and a modified PSO and novel heuristic info were used to attract swarm to find optimal linear feature transformation factors. Features on dataset transformed were further refined by binary PSO, and a grid method was utilized to obtain SVM with high accuracy. Experiments on madelon of neural information processing system (NIPS) 2003 and ten data sets of university of California Irvine (UCI) verify this method has higher accuracy on 4 data sets than original C-SVM.
A new algorithm of the network anomaly traffic detection by using time windows comparing method is proposed based on the feature of enterprise information technology (IT) network traffic. Then, combining the new algorithm, the present static detection algorithm and the dynamic detection algorithm, the integrated detection model of network anomaly traffic is proposed. The model can check the network anomaly traffic by comparison methods by different ways and from different views. Finally, the availability of model is validated by implementation and testing of model in real environment.
LI Lin;YUAN Chao-wei;YANG Rui-zhe
A new analysis method which chooses the performance of the irregular modulation mapping schemes in bit interleaved coded modulation with iterative decoding（BICM-ID） system is proposed based on the extrinsic information transfer （EXIT） chart, the difference from EXIT chart analysis method is the proposed method utilizes the corresponding relationship between the extrinsic mutual information and the whole signal to noise ratio region . The simulation results show that the proposed method can effectively choose the different irregular modulation mapping schemes in the different signal to noise ratio region and can roughly evaluate the scope of system bit error rate curve in BICM-ID system.
Under fraction load （FL） and insufficient guard interval (GI), a novel adaptive resource allocation (ARA) scheme is proposed in the orthogonal frequency division multiple access (OFDMA) system. It utilizes the flexibility of radio resource management to mitigate the inter-symbol interference and inter-carrier interference due to the insufficient GI under FL instead of the complicated interference reduction algorithms at the receiver. System-level simulations show that, the proposed ARA improves system performance than the conventional resource allocation (CRA) under FL, whereas the performance of proposed ARA approaches the CRA under the full load due to the limitation of resource allocation flexibility.
A novel two-level multiple neural networks-based collaborative filtering recommendation algorithm (TMNN-CFA) for rating prediction is presented. By cooperating the multiple back propagation (BP) networks together, the higher layer neural network propagates conversely the output deviation until to the lower layer neural networks to amend the network weights and based on which, item recommendation is accomplished in the forward process of two layers networks relying on the factors such as ratings, etc. Experiment results on the standard Movielens show that TMNN-CFA method is effective and feasible for item recommendation.
In order to analyze and compute the multi-path efects on angle error of microwave landing system (MLS), a new numerical method based on the geometric optical and ray-tracing theory is proposed. In the new method, through the geometric optical and geometric diffraction theory, the signal level ratio of multi-path signal relative to direct signal can be effectively acquired. And through ray-tracing theory, the phase displacement can be computed to achieve the angle error pattern. Through the simulation, the angle error pattern is devided into in-beam region and out-of-beam regions. It is found that the in-beam error is key point of angle error. The new method is applied to simulate the multi-path scenario in an airport. And the results are compared with results of the experience methods to validate its feasibility.
Broadcasting is a common operation for establishing on-demand routing protocols and updating routing table. A straightforward broadcasting by flooding results in serious redundancy and contention. To increase broadcasting efficiency of Ad hoc network, a dynamic broadcasting algorithm for on-demand routing protocol is proposed, which computes the rebroadcast probability of message based on neighbor node density and relative distance. Consequently the rebroadcast probability of redundancy message is reduced and the contention problem is alleviated. By applying this algorithm in network simulation platform, the simulation results show that the proposed algorithm improves the protocol efficiency and reduces the redundant information.
3G oriented online charging system （OCS） concept is briefly introduced at first. Then the major diameter credit control （DCC） protocol performance benchmark data to be handled by OCS is described. The composed full-services traffic, user credit status, service duration/traffic distribution are analyzed and the calculation formula for DCC protocol load is presented. The two load control models including tested system unresponsive and responsive model are proposed for different load test cases. The integrated load test model was adapted in 3G OCS development in China Telecom, and the test results show the model can support high performance load test for DCC protocol.
To increase transmission efficiency, a soft-information-based Turbo equalization algorithm is proposed for the space-frequency block code orthogonal frequency-division multiplexing (SFBC-OFDM) system without cyclic prefix (CP). In our algorithm, processing modules as inter-symbol interference cancellation, soft cyclic reconstruction, diagonalized SFBC decoder and minimum mean square error frequency domain equalizer are proposed to work in an iterative way to suppress interferences. Simulation results show that this algorithm can improve the bit error rate performance of SFBC-OFDM system without CP.
The structure of optical code division multiple access (OCDMA) system based on Chebyshev-map chaotic codes is investigated in passive optical networks. The system can support more subscribers by providing the different codes at same wavelength channel. The scheme on chaotic spread spectrum sequences using Chebyshev-map is presented, which is taken as subscriber address codes. The performance of this system is analyzed. The simulation results show that the chaotic sequences have good correlation properties, and restrain the influence of multiple access interference. The bit error rate performance of the proposed system is improved by about 1～3 orders of magnitude better than that of single OCDMA system, and obtains about 1 order of magnitude improvement relative to traditional optical orthogonal codes.
Given the same due date and no idle time in processes, a heuristic algorithm for the earliness and tardiness penalties scheduling problem in permutation flow shop is proposed. Firstly, the algorithm gets the relatively good sequence of the jobs according to the sorted sequence regulation of 7 jobs. And then, the NEH algorithm is used to get the optimized solutions. At last, the optimal solutions are achieved by comparing the neighborhood searching results for every solutions. In the middle scale and small scale problems, the extent of similarity between the solutions got by the heuristic algorithm and the optimal solutions is 99%. Meanwhile through testing the standard deviation of the large scale samples, it is shown that the stability of the solutions is very good. In large scale problems, the heuristic algorithm appears good optimized performances and its calculated results are better than those got by ILOG-constraint programming(CP) optimized engine.
In order to enhance the anti-interception performance of modulation-hopping technology, a method to design wide-gap chaos modulation hopping pattern based on time of day (TOD) is proposed. According to the deficiency of the cycle extraction algorithm in the balance, an improved algorithm combining the cycle extraction and reshaping operation is presented to design Chaos modulation hopping pattern. The analysis and simulation results verify this method has better properties in balance. Furthermore, in order to perfect the anti-interception property, symmetry algorithm and random shift replace algorithm are introduced to widen the modulation hopping pattern gap. Simulation results prove that symmetry algorithm is applicable for generating the long periodic hopping pattern used in valid data communications, while random shift replace algorithm is proper for the short periodic pattern with the good balance in the range of maximum effective length, and it can be used in the situations of short-cycle emergency communications.
The existing information security risk assessment approaches commonly ignore the relations among the assessment factors. To overcome the ambiguity and complexity of relations among the factors, and the difficulty of acquiring the correlative fusion weights of attributes in practice, a hierarchical risk assessment model based on Shapley entropies and Choquet integrals is proposed. By introducing the Shapley value concept of nperson cooperative game theory into the assessment model, this approach solves the information security risk assessment problem under poor information conditions by using the analytic hierarchy process with Choquet integrals from bottom to top, which is based on the maximum Shapley entropy principle. The effectiveness of the proposed approach is illustrated via an actual information security risk assessment for four subnets of the intranet in a community.
The trim beam number is an important parameter in the scheduling model of the cotton spinning manufacturing process. Because of the complexity of the trim technique, the actual trim beam number is difficult to obtain before scheduling. A prediction algorithm using a modified support vector machine (SVM)-based feature selection method and feed forward neural network (FFNN) is presented for predicting the trim beam number. In the algorithm, the proposed feature selection method is adopted to pick up critical features that affect the trim beam number, and FFNN is adopted to predict the trim beam number based on the critical features. Numerical computational results show that the proposed algorithm is effective. The algorithm also successfully applies in the related problems in practical cotton textile manufacturing system.
In order to implement efficient encoding, a class of nonbinary quasi-cyclic low-density parity-check (QC-LDPC) codes is proposed, which can be encoded with Richardson-Urbanke algorithm. The right part of the parity-check matrix has weight-2 columns, which allows the construction of both regular and irregular codes. Constraints on the parity-check matrix ensure a linear encoding complexity. Simulation results show that the proposed codes, when combined with higher order modulation, perform favorably with the codes constructed by the progressive-edge-growth (PEG) constructed code and close to the Shannon limits.
A new ultra wideband (UWB) synchronization method is presented. This method modulates the synchronization signals and information signals respectively on tow carrier waves which are orthogonal mutually, and then combines them to transmit. After the hybrid signal is received, it will be decomposed according to the characteristic of orthogonal waves to get the synchronization signals and information signals. The information transmission system is designed by using this synchronization method. We also simulate the system and analyze the performance. The result of the simulation indicates that the system uses this synchronization method can transmit the synchronization signals and information signals at the same time and improve the synchronization performance obviously.
To the applied limitation of the existing threshold signature schemes based on the (t, n) threshold structure, a new threshold signature scheme based on the general access structure is proposed through combining the secret sharing scheme based on the general access structure and the threshold signature scheme based on the (t, n) threshold structure. The generation and verification processes of the partial signature and the group signature are introduced in detail. Both the validity and security of the proposed scheme are proved.
The main construction methods of the forward error correction (FEC) code type for optical transmission systems are proposed and a novel super forward error correction (Super-FEC) concatenated code of RS(255,239)+BCH(2?040,1?930) to be applied for optical transmission systems is constructed after the characteristics of the concatenated code for optical transmission systems and the two super forward error correction(Super-FEC) code types in ITU-T G.975.1 ere analyzed. The simulation results show that this novel concatenated code, comparing with the RS(255,239)+CSOC(k0/n0=6/7,J=8) code in ITU-T G.75.1, has a lower redundancy and the better error-correction performance. Furthermore, its net coding gain (NCG) is respectively 045?dB, 048?dB more than that of BCH(3?860,3?824)+BCH(2?040,1?930) code and RS(255,239)+CSOC(k0/n0=6/7, J=8) code in ITU-T G.75.1 and 281?dB more than that of RS(255,239) code at the third iteration for the bit error rate of 10-12. Therefore, the novel Super-FEC concatenated code can better be used in optical transmission systems.
A proportional fairness resource allocation algorithm, including both subcarrier assignment and power allocation, is proposed in multicast orthogonal frequency division multiplexing (OFDM) systems. The former is achieved by modifying the existing unicast subcarrier assignment algorithm, whereas the latter is implemented as allocating power via dual-bisection iteration, which is capable of obtaining the optimal allocation results when subcarrier assignment is determined. Simulation results demonstrate that the proposed resource allocation algorithm can achieve perfect fairness among users while maintaining the efficient capacity performance. It is also shown that multicast results have remarkable capacity enhancement comparing to unicast despite proportional fairness constraint. However, the capacity gain has a close relation with rate proportion and the number of multicast group.
An decoding structure of Turbo code based on parallel prediction control (PPC-Turbo) is presented. The value of № (n+1) exterior information is predicted by prediction control algorithm (PCA). Compared to the traditional decoding algorithm, the PPC-Turbo structure based on PCA could reduce the complexity and delay of the decoding algorithm. The trend of the exterior information prediction is compared by the single predi
ction 6～9thand the compound prediction 6+8th,7+9th exterior information values in sequence. It is verified the delay reduction by the compa
rison of decoding delay under different frame length. The bit error rate performance curve is also compared between PPC-Turbo and traditional Turbo structure. The simulation results show that the process of the component decoder could be used instead.
Reachability analysis, which is widely used to generate executable protocol conformance test sequences, often causes state explosion problem. To overcome the problem, a heuristic state configuration exploration method is proposed. Using a weight extend finite state machine as a protocol model, the new method employs the heuristic state configuration exploration to replace the traditional strategy of breadth-first-search. In this way, the generation of executable protocol conformance test sequences is reformulated to find the minimum weight path in the whole state configuration space of a protocol implementation under test . The experiment results show that the proposed method is superior to breadth-first-search algorithm in time and space complexity.
A game model based on Vickrey-Clarke-Groves (VCG) mechanism for dynamic spectrum allocation is presented, to solve the complexity problem of the dynamic spectrum allocation and reduce information exchange during the dynamic spectrum allocation. Further, a partially observable Markov decision processes (POMDP) reinforcement learning algorithm is presented. Through the observation and statistics of historical information, the secondary users enhance the reward value of bidding strategy by continuous learning, so as to obtain the optimal bidding strategy
. Finally, the POMDP reinforcement learning algorithm is transformed into optimal strategy learning algorithm of belief Markov decision processes(MDP), which is solved by using the value iteration algorithm. The simulation results reveales that the POMDP reinforcement learning algorithm can significantly improve the performance of dynamic spectrum allocation.
A kind of distributed test and control network is studied. By analyzing the change of the network connectivity when the nodes fail, the minimum average degree of varying network is suggested as the aim of reconfigurating network. What follows is a new network reconfiguration strategy and its application. Finally, this strategy is proved to be effective in connectivity of reconfigurating network when the former structure of the network can not be recovered. Besides, it's valid in an aviation and aerospace test and control network.