- EI核心期刊
Many recently proposed subspace clustering methods suffer from two severe problems: First, the algorithms typically scale exponentially with the data dimensionality or the subspace dimensionality of clusters. Second, the clustering results are often sensitive to input parameters. A fast algorithm of subspace clustering using attribute clustering is proposed to overcome these limitations. This algorithm first filters out redundant attributes by computing the gini coefficient. To evaluate the correlation of each two non-redundant attributes, the relation matrix of non-redundant attributes is constructed based on the relation function of two dimensional united gini coefficients. After applying overlapping clustering algorithm on relation matrix, the candidate of all interesting subspaces is achieved. Finally, all subspace clusters can be gotten by clustering on interesting subspaces. Experiments on both synthesis and real datasets show that the new algorithm not only achieves a significant gain of runtime and quality to find subspace clusters but also is insensitive to input parameters.
Three approximate methods (called as scaling, rectangle and trapezia methods respectively) for performance analysis of orthogonal space-time block code (OSTBC) with M-ary phase shift keying (MPSK) and M-ary quadrature amplitude modulation (MQAM) over flat Rayleigh fading channels are proposed, and their approximate performance expressions are given. The analysis of approximate results shows that OSTBC can obtain full diversity orders. The simulation proves that trapezai method is the best of them. When the segments for trapezia method become very small, the approximate performance curves perfectly agree with the theoretical curves. Under high signal to noise ratio , all approximate curves based on the three methods approach to the exact performance curves, and thus satisfy practical engineering requirements.
Due to the base of the robot is in the floating state in space, the conventional method to calculate the joint torque of the ground robot can’t be directly used to free-floating space robot. The tangent space and normal space of the robot motion curve is constructed by using Newton-Euler orthogonal method, which satisfies the restrictions of linear and angular momentum conservation of free-floating space robot system. An analytical expression of joint actuation torque in the floating state is given by using this method. Finally, a two-link space manipulator is used to verify the application of this method.
A 4SPS-2CCS generalized Stewart platform is put forward and its forward displacement is analyzed. Based on quaternion, the model of the forward displacement solutions is built up. In order to obtain the number of solutions of this kind mechanism, Mourrain variety is used. The results show that the number of the forward displacement solutions of the parallel mechanism has at most 160. In addition, all the 160 solutions have been obtained by using homotopy continuation method. The example shows that the upper bound of 160 solutions can be reached in the general case. In stead of using Euler angle rotation matrix, the use of quaternion can reduce the Bezout number of the equations and the calculation time to increase the efficiency, and make the computer program easy to write. The result is verified by a numerical example.
A new state-dependent blocking probabilities analysis model is proposed for multi-fiber WDM (wavelength division multiplexing) network. By processing the idle states of each wavelength of the independent links as the markov chain, this model can be used to analyze the multi-fiber routing network with First-fit wavelength assignment (FF-WA) and fix routing algorithms, and break the limitation of the previous models which can only handle the random wavelength assignment algorithm. Numerical analysis proves that the new model can estimate the blocking probabilities in wavelength-routed WDM network more accurately, and capture the utilization changes of different wavelength.
Abstract: The existing proxy ring signature schemes share the disadvantage of low efficiency in computation. An efficient proxy ring signature from bilinear pairing is proposed, which is obtained from Chow’s recently proposed identity-based ring signature. Furthermore, the proposed scheme can prevent the original signer from generating a valid proxy ring signature, thus the profits of the proxy signer are guaranteed. Strong unforgebility of the proposed scheme is proven to rely on CDH problem using forking lemma for ring signature under random oracle model. No pairing operation is required for the generation of a proxy ring signature and only two pairing operations are required for the verification. Compared with the existing schemes, the proposed scheme is more efficient.
In this paper modeling the base station of wireless communication networks with finite user population discrete time loss and delay system. And using the method of Markov chain we derive the performance index such as probability of loss in one slot, the mean waiting time of the mobile stations, the mean throughput of the system, and the utilization ratio of channels. And through the numerical simulation, we discuss the effect of the model and the applications on design and performance analysis of digital wireless communication networks.
An efficient rate-distortion (RD) optimized macroblock mode decision algorithm for H.264-Scalable video coding (SVC) extension in packet loss environment is presented. The end-to-end distortion is estimated according to the hierarchy B picture coding structure of SVC and the error-propagated distortion stored in a distortion map. A parameter on temporal level is added to the RD optimized macroblock mode decision algorithm in current H.264-SVC test model to improve the error resilience performance of the encoded scalable streams. The simulation results show that the proposed method outperforms the RD optimized macroblock mode decision algorithm in current H.264-SVC test model.
An adaptive selective repeat automatic repeat request scheme in correlated Rayleigh fading channel is proposed. For this algorithm, we adapt the transmission modes or packet size to the channel state variation. Closed form expression of throughput efficiency of ideal and non-ideal adaptive selective repeat ARQ is derived using multi state Markov system model defined according to both channel state and transmission mode. Further,Optimum packet length corresponding to transmission mode and optimum channel state transition estimation parameters are explored. Simulation results demonstrate with the optimum packet lengths and optimum channel state transition estimation parameters, the throughput efficiency is very close to ideal adaptive SR ARQ and higher than those with non optimum parameters schemes in a wide range of signal to noise ratio.
Lead time is an elementary factor in supply chain management, which has become a focus of recent research. Petri net is a quite good modeling technique for complex logistics systems, which is characterized by visual tools and strict math basis. This paper gives defination of the Generalized Stochastic Petri Net, and apply GSPN method to caculate the supply chian lead time and to analyze the perfomances of different activities derived fom the purchase process.
Orthogonal frequency-division multiplexing (OFDM) combined with multiple-input multiple-output (MIMO) is an efficient technique to improve wireless communication performance and capacity. To improve the performance of MIMO-OFDM system, a novel pilot sequence taking advantage of the orthogonality of space frequency block code (SFBC) as well as the corresponding simple channel estimation algorithm, is presented in this paper. The theoretic mean square error (MSE) of the algorithm is deduced and the complexity is also given. The performance of the proposed scheme is verified by simulations.
A norm and correlation based antenna selection algorithm (NCBA）and its modified version (NCBA-M) are proposed for the downlink transmit antenna selection in distributed MIMO systems. The performance of the proposed algorithms are compared with that of two low complexity antenna selection algorithms, which are norm based antenna selection algorithm (NBA) and large scale fading based antenna selection algorithm (LFBA) respectively. The simulation result shows that when no transmit antenna correlation in the same antenna port is considered, NBA, LFBA and NCBA have similar performance to the optimal antenna selection algorithms. However, NBA and LFBA result in a high capacity loss in scenarios with high correlation among transmit antennas, and NCBA-M can still approach the performance of optimal antenna selection algorithm.
Dynamic limited polling protocol based on length of queue (DPBQ) is proposed. This protocol adaptively controls the polling process according to the real-time queue length of each stations and serves the one with the longest queue firstly in each polling cycle. By means of imbedded Markov chain theory and vacation queue model, the theoretical model and performance analysis of that protocol are given. By comparing the polling system of DPBQ and basic polling system with limited service, it is shown that DPBQ system can decrease the mean waiting time of packet, so it can guarantee the delay performance better.
The Vertical Beil Laboratories Layered Space-Time(V-BLAST) is a multiple transmit and receive antenna system that provides very high capacity comparing to single antenna systems. Traditional detection algorithm for V-BLAST is based on Order Successive Interference Cancellation (OSIC) .the error propagation resulting from the low-diversity substreams constraints the overall system performance greatly using the OSIC algorithm. Using the algorithm that low-diversity substreams are iteratively decoded by using decisions form high-diversity substreams the performance is highly improved over the traditional one, but this algorithm needs too much iterative detection. For example, to the MIMO system with n transmit antennas the number of iterative detection are equal to n. The proposed new iterative detection algorithm for V-BLAST needs only three loops iterative detection to get the same high performance. For symmetric system where there is an equal number of transmit and recieive antennas simulation result show that there is a significant performance improvement comparing to the V-BLAST Order Successive Interference Cancellation (OSIC). On the other hand, since the number of iterative detection of the new algorithm is not very large, the complexity of the new algorithm is receivable.
A novel trust model for p2p networks based on recommendation is proposed in this paper, which employs the credibility to weigh the refers from different recommenders, as well mathematic analysis for credibility and some measures taken to defense against several malicious attacks are presented. Considering all kinds of malicious peers and peers’ dynamic behaviors in P2P environments, the computation of the trust value for a peer consists of two parts: reputation value and penalty value. Subsequent experimental results show that, compared to the current some trust models, the proposed model is more robust on trust security problems and more advanced in successful transaction rate.
In optical networks with mesh topology, two links in the network may fault at the same time. The way to protect the dual failures with p_Cycles was analyzed, and the configuration strategy of p_Cycles in mesh networks with at least 3-degrees was then discussed. Furthermore, a heuristic algorithm based on the distributing of the working resource and the effective straddling links was proposed. And we simulated our proposed algorithm by OPNET Modeler in COST239 network under static traffic model. The simulation results show that our proposed algorithm has shorter restore time and lower capacity redundancy.
Based on the decode-and-forward relaying mode, the joint resource optimization was analyzed for both power and bandwidth in two-hop OFDM-relay system. Taking the maximum end-to-end data rate as optimization criterion, the bandwidth allocation results of the two-hop is presented and the power allocation according to the water-filling algorithm in this situation is also proposed. At the same time, a matching scheme for two-hop sub-carriers was proposed based on the “Immunity Genetic Algorithms”(IGA). The simulation results show that compared with the uniform resource allocation scheme, the proposed joint optimization scheme can improve the system capacity, and the effect from power optimization is stronger than that of bandwidth. The system performance can be enhanced further by the matching scheme for sub-carriers.
A scheduling algorithm based on the two dimensional utility function is proposed to guarantee the packet drop ratio and play-out outage ratio of streaming application in the mixed service scenario. The simulation results show that guarantee capability for the drop and play-out outage ratio of the proposed scheme outperforms the modified-largest weighted delay first algorithm and exponent proportional fair algorithm, which only take the head-of-line delay into consideration.
In Motion-Compensated Temporal Filtering (MCTF) based 3D wavelet/subband scalable video coding, quality scalability or rate scalability is implemented by extracting corresponding sub-streams from a full-resolution embedded code stream. This paper first expounds the theoretic fundamentals for the commonly used bitplan truncation algorithm and points out its sub-optimality. An optimal subband rate allocation algorithm in mean-squared error (MSE) minimization sense is proposed, achieving persistent optimal rate-distortion per-formance under any given rate constrains. Comparative experiment results also show that the choice of bit alloca-tion strategies has serious influence on the performance of the 3D subband video coding.
The basic principle and receiver algorithms of spatial multiplexing-space time block code (SM-STBC) system are first introduced. The theory analysis on zero-forcing (ZF) detection is presented, based on the probability density function (PDF) of the output signal to noise rate (SNR). Therefore, the diversity gain and un-coded bit error rate (BER) expressions are derived and verified by computer simulation. Theory analysis and simulation results both demonstrate that through introducing STBC to the SM systems, SM-STBC can achieve much more diversity gain, i.e. better BER performance, at the price of loss of partial multiplexing gain.
The channel impulse response (CIR) leakage in OFDM channel estimation is analyzed. A new one- dimensional (1D) recursive least square (RLS) adaptive channel estimation algorithm is proposed in time domain. To reduce the CIR leakage, the input signals and reference signals of the RLS filter are weighted by a matrix that is independent to the channel statistics. Keeping the simplicity of the 1D-RLS algorithm, the proposed algorithm has much lower computational cost than the 2D-RLS counterpart. The simulation results show that the proposed algorithm improves the MSE performance of the 1D-RLS channel estimation significantly in the non-sample-spaced multipath fading channel. The MSE performance can even approach the result of the 2D-RLS algorithm.
Using the incomplete information static game theory, the telecom universal service fund allocation mechanism was studied. One balanced bid under the lowest bidder auction mechanism was put forward by building game theory models. The result shows that the bid in the balance decreases as the number of universal service fund participant increases. The theoretical basis was provided for the development of telecom universal service fund allocation system.
An improved adaptive power control algorithm in CDMA systems is presented in this paper. The proposed algorithm is based on SIR-measurement method. The step-size of variable power is decided according to both channel fading characteristics estimated by a Kalman filter and the previous power control instructions recorded by a memory unit. The simulation results demonstrate that the proposed algorithm will be able to overcome the fast and deep fading and improve the performance of CDMA systems.
A channel estimation method of high precision in OFDM system, based on least square (LS) algorithm, is proposed. The valid path taps are effectively chosen from the estimation value of channel impulse response (CIR) gotten by LS algorithm, according to the relationship of channel in time domain and then the time domain estimation results are transformed into frequency domain. To utilize the relationship of channel in frequency domain, a linear minimum mean square error (LMMSE) combiner is used to refine the estimation results. The simulation results in the frequency band of 2.4 GHz show that the proposed method achieves a good bit error rate (BER) performance close to the theoretical bound of ideal channel estimation.
Multi-Carrier Proportional Fair Scheduling (MC-PFS) and Max C/I (MCI) improved algorithms implemented by Genetic Algorithms (GA) are proposed for OFDMA systems. According to the criterion of proportional fairness, MC-PFS searches optimal subcarrier allocation scheme on basis of GA. Simulation results show that near upper bound performance is achieved through GA based MC-PFS. By introducing MCI initial allocation, improved algorithm increases system throughput compare to the upper bound (PF).
Based on the architecture of Hierarchical Mobile IPv6, a Hierarchical Mobile Multicast (HMM) algorithm is proposed to resolve the motion of multicast receiver. By using hierarchical structure, local tunnel and inner multicast, the results of simulations show that the rebuild ratio of multicast tree is decreased in HMM. As the same time, delay of recovering multicast service, average transmission cost to destinations and total transmission cost of entire multicast network in HMM are near-optimal, which has good application foreground.
The MAC protocol applied for OFDMA system — Centralized Mini-slot Packet Reservation Multiple Access (CMPRMA) protocol is proposed. The proposed protocol divides the radio resources into mini-time-frequency blocks, which are the basic transmission units. As the control center, Base station controls the resource allocation dynamically for users’ access and data transmission. It not only inherits the advantages of MPRMA prototype, which can support real-time traffic well, but also can supply the resource reservation scheme for data traffic，and can support efficient transmission for multi-users and multi-traffics.
A high speed software decoder based on BREW platform was proposed. The decoding algorithm for H.264 was improved. The whole architecture, memory allocation and key functions of the code transplanted to BREW platform were recomposed and optimized. Experimental results show that, on the premise of guaranteeing the video quality, this decoder implemented on mobile terminal using Arm9 chip meets the requirements of continuous playing.
DDoS attack detection based on Self-Similarity of Network Traffics is a new-fashioned method based on analysis of parameter of self-similar. With the study of the feature of Packet Capture and wavelet analysis, we propose a fast online detection method aiming at DDoS attack detection based wavelet transform, and show its performance through experiment.
ZHOU Cui-lian1,2, CHANG Jin-de3
An Euler-Bernoulli beam equation subject to a special boundary feedback problem is considered. This problem is in sharp contrast to the general principle in engineering practice that the conjugate variables cannot be assigned simultaneously at the same boundary. The Riesz basis approach is adopted in investigation. It is shown that the closed-loop system is well-posed in the usual energy state space and the trajectory approaches to the zero eigenspace of the system as time goes to infinity. Therelease of the engineering restriction gives more freedom in the design of boundary controls for thesuppression of vibration of flexible structures.