- EI核心期刊
An integrated routing algorithm with load balancing consideration——loadbalancing algorithm is proposed to ease the existing problems of the routing scheme used in Internet protocol over wavelength division multiplex network. It sets weight of links (both for the virtual link and physical link) with consideration of the influence of the node load, and designs objective function taking into account of the resource utilization. Simulations show that the variance of load of node is small and the connection blocking probability is reduced too.
The problem of intra-channel diversity in multi-channel routing can be relived efficiently using new proposed technology called “Network coding”, and hence spectral efficiency can be increased. In this framework, a novel routing metrics WCECTT (Weighted Cumulative Coding-aware ETT) which fairly quantify the network coding gain while considering the wireless transmission conditions and intra-frequency interference is proposed. This metric can be used to determine the optimum path from the coding-available paths and the coding- unavailable ones. We implement the WCECTT routing in wireless mesh system in C++ and conduct extensive evaluation, and the simulation results illustrate that compared with the ones which oblivious of coding gains using ,transmission efficiency can be sharply increased in the whole network and achieves 10%-20% throughput gains.
A reconfigurable processor is presented to execute the loop automatically in reconfigurable cell array. Data distribution and asymmetric first in first out buffer (FIFO) can speedup the data transfer with 8 times. The hardware architecture is verified on the platform of fieldprogrammable gate array (FPGA) with some kernel algorithms of multimedia applications such as integer invert discrete cosine transform (IDCT) and motion estimation of advanced video coding of moving pictures experts group4 (H.264) and IDCT of moving pictures experts group2(MPEG2). With a same scale of reconfigurable array, the performance will be 35 times higher than the similar researches.
Wyner-Ziv video coding is a typical distributed video coding paradigm. Current solutions in the literature perform rate control at the decoder by making use of the feedback channel (FC), and this will definitely cause huge complexity for decoding and constrain its application. Based on analyzing the impact that different motion activity would have on video coding rate, an effective rate control algorithm is proposed by utilizing the correlation noise statistical model, characteristic of feedback channel and temporal correlation. First, the encoder estimates the minimal number of rate chunks, then adjust minimal rate chunks according to the average number of requests and temporal correlation among coefficient bands. Then it would be easily to obtain suitable initial rate chunks, and consequently control the requesting times and complexity of decoder. Experimental results showed that this algorithm greatly reduce decoding complexity than others which have nearly the same coding efficiency.
An accurate macroblock (MB) layer rate control scheme based on Sobel edge operator for H.264/AVC is proposed. Utilizing the Sobel edge operator, the prediction accuracy of both mean absolute difference (MAD) and header bits can be improved. Meanwhile, a more suitable Rate Quantization (RQ) model history data point selection method is also developed to enhance the accuracy of RQ model. Extensive simulations show that the scheme can control bit rates accurately with the average peak signaltonoise ratio (PSNR) up to 0.8dB when compared with the conventional H.264/AVC rate control algorithms.
Endtoend performance of multihop wireless communication systems with nonregenerative relays operating over independent not necessarily identically distributed Weibull fading channels is presented. Based on the inequality between harmonic and geometric means, the endtoend signaltonoise ratio (SNR) is bounded upper. Novel expressions are derived for the momentgenerating, probability density, and moment functions of the bound in closed form. Using these statistical results, convenient closedform bounds are obtained for the average endtoend SNR, the channel capacity, and the average biterror probability of multihop wireless communication systems. Numerical results show the tightness of the proposed bounds.
Cooperative scheduling between base stations is proposed to reduce intercell interference in multiple antenna cellular communication systems. Based on interferencelimited twocell model, cooperative opportunistic scheduler is proposed and analyzed. The interference signal power is proven to be exponentially distributed. The concise expression of the approximated average rate and the multiuser diversity gain are derived by utilizing extreme value theory. Furthermore, the impact of several factors on the cooperative opportunistic scheduling in practical cellular systems is discussed. The multiuser diversity gain of noncooperative opportunistic scheduling is shown as withusers per cell, whereas that achieved by the proposed cooperative opportunistic scheduler is. Simulation shows that the proposed cooperative opportunistic scheduler can effectively reduce intercell interference and achieve significant gain.
Iterative tree search (ITS) is an efficient soft detection scheme in multipleinput multipleoutput (MIMO) systems. However, there often exists in ITS that loglikelihood ratio (LLR) values of some detected bits cannot be evaluated. Although it can be somewhat solved by setting the LLR magnitude for these bits to a constant value, the system performance would be degraded. To overcome this drawback, an improved soft detection scheme is proposed. The scheme ensures that the LLR value of each bit can be calculated and provides highly reliable LLR by appropriately choosing some discarded paths, retaining them and extending them to fulllength paths. The rule of choosing suitable discarded paths is presented and the lowcomplexity approach of extending them to full length is given. Simulations show that the proposed algorithm can obtain better performance than ITS and can achieve good performancecomplexity tradeoff.
The working principle, performance parameters, behaviors and main characteristics of intersecting cortical model (ICM) are analyzed. The expressions of internal activity without coupling, firing time and firing periodicity are derived from the mathematical description of ICM. The ICM features of variable threshold, nonlinear pulse modulation, synchronous pulse bursts, pulse capture, dynamic pulse bursts, autowave and synthesized spacetime are summarized. Then, the applications of ICM in automatic image segmentation, automatic edge extraction, feature extraction and image smoothing are introduced respectively. Experiments show that ICM can be used to segment images, retrieve images and denoise images effectively.
An improved cooperative medium access control (MAC) protocol, called ICoopMAC, is proposed for wireless networks. The protocol adopts a new frame control format and is completely compatible with the legacy 802.11 distributed coordination function (DCF). To adapt to dynamical channel variation, joint feedback of instantaneous channel state information from both destination node and relay node is used. Moreover, a new network allocation vector mechanism is adopted to reduce access delay at the stations caused by collisions. It is shown that ICoopMAC can increase the throughput of the whole network and reduce the average packet delay, compared to CoopMAC protocol.
A new class of binary sequence pair sets with zero correlation zone (ZCZ) is constructed by interleaving characteristic sequences from quadratic difference set (4t-1,2t-1,t-1)DS and sequence pairs from difference set pair (N,p,q,e,λ)DSP. This method solves the problem that some nonbinary elements exist in the binary sequence pair sets with ZCZ constructed so far. Meanwhile, by selecting the appropriate difference set pairs, the proposed binary sequence pair sets with ZCZ can achieve satisfactory energy efficiency in mismatched filtering.
A network selection scheme based on Hilbertspace vector norms is presented. It defines the vector norms as satisfaction function to revise the weighting factors of metrics given by analytic hierarchy process (AHP). Meanwhile, the scheme exploits vector geometrical relationships in Hilbert subspace basing on revised weights to measure similarities between vectors of network ability and user target. It can be used to choose the most suitable network for users, thereafter, it constructs a flexible space based on the both thresholds of angle and distance between network ability and user target vectors to control the handover frequency. Simulations fully indicate that the proposed scheme can select the appropriate network for various services accurately and reduce the frequency of vertical handover to maintain service constancy.
In order to overcome the shortcoming in cryptographic protocols for radio frequency identification (RFID) using the backend database model, a lightweight authentication protocol for RFID was proposed without the need for a bakeend database based on the universally composable security model. The protocol provides anonymity, untraceability and mutual authentication. The implementation of the protocol requires only the use of pseudorandom function and has higher efficiency. Universally composable security guarantees that the protocol holds secure even when running in an arbitrary and unknown multiparty environment.
A global optimal algorithm based on particle swarm optimization (PSO) is presented to resolve services selection with quality of service (QoS) global optimal in web service composition. This method will transform a web service selection global optimal problem into an optimization problem based on web service QoS. This algorithm is utilized multiobjective particle swarm optimization intelligent optimization principle to produce a set of optimal Pareto with constraint principle by optimizing multi QoS parameters simultaneously. Experiments show the feasibility and efficiency of this algorithm.
To improve performance of data transmission, a new system model for peertopeer stream (NSMP2PS) is built. The architecture and design principles for NSMP2PS are described. Introducing the network coding mechanism, the storeandforward capability of data is improved. Using the management strategy for better nodes and failure nodes algorithms, the NSMP2PS can be used to avoid the problem of unbalanced load for multicast tree. On the other hand, the optimal multicast tree structure and the continuity of players can be guaranteed by its selfoptimizing process nodes. The incentive mechanism is introduced to encourage more selfish nodes transmitting and sharing resources and data, therefore, the data quality and the overall performance of live system are also improved. 68-72
The algebraic standard form of Boolean functions with optimal algebraic immunity was researched. The sufficient conditions satisfied by the Boolean functions with optimal algebraic immunity were obtained. The Boolean functions with optimal algebraic immunity were constructed.
A blind scheme to estimate channel frequency response(CFR) in oversampled singlecarrier frequencydomain equalization(SCFDE) systems based on linear prediction algorithm(LPA) is developed. Compared with conventional LPA based timedomain channel estimation approach, this method obtains the closedform solution for CFR directly from tap weights of the prediction filter, rather than crosscorrelation of innovation and measurements. The proposal is robust to channel order overestimation. Its performance is better than that of conventional LPA based timedomain channel estimation approach. Futhermore, it ensures that the channel estimation in timedomain appears finite supporting. Simulations confirm our analysis.
When introducing and analyzing routing protocols in intermittently connected wireless networks (ICWN), an enhanced probabilistic routing protocol is proposed using history of encounter and transitivity (EPROPHET) for ICWN. EPROPHET forwards messages to the nexthop according to the contact frequency and contact duration. Simulation indicates that EPROPHET saves the buffer space, improves the packet delivery ratio, and decreases the endtoend delay.
For suppressing Gaussian noise and narrowband interference, a Gaussian noise and narrowband interference suppression algorithm called cyclic side correlation replacement (CSCR) is presented. CSCR is on the basis of the further analysis of side correlation replacement. CSCR algorithm is used for reference SCR in cyclic spectrum. Using cyclic spectrum could suppress Gaussion noise effectively, CSCR algorithm reconstructs the expectation signal by replacing the value of symmetric band with the interference signals. In addition, Gaussian noise is suppressed on the axis of nonzero cyclic frequency because of the noncyclostationary character, so Gaussian noise and narrowband interference are both suppressed at the same time. The algorithm is suitable for the modulation mode that has the symmetry of spectral correlation density. At last, through the computer simulation for phase shift keying (2PSK) signal, it shows that the cyclic side correlation replacement algorithm can effectively suppress Gaussian noise and narrowband interference.
An optimal economic replacing algorithm is introduced, it can avoid disadvantages of emphasizing particularly on performance criticality but neglecting cost occurred in the replacing algorithms based on the age and the performance degradation degree. An optimal performance inventory control algorithm is put forward, it can overcome the demerits of ignoring the technical depreciation and capital overstocking processed by the inventory control algorithm based on the minimal total expense. Simulations show that they can largely reduce the total cost and the capital overstocking.
The performance of selective transmit diversity (STD) with feedback delay is investigated under timevarying Rayleigh channel. First, the closedform expressions of the receiver output signaltonoise ratio (SNR) distribution are derived and validated via simulation. Based on these results, the bit error rate (BER) of BPSK modulation is obtained. It is shown that the BER of STD system consists of two parts: the BER of an ideal STD system without feedback delay plus an error floor of diversity order one representing the contribution of delayed feedback. No matter how small the feedback is, the error floor always exists. In consequence, the theoretical diversity order is one for any STD system with none zero feedback delay.
In order to achieve a balance between the system transmit rate and fairness among relays, a loadbalance based relay selection scheme is proposed in the systems with multiple decode andforward (DF) relays. Moreover, the average transmission rate performance and the relay selection gain of the system are analyzed based on the extreme value theory. It can be seen that the requirements of both transmit rate and fairness can be satisfied with the proposed relay selection scheme. The system can achieve the relay selection gain and average transmission rate logarithmic to the relay node number.
The statistical property of timing metric is analyzed in the presence of narrowband interference. The missing and false detection probabilities of training sequence are addressed according to the probability distributions of timing metric. It is shown that the performance of timing metric estimation can not be influenced by signal to interference ratio (SIR). In normal condition, the noise is the key factor which affects estimation performance when SIR is high(higher than 10?dB). The mean and the variance of timing metric is hardly affected by jamming bandwidth in a given SIR. The missing detection probability of training sequence, affected by narrowband interference (NBI), becomes high when SIR is lower than 15?dB. The false detection probability is not related to signaltonoise ratio (SNR) and SIR.
An antipartition selfdeployment algorithm (APSDA) is proposed to solve the problems of networkpartition, coverageholes and lacking of keysensing which disturb conventional virtual force algorithm (VFA) deployment approach in sparse network environment. By constructing the virtual repulsive force between nodes, the movement equation and the virtual attractingfield in the sensing area, the capability of selfrepair, keycoverage and antipartition are achieved. Simulation shows that the performance of network is enhanced by APSDA.
Bitplane complexity segmentation (BPCS) steganography replaces bitplane pixel blocks of high complexity with secret data. The continuity of the complexity histogram of the pixel block is affected by the embedded data. Based on the analysis of the security flaw in the BPCS steganography, a detection algorithm against BPCS embedding is proposed. The algorithm can reliably detect the existence of hidden data. It is easy to implement and computationally efficient, and suitable for realtime detection.
To eliminate effects of the feedback delay and mobile velocity in broadcast channel, the channel statistics and outdated channel information is used to estimate the current channel state information at transmitter (CSIT) jointly. Based on the obtained imperfect CSIT, a robust linear zero forcing (ZF) precoding is proposed. Simulation shows the performance of robust method is preferable. Furthermore, the relationship between system performance and feedback delay or mobile speed is presented.
A simple but powerful algorithm for automatically clustering Chinese cotopic words is presented. The method first uses punctuation ‘、’ to split and extract paratactic Chinese words within sentences from a corpus and constructs a cocitation graph by treating Chinese words as nodes. Second, the method generates several locality sensitive Hashing (LSH) signature combinations for each node in the cocitation graph. Those nodes shared at least one LSH signature combination, are grouped together and most of them may belong to the same topic. The main advantages of the algorithm are the fast speed of calculation and high convenience of implementation in parallel. Experimental results indicate the high efficiency and good clustering effect.
A multiple access network selection method based on the universal model for multiobjective optimization is presented. Parameters from upper layer and lower layer related with network selection are classified systematically. A new multiattributes decision making algorithm combined sorting in genetic algorithm and HookeJeeves method is also proposed. Simulation shows that the parameters are identified effectively, and the network selection mechanism we proposed based on the multiobjective optimization modeling can select the appropriate network for users.