Time filter

Source Type

Zhao X.,State Key Laboratory of Rail Traffic Control and Safety | Zhao X.,University of Jinan | Xu W.,Beijing Jiaotong University
Computational Intelligence and Neuroscience | Year: 2015

Affinity propagation (AP) algorithm, as a novel clustering method, does not require the users to specify the initial cluster centers in advance, which regards all data points as potential exemplars (cluster centers) equally and groups the clusters totally by the similar degree among the data points. But in many cases there exist some different intensive areas within the same data set, which means that the data set does not distribute homogeneously. In such situation the AP algorithm cannot group the data points into ideal clusters. In this paper, we proposed an extended AP clustering algorithm to deal with such a problem. There are two steps in our method: firstly the data set is partitioned into several data density types according to the nearest distances of each data point; and then the AP clustering method is, respectively, used to group the data points into clusters in each data density type. Two experiments are carried out to evaluate the performance of our algorithm: one utilizes an artificial data set and the other uses a real seismic data set. The experiment results show that groups are obtained more accurately by our algorithm than OPTICS and AP clustering algorithm itself. © 2015 XiuLi Zhao and WeiXiang Xu. Source

Yang J.,Beijing Jiaotong University | Zhang H.,State Key Laboratory of Rail Traffic Control and Safety | Ling Y.,Beijing Jiaotong University | Pan C.,Beijing Jiaotong University | Sun W.,Beijing Jiaotong University
IEEE Sensors Journal | Year: 2014

Many applications of wireless sensor network (WSN) require the execution of several computationally intense in-network processing tasks. Collaborative in-network processing among multiple nodes is essential when executing such a task due to the strictly constrained energy and resources in single node. Task allocation is essential to allocate the workload of each task to proper nodes in an efficient manner. In this paper, a modified version of binary particle swarm optimization (MBPSO), which adopts a different transfer function and a new position updating procedure with mutation, is proposed for the task allocation problem to obtain the best solution. Each particle in MBPSO is encoded to represent a complete potential solution for task allocation. The task workload and connectivity are ensured by taking them as constraints for the problem. Multiple metrics, including task execution time, energy consumption, and network lifetime, are considered a whole by designing a hybrid fitness function to achieve the best overall performance. Simulation results show the feasibility of the proposed MBPSO-based approach for task allocation problem in WSN. The proposed MBPSO-based approach also outperforms the approaches based on genetic algorithm and BPSO in the comparative analysis. © 2013 IEEE. Source

Ou B.,Beijing Jiaotong University | Ou B.,Beijing Key Laboratory of Advanced Information Science and Network Technology | Li X.,Beijing Institute of Technology | Zhao Y.,Beijing Jiaotong University | And 4 more authors.
IEEE Transactions on Image Processing | Year: 2013

In prediction-error expansion (PEE) based reversible data hiding, better exploiting image redundancy usually leads to a superior performance. However, the correlations among prediction-errors are not considered and utilized in current PEE based methods. Specifically, in PEE, the prediction-errors are modified individually in data embedding. In this paper, to better exploit these correlations, instead of utilizing prediction-errors individually, we propose to consider every two adjacent prediction-errors jointly to generate a sequence consisting of prediction-error pairs. Then, based on the sequence and the resulting 2D prediction-error histogram, a more efficient embedding strategy, namely, pairwise PEE, can be designed to achieve an improved performance. The superiority of our method is verified through extensive experiments. © 1992-2012 IEEE. Source

Liu Y.,China Mobile | Tan Z.,State Key Laboratory of Rail Traffic Control and Safety | Wang H.,China Mobile | Kwak K.S.,Inha University
IEEE Transactions on Vehicular Technology | Year: 2011

In this paper, an order recursive method is proposed to solve the joint estimation of channel impulse response (CIR) and carrier frequency offset (CFO) for orthogonal frequency-division multiplexing (OFDM) transmission. As long as one can obtain the solution for Qth-order Taylor expansion, the solution for (Q + 1)th order can also be obtained via a simple recursive relation. The proposed recursive algorithm actually provides a method to handle any Qth-order Taylor expansion, instead of just the second order adopted in the technical literature. Significant improvement can be observed by adopting higher order approximation. Analytical mean-square-error (MSE) performance results are given, demonstrating the efficiency of the proposed algorithm. © 2011 IEEE. Source

Zhang H.,Beijing Jiaotong University | Zhang H.,State Key Laboratory of Rail Traffic Control and Safety | Zhang M.,Beijing Jiaotong University | Sun W.,Beijing Jiaotong University
Yi Qi Yi Biao Xue Bao/Chinese Journal of Scientific Instrument | Year: 2013

To test whether the performance of security message transmission meets the requirements of vehicular to infrastructure high speed communication based on IEEE802. 11p, the communication between single roadside unit and mobile vehicle flow is simulated using Nakagami and Poisson distributions that accord with actual road characteristics to describe the wireless channel fading and moving vehicular data flow, respectively. The vehicular data sending interval, number of vehicles and vehicular speed are taken as the variables, the relationships between throughput and number of vehicles, data sending interval and packet loss rate, data sending interval and average delay, as well as vehicular speed and number of delivered data bytes for four different access categories of messages are obtained. The results show that, for different access categories of messages, there is an optimal point between the number of vehicles and throughput where both real time transmission and high network capacity are achieved, and maximal number of vehicles and throughput are realized. The packet loss rate and average latency will increase with the increasing of the vehicular data sending interval. The high priority messages can achieve low packet loss rate and low average latency under high vehicular data sending interval. When vehicular speed increases, the number of delivered bytes for security message will decline and is inversely proportional to the vehicle speed. The conclusion from simulation results is that the protocol of IEEE802. 11p is able to ensure communication service quality for high priority security messages. Source

Discover hidden collaborations