Time filter

Source Type

Yang J.,Shanghai Second Polytechnic University | Wang Z.,Nanjing Institute of Communications Engineering
ICIME 2010 - 2010 2nd IEEE International Conference on Information Management and Engineering | Year: 2010

A blind source separation algorithm by the accurate diagonalization of second-order statistics is presented for the mixed speech-noise signals. The new algorithm uses the autocorrelation covariance matrix and its one-sample delayed counterpart of the two prewhitened noisy data and forms two 2×2 positive definite symmetry matrices pencil. A tangent algorithm for computing the generalized singular value decomposition is then exploited for simultaneously diagonalizing these two matrices accurately, and its theoretic proof is also presented. Compared with JADE algorithm and SOBI algorithm, the new algorithm is of simple computation and more computation precision. And it overcomes the limitations of their both incapable accurate diagonalization for fourth-order cumulant matrices and time-delayed covariance matrices, separately. Under the conditions of white Gaussian and colored noise, computer simulation results show that the performance of the new algorithm separating speech from noisy data is superior to those of JADE algorithm and SOBI algorithm. © 2010 IEEE.

Wu X.,Nanjing University of Posts and Telecommunications | Wu X.,Nanjing Institute of Communications Engineering
2012 International Conference on Wireless Communications and Signal Processing, WCSP 2012 | Year: 2012

It was recently shown that the multiplicative-repetition codes can achieve surprising excellent performance when concatenated with non-binary LDPC codes. However, the role of multiplicative-repetition codes is not well understood. In this paper, we provide important insights into the performance of the multiplicative-repetition codes under the maximum-likelihood decoding over binary erasure channels (BECs). It is shown that the design of good multiplicative-repetition code with repetition degree of T under ML decoding is equivalent to optimizing the distance profile of the minimal codewords for the code of the rate T-1/T Furthermore, we also provide important insights into the design of polar codes, as it can be viewed as a special class of multiplicative-repetition codes. © 2012 IEEE.

Li D.,Nanjing University of Posts and Telecommunications | Zhong W.,Nanjing Institute of Communications Engineering
IET Networks | Year: 2014

Recent results have shown that employing NC may have an adverse effect and dimmish the advantage of NC in CC because of the existence of NC noise. The authors propose space network- coded cooperation (SNCC) in which sources transmit in turn, and relays overhear and process the data they have received from different sources. Then, they broadcast the coded data to corresponding destinations in a single time instance. In this spirit, NC can be viewed as being performed in the space, and each destination can average the NC noise that comes from different relays. The authors derive a closed-form expression for the NC noise of SNCC, and show that SNCC has a potential to significantly reduce the NC noise if sessions in the network are well organised into disjoint groups/coalitions to use SNCC. Towards the end, the formation of SNCC coalitions is interpreted as a coalition formation game. Algorithms are proposed to get the solution of the game. Finally, through numerical simulations, the authors demonstrate the effectiveness of the proposed SNCC scheme and algorithms in achieving a high spectrum efficiency for CC. © 2014 The Institution of Engineering and Technology.

Xu B.,Nanjing Institute of Communications Engineering | Shen L.,Nanjing Southeast University | Yan F.,Nanjing Southeast University | Zheng J.,Nanjing Southeast University
Wireless Communications and Mobile Computing | Year: 2011

Node positioning is a useful service for a vehicular communication system. The global positioning system (GPS) is the most popular positioning technique widely used in vehicular communication systems. However, GPS cannot work effectively in many situations, for example, in a tunnel or under a bridge. For accurate positioning in such situations, it is desirable to have more reliable positioning techniques. In this paper, we propose a novel method for node positioning in vehicular communication systems. Unlike existing positioning methods that employ a roadside node as an observation station, the proposed positioning method allows a vehicular node to play the role of an observation station, and estimates the vehicular node's absolute coordinates based on the measurement of the Doppler frequency shift, the velocity of the vehicular node, and the known absolute coordinates of roadside nodes. To improve positioning accuracy, it introduces extended Kalman filtering (EKF) to mitigate the effects of channel noise and multipath interference. Analytical and simulation results show that the proposed positioning method can significantly improve the positioning accuracy compared with differential GPS and reduce the positioning delay. Copyright © 2009 John Wiley & Sons, Ltd.

Sun X.,Nanjing Southeast University | Wu X.,Nanjing Southeast University | Wu X.,Nanjing Institute of Communications Engineering | Jiang M.,Nanjing Southeast University | And 2 more authors.
IET Communications | Year: 2012

The authors study the optimisation of the rate-compatible (RC)-irregular-repeat-accumulate (IRA) codes for asymmetric Slepian-Wolf coding to deal with flexible correlated sources. By assuming two different additive white Gaussian noise channels, the authors formulate a non-uniform extrinsic information transfer method to optimise the degree distribution of RC-IRA codes. The parity check matrices of RC-IRA codes are constructed from a mother matrix by combining puncturing and check-splitting, which achieves a good trade-off between the lowest-rate codes and the highest-rate codes. Simulation results show that the optimised non-uniform error-correcting code performs very close to the Slepian-Wolf limit over the target compression rate range and outperforms the prior results. © 2012 The Institution of Engineering and Technology.

Discover hidden collaborations