Time filter

Source Type

Liao C.-H.,Tsinghua University | Liao C.-H.,National Key Laboratory on Blind Signals Processing | Zhou S.-D.,Tsinghua University | Zhu Z.-L.,National Key Laboratory on Blind Signals Processing
Xi Tong Gong Cheng Yu Dian Zi Ji Shu/Systems Engineering and Electronics | Year: 2010

A maximum-likelihood (ML) based algorithm for joint timing estimation of both useful signal and co-channel interference is proposed. The likelihood function is first derived, and then the timings of two signals are identified in the two-dimensional timing space by using a scale searching method. Compared with the traditional algorithm for single-signal timing estimation, the proposed algorithm requires more observation data. The algorithm is little sensitive to frequency offsets and shows small performance attenuations for a frequency shift of 2% of the symbol rate. The simulation results verify the effectiveness of the proposed algorithm.


Liao C.,Tsinghua University | Liao C.,National Key Laboratory on Blind Signals Processing | Wan J.,National Key Laboratory on Blind Signals Processing | Zhou S.,Tsinghua University
Qinghua Daxue Xuebao/Journal of Tsinghua University | Year: 2010

Separabilities of mixed signals and algorithms for single-channel blind separation (SCBS) are evaluated using the blind separation performance bound. This paper focuses on the theoretical SCBS bound for two co-frequency modulated signals, the multiphase phase-shift keying (MPSK) signal and the quadrature amplitude modulated (QAM) signal. An upper bound of the symbol error rate was derived in a closed form using Forney's method and the maximum-likelihood criterion for the joint symbol sequence detection of two signals. Simulations show that the upper bound agrees well with the Viterbi result in ideal situation and that the parameter differences between the two source signals remarkably affect the SCBS performance, with the performance being better when the parameter differences become larger and with the delay difference being the main factor affecting the separation performance.


Zhang X.,University of Electronic Science and Technology of China | Hu J.-H.,University of Electronic Science and Technology of China | Zhu Z.-L.,National Key Laboratory on Blind Signals Processing | Yuan Y.,National Key Laboratory on Blind Signals Processing
Dianzi Keji Daxue Xuebao/Journal of the University of Electronic Science and Technology of China | Year: 2015

In this paper, a symbol rate estimation algorithm based on cyclic spectrum is proposed for single- channel two co-channel mixed OFDM signals, which analyzes the character of cyclic spectrum of a single OFDM signal and the mixed OFDM signal respectively. The receiver can identify the mixed OFDM signal and simultaneously estimate the symbol rates of two co-channel OFDM signals from the difference of cyclic spectrum. The main feature of the proposed approach is needless to know any prior information of the mixed signals, which can be applied to intermediate frequency signal. Simulation results demonstrate good availability of the proposed method, and verify that it is not sensitive to the noise © 2015, Editorial Board of Journal of University of Electronic Science and Technology of China. All right reserved.


Yang Y.,Zhengzhou University | Zhang D.-L.,Zhengzhou University | Peng H.,Zhengzhou University | Tu S.-L.,National Key Laboratory on Blind Signals Processing
Tien Tzu Hsueh Pao/Acta Electronica Sinica | Year: 2015

For single-channel linear mixture of high-order modulated signals which have t he fixed frame length and sync code, a frequency offset estimation(FOE) algorithm based on the maximum likelihood criterion is proposed when the two signals are c oded asynchronously, and the effective range of FOE is given.FOE is calculated by optimising the objective function, which is obtained from cross-correlation comp utation, and the optimised process is accomplished by utilising hierarchical sear ch. To evaluate the performance of the proposed algorithm, Cramer-Rao Bound (CRB) of mixed-signal FOE is derived. Simulation results show that, the estimation perf ormance of the algorithm can come close to CRB performance with low Eb/N0.The proposed algorithm is also applied to the FOE of a mixed signal whose power is asymmetrical. ©, 2015, Chinese Institute of Electronics. All right reserved.


Yang Y.,Zhengzhou University | Zhang D.-L.,Zhengzhou University | Peng H.,Zhengzhou University | Tu S.-L.,National Key Laboratory on Blind Signals Processing
Tongxin Xuebao/Journal on Communications | Year: 2015

A Gibbs-sampler-based separation algorithm was proposed for single-channel co-frequency digital-modulated signals, which were received in non-cooperative ways. The probability density samples of unknown symbol sequences can be obtained by statistic method, and the computation complexity was not growing exponentially with the increase of channel order. With special emphasis on the separation algorithms of single pair of symbols, multiple pair of symbols and the tracking of channel responses. Moreover, the separation performance between Gibbs sampler and per-survivor processing (PSP) algorithm were analyzed and compared in detail. Simulation results show that, compared with PSP algorithm with L=4 for two co-frequency QPSK-modulated signals, Gibbs sampler can be nearly 17 times lower in computation complexity when the approximately performance is obtained. ©, 2015, Editorial Board of Journal on Communications. All right reserved.


Liao C.-H.,National Key Laboratory on Blind Signals Processing | Tu S.-L.,National Key Laboratory on Blind Signals Processing | Wan J.,National Key Laboratory on Blind Signals Processing
Dianzi Yu Xinxi Xuebao/Journal of Electronics and Information Technology | Year: 2014

An anti-frequency-offset algorithm is proposed by utilizing amplitude distribution feature for modulation recognition of conventional satellite modulations, such as QPSK, 16QAM and new modulations like 16APSK and 32APSK. The algorithm is based on adaptive construction of amplitude distribution template. After calculating the matching error between the amplitude distribution template and the actual amplitude distribution vector, the algorithm can recognize the modulation type by choosing the modulation type with the minimum matching error. This method does not need any prior knowledge about Carrier-to-Noise ratio (C/N), as well as threshold, and it is not sensitive to frequency offset. Because of these advantages, the algorithm is suitable for engineering application. Computer simulations show that the correct recognition probability is more than 98% when C/N is greater than 9 dB and 4000 symbols are used. It verifies the effectiveness of the algorithm.


Liao C.-H.,National Key Laboratory on Blind Signals Processing | Tu S.-L.,National Key Laboratory on Blind Signals Processing | Wan J.,National Key Laboratory on Blind Signals Processing
Tongxin Xuebao/Journal on Communications | Year: 2013

A per-survivor processing (PSP) based anti-frequency-offset algorithm was proposed for single-channel blind separation of burst GMSK mixed signals. The algorithm traces the phase changes caused by frequency offset and channel changes caused by other matters separately, which remarkably tones up the capacity against frequency offset. Simulation results show that the proposed algorithm can work well even when the unitary frequency offset (compared to symbol rate) grows up to 1.5×10-2, while the traditional PSP algorithm needs the unitary frequency offset less than 2.0×10-4. Furthermore, the proposed algorithm has no need for phase initialization, and a loss of only 1dB on performance can be achieved in comparison with the traditional PSP algorithm.


Liao C.-H.,National Key Laboratory on Blind Signals Processing | Tu S.-L.,National Key Laboratory on Blind Signals Processing | Wan J.,National Key Laboratory on Blind Signals Processing
Tongxin Xuebao/Journal on Communications | Year: 2011

An algorithm was proposed to exploit the codes in iterative single-channel blind separation and decoding of two co-frequency modulated signals with long-constraint channel codes. The output log-likelihood ratio (LLR) values of the decoder were used as a prior LLR fed back to the separator to improve the separation performance. Special emphases were laid on soft input soft output blind separation, maximum a posteriori (MAP) decoding and LLR interaction between separator and decoder. Simulation results show that the proposed algorithm can make a significant improvement in performance over the algorithms without iterations. For BPSK signals with (2, 1, 6) convolutional codes and random interleaving, a gain of about 2 dB in signal-noise ratio can be obtained after two iterations.

Loading National Key Laboratory on Blind Signals Processing collaborators
Loading National Key Laboratory on Blind Signals Processing collaborators