Tongfang Electronic Science and Technology Co.

Jiujiang, China

Tongfang Electronic Science and Technology Co.

Jiujiang, China
SEARCH FILTERS
Time filter
Source Type

Sun Y.,University of Electronic Science and Technology of China | Zhou Z.-P.,Tongfang Electronic Science and Technology Co. | Tang S.-L.,Tongfang Electronic Science and Technology Co. | Ding X.K.,Tongfang Electronic Science and Technology Co. | And 2 more authors.
International Conference on Signal Processing Proceedings, ICSP | Year: 2017

Hybrid source localization methods have proved to be able to improve the accuracy of position estimation. But several existing hybrid algorithms deal with two-dimensional (2D) scenarios only or require numerical search. This paper introduces a three-dimensional (3D) passive localization method using the active time of arrival (TOA) measurement and the one TOA and two angle of arrival (AOA) pairs observed at two stations. A closed-form solution is derived for the 3D source location with hybrid measurements from two stations. The experimental results demonstrate that the proposed method outperforms the state-of-art ones in terms of root mean-square error (RMSE), especially in large measurement error situation. © 2016 IEEE.


Wan Q.,University of Science and Technology of China | Yin J.-H.,University of Science and Technology of China | Zou L.,University of Science and Technology of China | Rao Z.-C.,TongFang Electronic Science and Technology Co. | Liu Y.-L.,China Electronic System Engineering Company
2013 IEEE 5th International Conference on Engineering Education: Aligning Engineering Education with Industrial Needs for Nation Development, ICEED 2013 | Year: 2014

In this lecture note, we relate some popular power density function (PDF) estimation methods to a unified and simple form to avoid student's tedious sense. It is relatively more easily to understand and facilitates the comparison with each other. First, we reveal the criteria to design different PDF estimators, including periodogram method, MVDR (minimum variance distortless response) method, AR (auto regressive) method and MUSIC (multiple signal classification) method. In this way, students will be able to clearly see the means of optimization of the various algorithms. Then we will derive the solution to different optimization problem. Finally, we express the PDF estimation as a unified linear form to show essential differences of these methods to students. © 2013 IEEE.


Fan R.,University of Electronic Science and Technology of China | Wan Q.,University of Electronic Science and Technology of China | Zhu H.,Tongfang Electronic Science and Technology Co.
Proceedings of 2011 IEEE CIE International Conference on Radar, RADAR 2011 | Year: 2011

A high-resolution range profile (HRRP) is the phasor sum of the time returns from different scatters on the target located within a resolution cell. In this paper, based on geometrical theory of diffraction (GTD) model, we analysis the whole process using traditional inverse discrete Fourier transform method (IDFT) to synthesis HRRP firstly. After that basis pursuit (BP), to obtain HRRP based on GTD model, is presented, which takes advantage of the fact that target HRRP is sparse in radar range window, so we can also obtain the synthesised range profile (SRP) correctly in presence of observation data loss. Compared with IDFT method, the SRP using sparsity as a prior (namely, using BP method) has higher resolution, lower sidelobe level and more robust to incomplete measurement data HRRP. Meanwhile, to avoid computational complexity in BP, we also presents greedy algorithm named Orthogonal Matching Pursuit (OMP) for approximating the solution of BP. Both theoretical analysis and simulation experiments illustrate the advantage of proposed method. © 2011 IEEE.


Ahmed H.I.,University of Electronic Science and Technology of China | Ahmed H.I.,Karary University | Qun W.,University of Electronic Science and Technology of China | Ramadan M.,Karary University | And 2 more authors.
Proceedings of 2016 IEEE Information Technology, Networking, Electronic and Automation Control Conference, ITNEC 2016 | Year: 2016

There is no doubt that the issue of estimate locations for unknown sources has been mattering a lot of debate in recent years, in the field of communication, Radar, Navigation, wireless sensor networks,...etc. In this paper, the multidimensional scaling (MDS) based on subspace method was applied to solve the addressed issue in the presence of distance information that acquired from TOA measurements. The Simulation results show the significant consequences compared with Cramer-Rao lower bound (CRLB) as well as the modified version of classical Multi-dimensional scaling, it became close to the lower bound of variance when the signal to noise ratio increased. © 2016 IEEE.


Gao Y.-F.,University of Electronic Science and Technology of China | Wan Y.-H.,TongFang Electronic Science and Technology Co. | Tang S.-L.,TongFang Electronic Science and Technology Co. | Xu B.-G.,TongFang Electronic Science and Technology Co. | Wan Q.,University of Electronic Science and Technology of China
2015 10th International Conference on Information, Communications and Signal Processing, ICICS 2015 | Year: 2015

This paper investigates a tensor-decomposition-based algorithm for the arrival angle estimation. With the assistance of multiple invariance modeling and uniqueness property of Canonical Polyadic Decomposition (CPD), the proposed method is capable of performing well even in worse conditions. Furthermore, by exploiting the feature of Vandermonde structure, a modified preprocessing step has been incorporated into this algorithm which can achieve a better performance than the standard CPD and classic subspace methods. Numerical experiments have demonstrated the feasibility and effectiveness of the proposed method. © 2015 IEEE.


Guo X.,University of Electronic Science and Technology of China | Guo X.,TONGFANG Electronic Science and Technology CO. | Chu L.,University of Electronic Science and Technology of China | Li B.,University of Electronic Science and Technology of China | And 3 more authors.
Progress in Electromagnetics Research | Year: 2014

We present a robust multiple-channel vector-matching localization approach (MCVM) based on signal strength difference (SSD) fingerprinting of ZigBee Network. Compared with some existing algorithms, our presented approach has threefold advantages: firstly, far fewer numbers of received signal strength(RSS) measurements and reference nodes are needed; secondly, it shows more robustness to the fluctuation of RSS; thirdly, it requires low time-consuming signal strength collection surveys in the location space. We demonstrate the performances of our algorithm experimentally using different numbers of channels, reference nodes and training points. The Cramer-Rao Low Bound(CRLB) of SSD is derived in order to compare the performance of the different localization methods addressed. The experiment results show the efficacy of our proposed approach.


Guo X.,University of Electronic Science and Technology of China | Wan Q.,University of Electronic Science and Technology of China | Xu B.,TONGFANG Electronic Science and Technology Ltd Company | Rao Z.,TONGFANG Electronic Science and Technology Ltd Company
2013 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2013 - Proceedings | Year: 2013

Spacial variability of Received Signal Strength measurements (RSS) is the fluctuation of measured RSS at a fixed location in difference time, which is the main reason for low accuracy and bad robustness of some existing indoor localization techniques. In this paper, we proposed a novel approach to eliminate space variability of RSS in indoor localization. The merits of our proposed method is twofold. First, our proposed method does not require the process of RSS fingerprinting constructing, which is a huge burden of localization methods based on RSS. Secondly, our proposed method is more robust than some existing methods, such as TOA-based method and RSS-based method, in big measurement error case. The efficacy of our proposed method are proven by means of some experimental results. © 2013 IEEE.


Zhang K.,University of Electronic Science and Technology of China | Ding X.K.,TongFang Electronic Science and Technology Co. | Wan Q.,University of Electronic Science and Technology of China | Gui G.,University of Electronic Science and Technology of China
IET Conference Publications | Year: 2015

The time delay estimation between two signals has always been an important issue. In this paper, we propose a new fast time delay estimation method of cubic phase signal based on Simplified Fractional Fourier transform (SFFT). First, 2-order instantaneous cross terms is introduced to transform cubic phase signal to chirp signal form. Then time delay is evaluated in the Simplified Fractional Fourier domain by measuring the time differential of two peak values in Fractional spectrum. Simulation shows better and faster estimation performance compared to GCC method at relatively low SNR.


Wan Q.,University of Electronic Science and Technology of China | Li Y.,University of Electronic Science and Technology of China | Wan Y.,TongFang Electronic Science and Technology Co. | Tang S.,TongFang Electronic Science and Technology Co. | And 2 more authors.
2015 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2015 - Proceedings | Year: 2015

Gain and phase errors calibration of sensor array plays an important role in practical applications. In this paper, an approach to sensor gains and phases estimation is proposed in the presence of multipath propagation. First, spatial spectrum of multipath signals is constructed based on the singular value decomposition of data covariance matrix. Then, direction-of-Arrival (DOA) of the non-line-of-sight (NLOS) signal is estimated through spectral peak searching while the exact sensor gains and phases are unknown. Finally, sensor gains and phases are determined by using both the known DOAs of LOS signals and the estimated DOA of NLOS signal. Though the proposed method can deal with multipath signal, it is not limited to a specific array configuration. Simulation results are provided to demonstrate the performance of the proposed method. © 2015 IEEE.


Han W.,University of Electronic Science and Technology of China | Ding X.K.,TongFang Electronic Science and Technology Co. | Zou L.,University of Electronic Science and Technology of China | Gui G.,University of Electronic Science and Technology of China | Wan Q.,University of Electronic Science and Technology of China
IET Conference Publications | Year: 2015

EMVS is able to stably perceive signals with different forms of polarization, and achieve a precise parameter estimation. As tensor method used in this kind of arrays, the performance is greatly improved. On the other hand, the traditional signal model often cannot intuitively describe the uncompletely polarized signal. In this paper, BTD from tensor algebra was adopted, while the modeling and analyzing for uncompletely polarized signal were given. In addition, Poyntmg's relationship was used to calculate the direction parameters explicitly. Smaller error compared to ESPRIT method was shown in simulation, especially at low SNR.

Loading Tongfang Electronic Science and Technology Co. collaborators
Loading Tongfang Electronic Science and Technology Co. collaborators