[time]
2022-08-16T02:57:54+02:00

[track]
7

[team_name]
LGD

[team_institution]
Army Engineering University of PLA

[logolink]


[team_members]
Zhizhao Cui;Weiheng Dai;Ao Huang;Liang Chen;Xioaqi Dai;Yaoquan Liu;Xiaoyu Ding

[reference_person]
Zhichao Cui

[reference_email]
cuizhic@163.com

[description_short]
The system adopts TDOA ranging method for location. The system first obtains the measures' channel impulse response waveforms, and idientifies its channel state by an offline-trained support vector machines (SVM) classified model. The identified measurements in non-line-of-sight channels are discarded in the location algriothm. For the model training phase, the first 512 bins of the CIR waveform need to be intercepted, and then six signal features such as peak value, maximum value, rise time, kurtosis, RMS Delay Spread, and maximum excess delay are extracted. On this basis, training is performed according to the labels. 
Identified LOS measurements are used for location. The calculation method of the time is to use the time from the start time to the first peak of 0.2 times as the time difference between the base station and the mobile station, and multiply the time difference by the speed of light to obtain the distance.
The SVM algorithm belongs to one of the machine learni

[description_long_link]
https://workdrive.zoho.com.cn/file/hwvkyb6db8ff3f5fb455594336df2aadde18d

[publish_check_]
true

[results_check_]
true

[data_check_]
true

[pdf_check_]
false

[video_check_]
false