Detection And Parameter Estimation Of The Passive Moving Target In Mine Tunnel

Detection and parameter estimation are key technologies of passive positioningsystem. When a mining accident happens, the active positioning system has twodrawbacks: devices may get lost or damaged during the first-aid. In the thesis, I comeup with model of passive moving target positioning system in mine. Firstly, this thesisanalyzes characteristics of radio propagation characteristic by using ray tracingmethod in mine tunnel and draws a conclusion that characteristics of wirelesscommunication in mine are dense multipath, serious fading and delay spread in nslevel. In consideration of these characteristics, this thesis studies moving targetdetection and parameter estimation algorithms.Research on detection technology of moving target includes:(1)analyzing idea ofsuppressing strong interference, designing null matrix Based on Eigen valuedecomposition and simulating shows that can effectively suppress multipathinterference signal whose SNR is-10dB. Given characteristics of wirelesscommunication in mine are dense multipath and serious fading, due to null matrix cannot suppress low SNR fading multipath interference, this thesis designs and simulatesalgorithm based on the minimum norm, the simulation shows that this algorithm cannot suppress low SNR fading multipath interference whose SNR is-20dB.(2)Detecting echo signals after the above two algorithms, simulation shows when thetarget signal’s SNR is-15dB, both methods can accurately detect the target signal,when the target signal’s SNR is-25dB, null matrix can’t accurately detect the targetsignal but algorithm based on the minimum norm still works. This illustrates thatdetection sensitivity of algorithm based on the minimum norm is higher and simpler.Research on parameter estimation algorithms of moving target includes:(1)analyzing limitations inverse filter algorithm, MUSIC and its improved algorithm andfourth-order cumulants for TOA estimation using theory and simulation in mine, theresults show that MUSIC and its improved algorithm has higher resolution for DOAand TOA estimation but can’t suppress strong noise, fourth-order cumulants cansuppress Gaussian noise but resolution is low. Combining with characteristic is delayspread in ns level, so moving target signal is drowned in noise, this thesis proposesfourth-order cumulants Root-WMUSIC algorithm, simulation shows that thealgorithm has high resolution under low SNR and it can be applied to system better.