Multi-spectral echo signal processing for improved detection and classification of radar targets

Michael Vogt, Neumann, Thorsten, Michael Gerding, Christoph Dahl, Ilona Rolfes

Ger­man Micro­wa­ve Con­fe­rence (GeMiC 2016), pp. 309-312, doi: 10.1109/GEMIC.2016.7461618, Bo­chum, Ger­ma­ny, March 14-16, 2016


A major chal­len­ge in radar based re­mo­te sen­sing and ima­ging is to iden­ti­fy and to de­tect radar tar­gets, and also to ac­cu­ra­te­ly de­ter­mi­ne their lo­ca­ti­ons and sizes. This es­pe­ci­al­ly ap­p­lies in the case of mul­ti­ple, spa­ti­al­ly di­stri­bu­ted radar tar­gets, for ex­amp­le in radar ima­ging, radar level me­a­su­re­ment of bulk so­lids in silos, au­to­mo­ti­ve ra­dars, and others. In this paper, a con­cept for mul­ti-spec­tral ana­ly­sis and pro­ces­sing of echo si­gnals is pre­sen­ted and di­s­cus­sed, which has been in­ves­ti­ga­ted for scan­ning radar ap­p­li­ca­ti­ons. The pro­po­sed tech­ni­que is based on fil­te­ring the radar echo si­gnals in mul­ti­ple fre­quen­cy sub-bands and in­co­her­ent­ly com­bi­ning the fil­te­red si­gnals. Fur­ther­mo­re, the va­ri­an­ce of en­ve­lo­pe si­gnals is ana­ly­zed in order to allow for a dif­fe­ren­tia­ti­on bet­ween echoes from di­stri­bu­ted, ran­dom­ly ar­ran­ged scat­te­rers and from spa­ti­al­ly iso­la­ted sin­gle scat­te­rers. The mean over the stan­dard de­via­ti­on of the en­ve­lo­pe si­gnals ob­tained from the dif­fe­rent sub-bands is sug­gested to be used as an am­pli­tu­de-in­va­ri­ant pa­ra­me­ter for the clas­si­fi­ca­ti­on of radar tar­gets. Re­sults of an ex­pe­ri­men­tal eva­lua­ti­on of the con­cept using a me­cha­ni­cal­ly scan­ning 75 to 80 GHz Fre­quen­cy Mo­du­la­ted Con­ti­n­uous Wave (FMCW) radar sys­tem are pre­sen­ted. It will be shown that the pro­po­sed tech­ni­que enables to lar­ge­ly sup­press the echo si­gnal fluc­tua­ti­ons, which are given in sce­na­ri­os of spa­ti­al­ly di­stri­bu­ted radar tar­gets, and also to dis­tin­gu­ish bet­ween dif­fe­rent kinds of radar tar­gets.

[IEEE Li­b­ra­ry]