SOURCEFORGE » RALF

h1. Examples (radar)

Hereafter you will find : 

* experimental data set that were obtained with SDR-radar systems;

* processing output examples.

h2. Experimental radar data set from SDR-KIT 2400AD2 Ancortek 

Useful information about the SDR-KIT 2400AD2 Ancortek can be found at http://ancortek.com/

!{width:150px}resized-im2-radar.jpg! !{width:150px}resized-im1-radar-screen.jpg!

IQ data are recorded in MAT-file and can be read with the Matlab code attachment:"readdata_radar2400AD2.m"

IQ imbalance processing is likely to be needed.

table(centered).

|=. Data file |=. Waveform|=. Carrier frequency (GHz)|=. Bandwidth (GHz)|=. PRF (ms)|=. Samples per sweep (-)|=.Scenario description|

| attachment:"2017-02-09-10-34-32.mat"| LFMCW| 25| 2|1 | 128| RC car receding in corridor (non moving environment)|

||||||||

| attachment:"2017-07-27-16-57-05.mat"| LFMCW| 24.75| 1.5|4 | 128| No target, corridor with plants on ground in range-gate ca. 20-30 with hidden fan creating a windy environment|

| attachment:"2017-07-27-17-01-31.mat"| LFMCW| 24.75| 1.5|4 | 128| No target, corridor with plants on ground in range-gate ca. 20-30 with hidden fan off|

| attachment:"2017-07-27-17-07-46.mat"| LFMCW| 24.75| 1.5|4 | 128| RC car receding-closing in corridor with plants on ground in range-gate ca. 20-30 with hidden fan creating a windy environment 

attachment:"VID_20170727_170631_forge.mp4"|

||||||||

| attachment:"2017-07-27-13-57-54.mat"| LFMCW| 25|2|4|256| No target, corridor with plants on ground in range-gate ca. 54-82 with hidden fan creating a windy environment|

| attachment:"2017-07-27-14-05-00.mat"| LFMCW| 25|2|4|256| No target, corridor with plants on ground in range-gate ca. 54-82 with hidden fan off|

| attachment:"2017-07-27-13-42-41.mat"| LFMCW| 24.75| 1.5|4 | 128| RC car receding-closing in corridor with plants on ground in range-gate ca. 54-82 with hidden fan creating a windy environment 

attachment:"VID_20170727_134114_forge.mp4"|

h2. Examples of processing outputs

h3. Deramping processing

(section in construction)

h3. Doppler processing

(section in construction)

h3. Widedand processing

* Processing outputs of Bayesian sparse representation techniques [BidTAES2019].

In these figures, diffuse clutter is present at the selected range gates. The RC car is receding with a velocity approximately equal to $-v_a$.

The target is well estimated in the blind velocity with the sparse technique AROFF that jointly estimate the diffuse clutter and the target with its off-grid. In addition, only a few and low false estimations arise with AROFF unlike the 3 other sparse techniques.

|=. Outputs| Scenario|

|!{width:700px}movie-ssr-20170727134241.gif!|

dataset: 2017-07-27-13-42-41.mat

transparent background: coherent summation for wideband radar

diamond: output of sparse representation

* WON sparse technique assumes white noise and on-grid targets.

* WOFF sparse technique assumes white noise and off-grid targets.

* ARON sparse technique assumes autoregressive-colored noise and on-grid targets.

* AROFF sparse technique assumes autoregressive-colored noise and off-grid targets.

* The MMSE (Minimum Mean Square Error) estimator of target scene is depicted. 

$v_a$ is the ambiguous velocity|

h2. References

* [BidTAES2019] Stéphanie Bidon, Marie Lasserre, and François Le Chevalier. Unambiguous sparse recovery of migrating targets with a robustified Bayesian model. IEEE Transactions on Aerospace and Electronic Systems, 55(1) :108–123, Feb 2019. http://oatao.univ-toulouse.fr/22968/