AWR1642BOOST: medium range and high range resolution profile

Parth Raj Singh

Part Number: AWR1642BOOST
Other Parts Discussed in Thread: MMWAVE-STUDIO

With the help of mmWaveSensingEstimator 1.2.0, I have generated a profile (attached). I am not interested in velocity mesurement therefore I want to skift doppler FFT. However, According to the mmWaveSDK demo, there must be at least 16 doppler bins. I would like to know if the AWR1642Boost can be programmed with the profile I generated. I would be glad if someone can tell, what changes I should make in short range radar example of mmWave Automotive Toolbox for this porfile to work.

Fullscreen design_110m_3.8cm.json.txt Download

{
    "platform": "xWR1642",
    "num_rx": 4,
    "num_tx": 2,
    "tx_gain": 12.5,
    "rx_gain": 48,
    "frequency_range": "77 - 81",
    "maximum_bandwidth": 4000,
    "tx_power": 12.5,
    "ambient_temperature_degC": 20,
    "maximum_detectable_range": 110,
    "range_resolution": 3.8,
    "maximum_velocity_kmph": 3,
    "velocity_resolution_kmph": 11,
    "measurement_rate": 800,
    "typical_detected_object": "Car",
    "detection_loss": 1,
    "system_loss": 1,
    "implementation_margin": 2,
    "detection_SNR": 12,
    "maximum_radar_cube_size": 768,
    "maximum_RF_bandwidth": 5,
    "maximum_sampling_frequency": 6.25,
    "sensor_maximum_bandwidth": 4000,
    "maximum_allowed_bandwidth": 4000,
    "starting_frequency": 77,
    "maximum_velocity": 0.8333333333333334,
    "velocity_resolution": 3.0555555555555554,
    "valid_sweep_bandwidth": 3947.3684210526317,
    "idle_time": 7,
    "adc_valid_start_time": 4.2,
    "excess_ramping_time": 1,
    "periodicity": 1.25,
    "ambient_temperature": 293.15,
    "noise_figure": 16,
    "num_virtual_rx": 8,
    "non_coherent_combining_loss": 3,
    "rcs_value": 5,
    "combined_factor_in_dB": 38,
    "combined_factor_linear": 6309.57344480193,
    "inter_chirp_time": 12.2,
    "aux_comp_coeff_a": 78973684210.52632,
    "aux_comp_coeff_b": -44019942.10526316,
    "chirp_time": 557.4001332889036,
    "ramp_slope_init": 7.081750048679462,
    "ramp_slope_parameter": 147,
    "ramp_slope": 7.097125053405762,
    "aux_comp_T1": 1,
    "maximum_beat_frequency": 5.204558372497559,
    "sampling_frequency": 5.782842636108398,
    "number_of_samples_per_chirp": 3224,
    "total_sweep_bandwidth": 3993.632323004983,
    "idle_time_minimum": 7,
    "ramp_end_time": 562.7112799835086,
    "carrier_frequency": 77.02981078992607,
    "aux_comp_T2": 3.2,
    "adc_valid_start_time_2": 4.2,
    "lambda": 3.894596091091968,
    "max_chirp_repetition_period": 1168,
    "chirp_repetition_period": 1139,
    "num_range_fft_bins": 4096,
    "min_num_of_chirp_loops": 1,
    "max_range_for_typical_detectable_object": 737.1004795921523,
    "min_rcs_detectable_at_max_range": 0.0024798988652802523,
    "num_doppler_fft_bins": 1,
    "active_frame_time": 1.139,
    "range_inter_bin_resolution": 2.9910156249999997,
    "velocity_inter_bin_resolution": 3.0555555555555554,
    "radar_cube_size": 128
}

over 7 years ago

0 Cesar over 7 years ago

TI__Guru*** 149204 points

Hi,

Why are you interested in teh Short Range Radar example? Are you looking into an automotive Application?

Thank you
Cesar

0 Parth Raj Singh over 7 years ago in reply to Cesar

Prodigy 80 points

Hi,

Thank you for the response.

Actually, I have recently started working on AWR1642boost kit. Consequently, I have tested demo in mmwave sdk using mmWave demo visualizer and a python script; and I have also modified and tested SRR example in Automotive Toolkit. I am not looking for an automotive application however I am working on an autonomous robot application to generate an map of radius 100-200 meters w.r.t. the radar using SLAM.

SRR example is configured for 80m range and ~37cm range resolution which is close to my application requirements. Consequently, I am interested in it.

Regards

Parth

0 Nitin Sakhuja over 7 years ago in reply to Parth Raj Singh

TI__Mastermind 30405 points

Hi Parth,

The SDK Out-of-box demo processing chain is programmed for range, Doppler and angle of arrival processing and you cannot bypass Doppler processing with this chain. This is just the way the OOB demo software is implemented but you can implement a processing chain for just range and angle of arrival processing. One example of such a processing chain is implemented in the People Tracking and Counting Reference Design using mmWave, which computes a range-azimuth heatmap first followed by object detection and Doppler estimation. Please note that this chain is optimized for short ranges (6 to 14m) so it would not work with your configuration parameters, but it shows you an example of how to do range-azimuth processing instead of range-Doppler processing as given in the SDK OOB demo.

Coming back to your configuration parameters, the sensing estimator generates parameters for the Radar Front-end and if there are no errors shown on the Sensing Estimator, the parameters define a valid Radar front-end configuration. These parameters, while valid for the sensor front-end, may not work with the SDK Demo as the digital processing in the SDK demo may need to be modified for the processing requirements of the programmed front-end configuration.

You can use MMWAVE-STUDIO along-with the DCA1000 Data Capture card to capture raw data for your configuration and visualize the various plots provided by MMWAVE-STUDIO (range-doppler, range-angle, 1D FFT, 2D FFT) and tune the configuration. This way, you can enter the configuration parameters generated by Sensing estimator directly in MMWAVE-STUDIO sensor config tab and capture and visualize the raw data. You can then look into designing your own digital processing chain using the SDK OOB demo or SRR design, or People counting or Traffic Monitoring reference chains as reference.

Thanks

-Nitin

0 Parth Raj Singh over 7 years ago in reply to Nitin Sakhuja

Prodigy 80 points

Hi Nitin,

Thanks for your helpful response.

Regards,
Parth

Sensors

Sensors forum

AWR1642BOOST: medium range and high range resolution profile