Interference mitigation by dynamic false rate

1 . A method performed by a radar system, the method comprising:

receiving a radar signal at receive antennas;

processing the received radar signal with analog signal processing circuitry;

digitizing the processed radar signal using at least one analog-to-digital converter (ADC) to generate radar data;

executing, by a digital signal processor, a range fast Fourier transform (FFT) and a Doppler FFT on the radar data to generate range and Doppler data values;

selecting one or more training cells for noise power estimation, from which radar signals have been received;

calculating a mean noise power value for the one or more training cells;

for each of the one or more training cells, calculating a deviation value by determining a difference between the noise power of the training cell and the mean noise power value;

calculating a standard deviation of the deviation values for the one or more training cells;

determining a trained probability of false alarm (PFA) threshold offset by multiplying the standard deviation by a scaling factor k, wherein k is obtained from an inverse cumulative distribution function corresponding to a desired PFA; estimating a PFA threshold to be applied for a non-training cell;

setting a current PFA threshold offset;

comparing the current PFA threshold offset to the trained PFA threshold offset and a guard interval offset;

replacing the current PFA threshold offset with the trained PFA threshold offset only when an absolute difference between the current PFA threshold offset and the trained PFA threshold offset exceeds the guard interval offset;

controlling the analog signal processing circuitry to mix a radar signal with a carrier signal to shift a center frequency of the radar signal and to hop between frequency subbands used for transmission and receiving, and to perform amplification operations on radar signals output by the receive antenna;

digitizing the radar signal after said mixing and amplification using at least one analog-to-digital converter (ADC) to generate radar data; executing, by a digital signal processor, a range fast Fourier transform (FFT) and a Doppler FFT on the radar data to generate range and Doppler data values;

applying a cell-averaging constant false alarm rate (CA-CFAR) detection routine to the radar data using an adjusted PFA threshold offset; and

populating one or more multi-dimensional point clouds with a detected target cell corresponding to the non-training cell.

2 . The method of claim 1 , wherein:

the non-training cell is one of a range cell, an angle cell, and a Doppler cell;

wherein the training cells are one or more of range cells, angle cells, and Doppler cells;

and wherein the training cells are not the same type of cell as the non-training cell.

3 . The method of claim 1 , wherein the training cells are selected from among cells that neighbor the non-training cell.

4 . The method of claim 1 , wherein the selecting, calculating, and estimating steps are performed offline during a training phase.

5 . The method of claim 1 , wherein the selecting, calculating, and estimating acts are performed online in real time, and wherein estimating the PFA threshold offset comprises:

calculating a difference between a noise power of respective training cells and a mean noise power across all training cells;

determining an absolute value of the calculated difference; and

estimating the PFA threshold offset based upon the absolute value of the calculated difference.

6 . The method of claim 1 , performed on a frame-by-frame basis.

7 . The method of claim 1 , wherein the analog signal processing circuitry comprises a mixer configured to combine the received radar signal with a local oscillator signal prior to digitization by the analog-to-digital converter.

8 . The method of claim 1 , wherein the programmable guard interval offset is dynamically adjusted based on environmental conditions detected by the radar system, including interference level or ambient noise.

9 . The method of claim 1 , wherein the one or more multi-dimensional point clouds are transmitted to a central processing unit or computer for further processing, and the radar system is configured to store at least the point cloud and the corresponding adjusted PFA threshold offset in a memory or data store.

10 . A radar system comprising:

receive antennas that receive a radar signal; and

one or more processors configured to perform acts comprising:

processing the received radar signal with analog signal processing circuitry;

digitizing the processed radar signal using at least one analog-to-digital converter (ADC) to generate radar data; executing, by a digital signal processor, a range fast Fourier transform (FFT) and a Doppler FFT on the radar data to generate range and Doppler data values;

selecting one or more training cells for noise power estimation based upon the radar signal received at the receive antennas;

calculating a mean noise power value for the one or more training cells;

for each of the one or more training cells, calculating a deviation value by determining a difference between the noise power of the training cell and the mean noise power value;

calculating a standard deviation of the deviation values for the one or more training cells;

determining a trained probability of false alarm (PFA) threshold offset by multiplying the standard deviation by a scaling factor k, wherein k is obtained from an inverse cumulative distribution function corresponding to a desired PFA estimating a PFA threshold to be applied for a non-training cell;

setting a current PFA threshold offset;

comparing the current PFA threshold offset to the trained PFA threshold offset and a guard interval offset;

replacing the current PFA threshold offset with the trained PFA threshold offset only when an absolute difference between the current PFA threshold offset and the trained PFA threshold offset exceeds the guard interval offset;

controlling the analog signal processing circuitry to mix a radar signal with a carrier signal to shift a center frequency of the radar signal and to hop between frequency subbands used for transmission and receiving, and to perform amplification operations on radar signals output by the receive antenna;

digitizing the radar signal after said mixing and amplification using at least one analog-to-digital converter (ADC) to generate radar data; executing, by a digital signal processor, a range fast Fourier transform (FFT) and a Doppler FFT on the radar data to generate range and Doppler data values;

applying a cell-averaging constant false alarm rate (CA-CFAR) detection routine to the radar data using an adjusted PFA threshold offset; and

populating one or more multi-dimensional point clouds with a detected target cell corresponding to in the non-training cell.

11 . The radar system of claim 10 , wherein:

the non-training cell is one of a range cell, an angle cell, and a Doppler cell;

the training cells are one or more of range cells, angle cells, and Doppler cells; and

the training cells are not the same type of cell as the non-training cell.

12 . The radar system of claim 10 , wherein the training cells are selected from among cells that neighbor the non-training cell.

13 . The radar system of claim 10 , wherein the selecting, calculating, and estimating acts are performed offline during a training phase.

14 . The radar system of claim 10 , wherein the selecting, calculating, and estimating acts are performed online in real time, and wherein estimating the PFA threshold offset comprises:

calculating a difference between a noise power of respective training cells and a mean noise power across all training cells;

determining an absolute value of the calculated difference; and

estimating the PFA threshold offset based upon the absolute value of the calculated difference.

15 . The radar system of claim 10 , wherein the acts are performed on a frame-by-frame basis.

16 . A radar analysis system comprising:

one or more processors configured to perform acts comprising:

processing a received radar signal with analog signal processing circuitry;

digitizing the processed radar signal using at least one analog-to-digital converter (ADC) to generate radar data;

executing, by a digital signal processor, a range fast Fourier transform (FFT) and a Doppler FFT on the radar data to generate range and Doppler data values;

selecting one or more training cells for noise power estimation based upon a radar signal received at one or more receive antennas;

calculating a mean noise power value for the one or more training cells;

for each of the one or more training cells, calculating a deviation value by determining a difference between the noise power of the training cell and the mean noise power value;

calculating a standard deviation of the deviation values for the one or more training cells;

determining a trained probability of false alarm (PFA) threshold offset by multiplying the standard deviation by a scaling factor k, wherein k is obtained from an inverse cumulative distribution function corresponding to a desired PFA;

estimating a PFA threshold to be applied for a non-training cell;

setting a current PFA threshold offset; comparing the current PFA threshold offset to the trained PFA threshold offset and a guard interval offset;

replacing the current PFA threshold offset with the trained PFA threshold offset only when an absolute difference between the current PFA threshold offset and the trained PFA threshold offset exceeds the guard interval offset;

controlling the analog signal processing circuitry to mix a radar signal with a carrier signal to shift a center frequency of the radar signal and to hop between frequency subbands used for transmission and receiving, and to perform amplification operations on radar signals output by the receive antenna;

digitizing the radar signal after said mixing and amplification using at least one analog-to-digital converter (ADC) to generate radar data; executing, by a digital signal processor, a range fast Fourier transform (FFT) and a Doppler FFT on the radar data to generate range and Doppler data values;

applying a cell-averaging constant false alarm rate (CA-CFAR) detection routine to the radar data using an adjusted PFA threshold offset; and

populating one or more multi-dimensional point clouds with a detected target cell corresponding to in the non-training cell.

17 . The radar analysis system of claim 16 , wherein:

the non-training cell is one of a range cell, an angle cell, and a Doppler cell;

the training cells are one or more of range cells, angle cells, and Doppler cells; and

the training cells are not the same type of cell as the non-training cell.

18 . The radar analysis system of claim 16 , wherein the training cells are selected from among cells that neighbor the non-training cell.

19 . The radar analysis system of claim 16 , wherein the selecting, calculating, and estimating acts are performed offline during a training phase.

20 . The radar analysis system of claim 16 , wherein the acts are performed on a frame-by-frame basis.