Radar and method of updating background components of echo signal of radar

1. A method of updating background components in an echo signal of a radar, the radar being a frequency-modulated continuous wave radar and comprising an antenna, the method comprising:

the antenna receiving the echo signal;

generating digital data according to the echo signal;

generating M sets of N pieces of time-domain data in M respective time intervals from the digital data;

transforming the M sets of N pieces of time-domain data to a frequency domain to generate M sets of P magnitudes corresponding to P frequency bins, wherein the M sets of N pieces of time-domain data comprise spatial information of an object;

computing a pth reference value of M sets of magnitudes corresponding to a pth frequency bin of the P frequency bins;

computing a pth variance of the M sets of magnitudes corresponding to the pth frequency bin of the P frequency bins;

when a difference between the pth reference value corresponding to the pth frequency bin and the pth previous background component corresponding to the pth frequency bin exceeds a first threshold value and the pth variance corresponding to the pth frequency bin is less than a second threshold, using a reset mode as an operation mode;

when the difference between the pth reference value corresponding to the pth frequency bin and the pth previous background component corresponding to the pth frequency bin exceeds the first threshold value and the pth variance corresponding to the pth frequency bin exceeds the second threshold, using a frozen mode as the operation mode;

when the difference between the pth reference value corresponding to the pth frequency bin and the pth previous background component corresponding to the pth frequency bin is less than the first threshold value and the pth variance corresponding to the pth frequency bin is less than the second threshold, using a slow mode as the operation mode;

when the difference between the pth reference value corresponding to the pth frequency bin and the pth previous background component corresponding to the pth frequency bin is less than the first threshold value and the pth variance corresponding to the pth frequency bin exceeds the second threshold, using a fast mode as the operation mode; and

computing a weighted sum of the pth reference value and the pth previous background component corresponding to the operation mode to generate an adjustment value to update a pth background component of the P background components;

wherein M, N and P are positive integers exceeding 1, and p is a positive integer ranging between 1 and P inclusive.

2. The method of claim 1 , further comprising:

generating P sets of envelope data corresponding to the P frequency bins according to the M sets of P magnitudes corresponding to the P frequency bins;

computing P sets of difference data between the P sets of envelope data and P background components, the P sets of difference data corresponding to the P frequency bins, respectively; and

determining the spatial information of the object according to the P sets of difference data corresponding to the P frequency bins.

3. The method of claim 1 , wherein computing the weighted sum of the pth reference value and the pth previous background component corresponding to the operation mode to generate the adjustment value comprises:

setting a reference coefficient of the pth reference value to 0 and a background coefficient of the pth previous background component to 1 to generate the adjustment value in the frozen mode.

4. The method of claim 1 , wherein computing the weighted sum of the pth reference value and the pth previous background component corresponding to the operation mode to generate the adjustment value comprises:

in the slow mode, setting a background coefficient of the pth previous background component to be greater than a reference coefficient of the pth reference value; and

adding a product of the pth previous background component and the background coefficient and a product of the pth reference value and the reference coefficient to generate the adjustment value.

5. The method of claim 1 , wherein computing the weighted sum of the pth reference value and the pth previous background component corresponding to the operation mode to generate the adjustment value comprises:

in the fast mode, setting a background coefficient of the pth previous background component to be less than a reference coefficient of the pth reference value; and

adding a product of the pth previous background component and the background coefficient and a product of the pth reference value and the reference coefficient to generate the adjustment value.

6. The method of claim 1 , wherein computing the pth variance of the M sets of magnitudes corresponding to the pth frequency bin of the P frequency bins comprises:

computing a pth set of upper envelope data and a pth set of lower envelope data corresponding to the pth frequency bin of the P frequency bins; and

computing differences between the pth set of upper envelope data and the pth set of lower envelope data to generate the pth variance.

7. The method of claim 1 , wherein computing the weighted sum of the pth reference value and the pth previous background component corresponding to the operation mode to generate the adjustment value comprises:

setting a reference coefficient of the pth reference value to 1 and a background coefficient of the pth previous background component to 0 to generate the adjustment value in the reset mode.

8. The method of claim 7 , further comprising:

using the reset mode to compute the adjustment value during an initialization period of the radar.

9. A frequency-modulated continuous wave (FMCW) radar comprising:

an antenna configured to receive an echo signal; and

a processor coupled to the antenna and configured to generate M sets of N pieces of time-domain data in M respective time intervals from digital data, transform the M sets of N pieces of time-domain data to frequency domain to generate M sets of P magnitudes corresponding to P frequency bins, compute a pth reference value of M sets of magnitudes corresponding to a pth frequency bin of the P frequency bins, compute a pth variance of the M sets of magnitudes corresponding to the pth frequency bin of the P frequency bins, use a reset mode as an operation mode when a difference between the pth reference value corresponding to the pth frequency bin and the pth previous background component corresponding to the pth frequency bin exceeds a first threshold value and the pth variance corresponding to the pth frequency bin is less than a second threshold, use a frozen mode as the operation mode when the difference between the pth reference value corresponding to the pth frequency bin and the pth previous background component corresponding to the pth frequency bin exceeds the first threshold value and the pth variance corresponding to the pth frequency bin exceeds the second threshold, use a slow mode as the operation mode when the difference between the pth reference value corresponding to the pth frequency bin and the pth previous background component corresponding to the pth frequency bin is less than the first threshold value and the pth variance corresponding to the pth frequency bin is less than the second threshold, use a fast mode as the operation mode when the difference between the pth reference value corresponding to the pth frequency bin and the pth previous background component corresponding to the pth frequency bin is less than the first threshold value and the pth variance corresponding to the pth frequency bin exceeds the second threshold, and compute a weighted sum of the pth reference value and the pth previous background component corresponding to the operation mode to generate an adjustment value, so as to update a pth background component of the P background components;

wherein the M sets of N pieces of time-domain data comprise spatial information of an object;

wherein M, N and P are positive integers exceeding 1, and p is a positive integer ranging between 1 and P inclusive.

10. The FMCW radar of claim 9 , wherein the processor is further configured to:

generate P sets of envelope data corresponding to the P frequency bins according to the M sets of P magnitudes corresponding to the P frequency bins;

compute P sets of difference data between the P sets of envelope data and P background components, the P sets of difference data corresponding to the P frequency bins, respectively; and

determine the spatial information of the object according to the P sets of difference data corresponding to the P frequency bins.

11. The FMCW radar of claim 9 , wherein the processor is configured to:

set a reference coefficient of the pth reference value to 0 and a background coefficient of the pth previous background component to generate the adjustment value in the frozen mode.

12. The FMCW radar of claim 9 , wherein the processor is configured to:

set a background coefficient of the pth previous background component to be greater than a reference coefficient of the pth reference value in the slow mode; and

add a product of the pth previous background component and the background coefficient and a product of the pth reference value and the reference coefficient to generate the adjustment value.

13. The FMCW radar of claim 9 , wherein the processor is configured to:

set a background coefficient of the pth previous background component to be less than a reference coefficient of the pth reference value in the fast mode; and

add a product of the pth previous background component and the background coefficient and a product of the pth reference value and the reference coefficient to generate the adjustment value.

14. The FMCW radar of claim 9 , wherein the processor is configured to set a reference coefficient of the pth reference value to 1 and a background coefficient of the pth previous background component to 0 to generate the adjustment value in the reset mode.