Automotive Adaptive Cruise Control Using FMCW Technology distance error

Hi there,

I have a question about the example of Automotive Adaptive Cruise Control Using FMCW Technology example. I configure the speed of radar and car to be 0 and configure noise power to be 0 for the receiver (instead of using noise figure). In this case, we should have a noise free estimation and no error should be introduced. For 43m car distance, the code gives a distance estimation of 42.9997. I changed the car distance to 43.2 and the code gives 43.0578 meter. So i wonder which part of the code introduce the error?

[Arun]

Hi Dai, could you please share your modified code for this example?

Answer 1

[Pratyush]

Hi Yucheng,

In the MATLAB example of Automotive Adaptive Cruise Control Using FMCW Technology, even when you set the speed of the radar and car to 0 and the noise power to 0 for the receiver, you might still observe small errors in distance estimation due to numerical precision limitations in digital computation. These errors can stem from:

• Rounding errors in floating-point representations.
• Numerical errors from digital signal processing algorithms like FFT.
• Quantization errors from discretizing the continuous-time signal.
• Finite range resolution of the radar system.
• Simplifications or assumptions in the modeling process.

To find the source of the error, you should inspect the code, especially where the beat signal is generated, processed with an FFT, and converted from frequency to range estimate. Small discrepancies are common and often acceptable within the precision limits of digital simulations.