The project RADAR Target Generation and Detection deals with generation of a FMCW (Frequency Modulated Continuous Wave) and receiving the trasmitted wave back at receiver after hitting the target. The RADAR here has a set of pre-defined characteristics for which the other required variables namely the Bandwidth (B), Chirp Time (Tchirp) and Slope (slope) are calculated.
Source: https://www.udacity.com/course/sensor-fusion-engineer-nanodegree--nd313
The beat signal is generated by subtracting the received signal frequency from the transmitted signal frequency. This can be done by multiplying both the signals elementwise in the time domain.
Initially 1D FFT is applied on the mixed signal ( Transmitted signal multiplied with received signal). Further details can be found in the code block.
The figure above is the result of 1D FFT on the mixed signal.
Now 2D FFT is implemented on the mixed signal such that it outputs an image that has response in the range and doppler bins.
After receiving the above output, CFAR (Constant False Alarm Rate) algorithm is used to detect the actual target and isolate it from the noise around the Cell Under Test(CUT). This way we obtain a sharp peak which indicates the target signal and suppresses the noise in the environment.
Hence, a constant false alarm rate is then achieved leading to smooth functioning of the radar sensor.
NOTE:
More information about the implementation is available in the form of comments in the code block.
The entire coding is done using MATLAB. Hence it is recommended to install MATLAB if not already installed. Cheers!