fig 1. Deramp signals
The radar emits a modulated chirp s(t) of duration T in a frequency-band B towards the Earth’s surface, then, with a delay corresponding to the estimated return time of the emitted chirp, another which is slightly shifted in frequency. By mixing the returning and deramping chirps, the frequency shift can be estimated, which, using Fourier transforms, gives the time delay.
To obtain a resolution of 3.125 ns, pulses of this duration can be used; this is the approach used in laser ranging systems. Maintaining the strength of the return signal requires a certain amount of energy in the pulse, and with such short duration pulses, a very high transmit power is required. The approach commonly used in radar systems is to inject a short pulse into a dispersive delay line, which spreads the energy over time, generating a frequency modulated or chirp signal. When the echo is received, the radar passes the signal through an inverse matched filter which compresses the chirp signal back to a short pulse. This technique is called pulse compression. The compressed time resolution is inversely proportional to the chirp bandwidth.
A radar range resolution of 3.125 ns would require a very high radio frequency bandwidth, of the order of several hundred megahertz. To generate chirp signals of this bandwidth, the frequency of the signals from the dispersive delay line has to be multiplied. This entails difficulties in the receiver matched filter as the corresponding frequency division is not possible. Even if such a filter were available, video signal handling at a similar high bandwidth would be necessary.
To circumvent this problem, the full-deramp technique is used in combination with linear FM (Frequency Modulation) pulse compression. The return echo signal consists of many discrete chirps, each reflected from a different facet of the ocean surface, with slightly different delay times. The full-deramp concept consists in mixing this incoming signal with a replica of the transmitted chirp, slightly shifted in frequency. The deramp mixer generates signals which are the frequency difference between its two inputs. As both inputs have the same rate of frequency change, the output frequencies are constant tones. The input signals are linear so there is a mapping of time offset onto frequency offset. As a result, targets with a different range give echoes at different frequencies. Therefore, the range discriminator can be implemented with a bank of contiguous filters. The translation from the time domain to the frequency domain simplifies the signal processing stages as they are able to work with much-reduced bandwidth.