Pulse-doppler radar
Pulse-doppler is a radar system that functions by sending short pulses of radio energy and simultanously listens for the echo from objects using the same attenna. The time delay between pulse transmission and echo reception gives the range to an object. Pulse-doppler radar uses the doppler shift principle to determine the relative velocity of objects. The Doppler effect is the change in frequency of any signal due to the finite speed at which the signal travels compared to the motion of the object. For instance, sound travels at the fairly slow speed of around 300 m/s, which is why you hear the Doppler effect of an ambulance siren as it passes you at 3 m/s or so. Although this results in a small 1% change in frequency, the human ear is very good at detecting this change.
In the case of radar the speed of light is much faster than sound and thus the resulting shift much smaller. However modern electronics are even better at detecting this change than the human ear is for sound. Speeds as slow as a few centimeters per second can be easily measured, an accuracy typically much better than for the measurement of distance.
The major use of pulse-doppler is to separate moving objects from clutter. It's common for doppler radars to have a frequency range adjust control to reject low speeds. Another form color-codes returns by their speed.
Doppler measures the speed only along the direction from the reflection to the radar antenna. In order to measure the object's true speed and direction, the radar set or operator had to remember a return's location. Military organizations traditionally used a manual plotting board for this purpose. Computers in the radar systems have made this even more convenient.
The rate of transmission is divided into three main groups low, medium, and high. After a transmission the receiver is turned on and listens for returning echos in fixed time slots, called range gates. If there are -m- range gates before a transmission occurs again, then 1/m is the rate or pulse repetion rate PRF. A medium PRF MPRF pulse doppler radar system transmits many times a second so that the reflections can be accumulated coherently using a FFT. In which the reflected signal adds but noise does not. However MPRF radar systems have to deal with the ambiguity in range and doppler that occurs. Range ambiguity occurs when a returning echo does not arrive before another transmission is made. And doppler ambiguity occurs when the speed of the target is greater than the PRF rate. To resolve a targets true range in MPRF radar multiple PRF's are used. First a set of tranmit pulses are made at one PRF, usually a power of 2, say 64 then the received range gates are fed into a FFT. The input to the FFT is a matrix of range gates vs time slots and the output of the FFT a matrix of range gates vs ambiguous doppler frequenices. The matrix called an ambiguity matrix. The matrix is processed by a constant false alarm rate, CFAR, algorithm to derive a vector of range gates in which targets occur. Where the targets are peaks in the ambiguity matrix, and the vector a yes/no answer to the question, does this range gate have a target? The CFAR finds these targets by comparing a matrix value to the values surrounding it, for each value in the matrix.
The PRF is changed and another set of 64 pulses sent out and received. This is done for several PRF's producing several vectors of 'detects'. When the vectors are unfolded a target shows up as an alignment of detects in different PRF's. Unfolding occurs by repeating the range gate detects of a PRF head to tail over and over again. So for example if m is 11 and there is a detect at range gate 7 then after unfolding the detect shows up at 7, 18, 29, ... Another PRF with m of say 13 has a detect at 3 so that after unfolding its detects thay show up at 3, 16, 29 ... As you can see there is a correclation at unfolded range gate 29 in this example. A unfolding mechenism like the chineese remainder theroem is typically used. Doppler is unfolded in a similiar fashion, but it done after range unfolding, so the vector of detects must keep with it corresponding information like the filter bank in the ambiguity matrix where the detect orginated.