|
|
|
|
|
Matched FilterA matched filter is obtained by correlating with a time-reversed version of a signal (i.e. convolution) that we are looking for. Matched filters are commonly used in radar, in which a signal is sent out, and we measure the reflected signals, looking for something similar to what was sent out. Pulse compression is another name for matched filtering. As a general rule of thumb, pulse width determines bandwidth of the filter we use. For example, if we send out a pulse having a pulse width of one microsecond, we'd generally use a filter having a bandwidth on the order of one megahertz, whereas a pulse of width one nanosecond would generally be picked up with a filter having a bandwidth on the order of one gigahertz. Shorter pulses spread the energy over wider bandwidth and thus need wider band filters that also admit more noise. Thus a longer pulse allows the filter to be more selective in frequency, but at the same time, is inherently less selective in time. This tradeoff establishes an uncertainty principle along the cannonical conjugate variables of time and frequency, or equivalently, along range and doppler. A plot that is a function of range and doppler is known as a radar ambiguity function or radar ambiguity diagram. Example of matched filter As a specific example, suppose that we are looking for a signal with a frequency of plus one hertz. We correlate (compare) every incoming signal with a signal of minus one hertz in order to detect or measure the plus one hertz signal, since time-reversal of a positive frequency makes it into a negative frequency.
|
 |
|
| Copyright 2005-2009 OnPedia.com. All Rights Reserved |
|
|