Sampling Only When SNR > Some Limit
in [DSP]
|
From: Tim Wescott on 3 Aug 2010 14:51 On 08/03/2010 10:58 AM, Mark wrote: > >> >> * The technique that I know of is for narrow-band FM receivers: the >> audio is of a known bandwidth, and when the receiver is getting a strong >> signal the output of the detector has a spectrum within that band. So >> you just high-pass filter the detector output: when your radio starts >> seeing a lot of energy in that high band, it decides "ah! noise!" and >> cuts the audio off. >> >> -- >> > Actually a true noise blanker in a NBFM radio is even neater.... they > use small seperate receiver tuned to an unused nearby frequency. > This aux receiver is relativly wideband compared to the voice > channel. When an impulse noise event is detected in the wideband > receiver it opens a gate and BLANKS the path in the narrow band > receiver. And here is the neat part, the gate in the narrow band > receiver is BEFORE the narrow band filter because once the impulse > hits the narrow band filter, the game is over because the filter will > output the ..(wait for it) impulse response of itself which in > general is MUCH longer time duration compared to the impulse itself. > > So by opening or blanking the impulse and preventing it from hitting > the narrow band filter, you can make a big reduction in the noise > comming out of the filter with only a small loss of the desired > signal. > > This worked well to protect against ignition noise. It works well if you can count on unused bandwidth, and broadband noise. Broadband noise is pretty trustworthy, but the unused channel depends on the environment. I've also had good success in a data receiver by looking for an output from the bit integrator that is much larger or smaller in amplitude than expected -- if it is, then you can report that bit as an erasure or some degree of uncertainty to the FEC algorithm (if you have it). -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" was written for you. See details at http://www.wescottdesign.com/actfes/actfes.html |