FSK demodulation with repetitive impulse jammer
![](https://d23s79tivgl8me.cloudfront.net/user/profilepictures/25706.jpg)
Hi,
I demodulate FSK (46 ks/s,3600/4800Hz,1200 Bd) and I have regular fat jammer pulses every 221 samples. Filtering away the pulses outside the useful spectrum 2k6-5k8Hz does not do the job. A median filter flattens out the pulses but kills the modulation as well.
Any ideas?
THX Cheers
Detlef
![](https://d23s79tivgl8me.cloudfront.net/user/profilepictures/1584.jpg)
Hi Detlef, looks like a tough problem - no answers so far, and this forum has some of the best and brightest RF and DSP minds.
My only thought so far is frequency hopping ... somehow the transmitter and receiver would know they are trying but getting jammed, so they change (pseudo-randomly) to another band, and continue trying. Basically similar to the Ethernet approach: try, and if there's a collision back off, wait some interval, and retry. Treat the jammer as if it's another band user.
-Jeff
![](https://www.embeddedrelated.com/new/images/defaultavatar.jpg)
Hi Detlef,
I want to know the shape of the pulse noise spectrum and the method of demodulation you use.
I guess you can not reject only the noise and you should use special method.
After seeing the spectrum, I'll work out an idea.
Best regards.
James.