rejection of WLNA signal at input of colocated Wimax receive - RF Cafe Forums

 Post subject: rejection of WLNA signal at input of colocated Wimax receive

Posted: Thu Nov 19, 2009 11:01 pm 

I am designing a transceiver where WLAN and WIMAX radios are colocated (on the same PCB).
consider an usage scenariuo when the WLAN (2.45 GHz) systems is transmitting and the WiMAX (2.51 GHz) is receiving. due to finite isolation between the antennas (<15 dB) of WLAN and Wimax, the WLAN signal will either saturate the WIMAX LNA or may produce some intermodulation product by mixing with other signals.
In the standard the minimum rejection requirements of WLAN signal at WIMAX receiver input is not specified.

I calculated that -40 dBm of any interferer signal would be desired to keep all IMD products at LNA input(IIP3=3) below the Noise Floor.

From this I found out the rejection required for WLAN signal. For example it the receiver antenan see a maximum of 5 dBm WLAN signal then the filter before LNA should attenuate the signal to -40 dBm so that any IMD generated by this signal is less than Noise floor.

However it seems that such filters are extremly difficult to realize.

I was wondering if this is the correct approach of finding the filter rejection requirements.

I would be grateful if anyone can share their views on this.

reagrds,
Ashish

_________________
Ashish Bondia,
Design Engineer- RF

 Post subject: Re: rejection of WLNA signal at input of colocated Wimax receive

Posted: Fri Nov 20, 2009 9:36 am 

Hi ashish
There are two separate issues here:
1) mutual interference that might be filtered
2) unacceptable high saturation of the non-linear devices
Regarding (1) I am not familiar with the relative relationship between mutual interferers in these particular systems, but in general with digital modulation, it is possible to operate even co-channel with interferers that are only -10dBc. Thus typically one obtains a Bit Error rate of say 1 in 10^-8 for a carrier to noise ratio of 10dB. This arises mainly because the modulation is quasi-synchronous, and the interferer signal clock will be out of phase coherence and thus appear as just noise.
Regarding (2) the resulting non-linearity products might be capable of being out-filtered using an elliptic function filter with the rapid roll-off side of the BPF placed between the two carriers.

_________________
At bottom, life is all about
Sucking in and blowing out.