WARP Project Forums - Wireless Open-Access Research Platform

You are not logged in.

#1 2007-Jun-15 11:44:22

toolbox
Member
Registered: 2007-Jun-02
Posts: 18

Poor Video and Audio Quality

I have been able to run the uniMacServer and UniMacClient project. However, the received video and audio quality is very poor. The server and client are kept within 2 metres of each other and there are no obstacles in direct line of sight.

Is there a way of improving the reception by changing the transmitting power or adjusting the antennas positions etc? Can we use some error correcting codes?

There are no other nodes except a server and a client contending for the medium.

Interestingly, the audio quality reception using noMac is better than that of uniMac (in the same setup) perhaps because there is only one node contending for medium and moreover sequencing, acknowledgments are absent.

Kindly suggest on how to improve the reception.

Last edited by toolbox (2007-Jun-15 11:50:41)

Offline

 

#2 2007-Jun-15 15:53:40

chunter
Administrator
From: Mango Communications
Registered: 2006-Aug-24
Posts: 1212

Re: Poor Video and Audio Quality

It sounds to me like you are working in a busy medium. Currently, the threshold on energy for carrier sensing (CSMA) is fairly arbitrary and may be too low. If that is the case, the uniMac server senses the medium and simply refuses to transmit more often than not. The noMac project has no carrier sensing, so this problem isn't seen. There are a couple of things you can try:


-use the warpmac_setChannel function to move to a different part of the spectrum that might be less busy
-in the unimac code for the server, try changing warpmac_enableCSMA to warpmac_disableCSMA and see if the performance subjectively matches the audio quality of the noMac project
-if the performance is better, change it back to using warpmac_enableCSMA and change the threshold in that function in warpmac.c (http://warp.rice.edu/trac/browser/Platf … ac.c#L1194) to something higher, like 9000. This value is empirically driven; it's a research question in itself to find the optimal value


-Chris

Offline

 

Board footer