Traditional jamming is probably hard given the required power levels.
Lin and Noubir (google for "On Link Layer Denial of Service in Data
Wireless LANs") consider 802.11a and show that jamming just one
complete symbol will cause the frame to be retransmitted and so a
pulsed jamming signal could be effective.
You may want to take a look at the 802.16 working group document IEEE
C802.16e-04/406. The author reviews a number of MAC layer attacks but
also looks briefly at the carrier-sense PHY layer attacks. These
attacks are very low power when compared to traditional jamming
attacks and very effective. It looks like carrier-sense attacks would
be doable with a suitable radio front end and a USRP could very easily
do the job of generating the required carrier.