A radio can share the wireless spectrum of an existing radio by measuring the first radio’s transmissions and adjusting its own transmissions to avoid interference.
© 2010 iStockphoto/avant-g
Delivery of web content and video over next-generation wireless networks will require large amounts of bandwidth. The existing wireless spectrum in most countries, however, has already been fully allocated. Optimizing the use of this spectrum is therefore necessary to allow further development of wireless services.
One promising approach is called ’cognitive radio’, whereby a primary, licensed user and an unlicensed, secondary user share a wireless spectrum. However, adoption of this method has been hindered by the inability to avoid signal interference, which must be kept low. Now, Rui Zhang and Ying-Chang Liang, from the A*STAR Institute for InfoComm Research in Singapore and Feifei Gao, currently of Jacobs University in Germany, have proposed a practical and efficient scheme to determine and avoid interference on channels shared by multiple users.
The team had previously proposed an approach for minimizing interference called ‘cognitive beamforming’. Under this scheme, a secondary radio uses multiple antennas—each transmitting at different powers—to modify its transmission parameters in a manner that avoids interference. However, this proposal required perfect and complete information about the primary radio and its channels to avoid interference, which made the scheme significantly less practical.
Under Zhang and co-workers’ new proposal, this stringent requirement is avoided because the secondary radio can ‘learn’ about the primary radio by periodically sampling its transmissions. The secondary radio can then numerically construct an ’effective interference channel’ that allows it to estimate the interference its transmissions would cause, and to alter them to minimize interference. The new proposal also allows for simultaneous primary and secondary transmissions at the same frequency, in contrast to other cognitive radio schemes.
Zhang and his co-workers also showed that a trade-off exists between the time the secondary radio spends learning to reducing the interference it causes, and the time it spends actually transmitting data. They calculated the optimum time spent at each activity, so as to maximize the secondary radio’s transmission speed.
The proposed scheme can be extended to multiple primary radio receivers and channels, and, while it has not yet been implemented, is potentially relevant to any wireless system that requires supporting two radio networks on a single frequency, according to Zhang. More generally, it “breaks the fundamental gridlock inherent to the conventional operation mode of cognitive radios,” says Zhang, “and the general cognitive beamforming approach on which it is based has already motivated considerable follow-up research.”
The A*STAR-affiliated researchers mentioned in this highlight are from the Institute for Infocomm Research.
