Cognitive Radio: Regulatory Implications of Different Methods of Cognition

At the time of writing this article, three broad techniques for identifying whether bands are free from use have been suggested. These are sensing, beacons, and geolocation, which importantly can be used discretely or in combination to effect the desired level of confidence in the attainment of a low-interference environment. In this article I discuss the regulatory implications of these techniques.

Geographical Databases
An alternative to sensing is for a cognitive device to precisely know its location and have access to a database listing the frequencies it is allowed to use at each location. This overcomes the regulatory issues associated, with sensing, but leads to other regulatory issues, such as;
■ To what accuracy should the device know its location?
■ Who will maintain the database? Will there be one provider for all bands or a separate database per band? What will the commercial arrangements be? Will there be monopoly concerns?
■ What availability is needed for the database? Is it acceptable for it to be off-air for substantial periods?
■ How will devices download updated versions of the database? How frequently should they do so? What will the loading on the spectrum be as a result?
■ What about the dynamic use of the spectrum for frequencies that are only occasionally in use?
These issues are only now starting to be considered by regulatory authorities in countries such as the United States and the United Kingdom. The use of databases does provide one interesting advantage for the regulators: the ability to control license-exempt devices in a manner that has never been possible before. So, for example, a regulator that wished to remove all cognitive devices from a band in a specific geography could do so by setting all frequencies in the database to “busy.” The cognitive devices then would be unable to transmit in this geography. Alternatively, the regulator could set aside cognitive device–free.

Beacon Reception
This approach requires the transmission of a signal from some appropriate infrastructure providing information on which frequencies are available for cognitive use in the vicinity. Cognitive devices tune to this channel and use the information provided to select their preferred frequency. This can be thought of as centralizing the database described in the previous section and setting the location to that reachable from the specific transmission source. If the cognitive device is unable to find a beacon, it is safest for it not to transmit, since it may be within the area covered by the beacon but shielded from receiving it. Hence, if it were to transmit on a randomly selected channel it could cause interference.
While this resolves the sensing problem at the cognitive device it raises many other problems:
■ Who provides the beacon signal? What are the commercial arrangements, and if there is only one provider, are there competition concerns?
■ How is the information the beacon transmits kept up to date, especially where
the licensed services are changing rapidly?
■ What spectrum is used for the beacon?
■ What technical parameters and protocols are used by the beacon transmitter?
■ How to prevent the beacon signal from being received outside its intended coverage area and as a result being applied incorrectly, and conversely how to make sure that it is available to all cognitive devices in the target area?
■ Should there be separate beacons for separate frequency bands or one beacon for all the bands into which cognitive access is allowed?
■ Is it acceptable for the use of cognitive devices to be denied access if the beacon fails or is taken off-air for any reason?
In part based on these questions, the beacon reception approach has not generated much interest in the regulatory community to date and hence few, if any, of these questions have been explored at a regulatory level.