Proposal and analysis of new algorithms for wideband spectrum sensing in cognitive radio

The growing demand for available frequency bands in contrast to the scarcity of allocated spectrum slots, is promoting the development of smart wireless transmission terminals generally called cognitive radios. These devices are able to select the most suitable operating frequency and modulation scheme in order to minimize the probability of interference with other users operating in the same area. A key operation to reach this goal, is an accurate and reliable spectrum sensing, especially in environment characterized by the contemporaneous presence of many users, which could even transmit with a very low power. In order to implement this operation in cost-effective hardware, efforts for containing the computational burden should be also pursued. Several techniques for spectrum sensing are today present in literature, many of them require some knowledge about the context in which the system works, while further ones do not require any hypothesis on the signal to be detected. In this framework the authors propose and compare three algorithms based on a wideband analysis, possibly suited for the purpose. They all are thought for working in blind scenarios.