I. Introduction
Signal classification has many applications in wireless communications for both civilian and military purposes. Most of the time, it is posed as an M-ary hypothesis testing problem. In our system, however, we are not only focusing on the type of the signal, but also detecting its physical layer features. The proposed UCSD has the ability to demodulate signals without benefit of any priori information and forms the heart of a truly cognitive receiver. Our software-based design at IF or quasi-baseband realizes a universal classification, synchronization and demodulation algorithm to meet a cognitive radio receiver's requirements for automatic configuration.