Signal classification and cognitive sensor network

Abstract

Signal classification is an important subject for military radio communications. With the revolution of digitizing communications ever closer to the antenna, commercial cognitive radios with programmable modulation schemes and adaptive transmission rates have become a practical wireless communication platform. Signal classification techniques have attracted much attention recently by the cognitive network applications in developing the next generation radio receivers and sensor networks with built-in automatic signal detection and classification capabilities. However, the technical expectations and goals in commercial applications are quite different with that in military communication systems. A key research area is to develop new algorithms with the low-cost real-time adaptive demodulation capability. An overview of automatic signal classification techniques and the challenges in migrating current signal classification methods into the cognitive radios and sensor network will be discussed.