Specific Emitter Identification for Cognitive Radio with Application to IEEE 802.11

Abstract

Cognitive radio (CR) is believed to be an enabling technology for increasing spectrum efficiency. A CR collects spectrum usage information from not only its own spectrum sensing module, but also from peer CRs. The heavy dependence on spectrum knowledge from other CRs requires identification of malicious CR devices that could generate spoofed information. In addition, it also needs to track the users associated with problematic CR devices which unintentionally violate spectrum usage etiquette. The specific emitter identification (SEI) concept is applied to identification of such non-cooperative CR devices. In this paper, second-order cyclic features of OFDM signals are proposed as a means of increasing CR network security and stability through SEI. For this exploratory work, IEEE 802.11a/g signals from different WLAN cards are measured and classified using hidden Markov Models (HMMs).