Investigation of Steady State Features in Emitters Classification

الملخص

The security of the communication network has become an important issue. Because of most wireless communication networks are exposed to different types of penetrative attacks such as stray, spoofing, and interaction of communication signals. Radio Frequency (RF) Fingerprinting has provided a promising solution for communication network security. In this study, RF fingerprinting of steady state potion of Bluetooth (BT) has been applied to solve this problem. Set of Bluetooth (BT) signals has been collected from different mobile phones to generate a preliminary raw data set. Successive of preparation processes applied to the collected BT signal data set to generate signals’ features data. These processes are converting signals from text to digital, cantering and normalizing the digital BT signals, determination of steady sate portions, and Hilbert-Huang Transform (HHT) along with Empirical Mode Decomposition (EMD). By applying HHT, and EMD to the signals Time Frequency Energy Distributions (TFED) are obtained. By means of the signals’ energy envelopes and the signals’ steady state, and their TFEDs, signals’ features are extracted. The extracted features represent the input data set of classifiers. A learning machine technique is applied to classify and identify the transmitter device. Part of data set is used to learn the classifiers, while the rest of the data is used to test the classifiers performance. The performance of the classifiers is evaluated for different levels of signal to noise ratio (SNR). The results of this study demonstrate the usability of steady state of RF fingerprinting for BT signals at physical layer security of wireless networks, and the effectiveness of the applied processes and introduced classifiers.