Wireless Technology Identification Employing Dynamic Mode Decomposition Modeling

Significant growth in broadband wireless services, as well ever-increasing demand on the spectrum caused by Internet of Things (IoT) have overstretched limited available space for services. Heterogeneous networks (HetNets)—wherein multiple technologies (e.g., Wi-Fi, Bluetooth, Zigbee, LTE, and GSM) coexist share spectrum—are a promising solution enhancing sharing. An essential element developing coexistence protocols is correctly identifying anticipated to shift users between an effort optimize usage minimize interference. For research reported this paper, we analyzed performance our developed novel algorithm based dynamic mode decomposition (DMD) mathematical modeling identify differentiate among various technologies. More specifically, technique identified GSM LTE signals cellular domain, IEEE802.11n, ac, ax Wi-Fi Bluetooth Zigbee. The proposed DMD-based identifies time domain signature signal capturing embedded periodic features transmitted within signal. Performance accuracy were tested validated using experimental dataset collected series, raw-power measurements targeted Results showed that can classify individual coexisting with high —greater than 90% most cases. Furthermore, only short time— less one second—is required enabling implementation real-time practical networks. advantage over comparable techniques lower complexity (i.e., shorter processing training time, no channel estimation, time/frequency synchronization, need long observation-time intervals).