Wireless technology Identification employing dynamic mode decomposition spectrum

Broadband wireless services have experienced significant growth, resulting in an ever-increasing demand for the limited spectrum space available for wireless services. As a solution for enhancing spectrum sharing, heterogeneous wireless networks (HetNets) have emerged as a promising technology that involves the coexistence and sharing of spectrum by multiple wireless technologies, including Wi-Fi, Bluetooth, LTE, and GSM. Accurately identifying wireless technologies in a shared environment is a crucial element in the development of HetNets coexistence protocols. The main objective is to optimize spectrum usage and minimize interference among coexisting users equipped with different wireless technologies. In this paper, a novel algorithm based on dynamic mode decomposition (DMD) mathematical modeling was developed to classify and differentiate among various wireless technologies. The developed technique utilized DMD frequency spectrum to identify the time domain signature of a signal. Experimental validation using an empirical dataset collected during various time series as well as power measurements of the targeted technologies demonstrated that the proposed technique can identify individual wireless signals with an accuracy exceeding 95%. Moreover, the proposed technique requires a short time- less than one second- to identify a signal, making it suitable for real-time practical networks. Compared to comparable techniques, the developed technique has lower complexity (shorter processing time, no channel estimation, no time/frequency synchronization, and no need for long observation-time intervals).