To tackle cross-technology interference, researchers from IDLab – an imec research group at the University of Ghent and Antwerp – have developed two machine learning solutions. First, they identify which technologies are in the air (technology recognition) and how they are using the available bandwidth (traffic recognition). The next step is to develop an algorithm for automatic finegrained spectrum allocation, which can ensure that the full potential of the spectrum is used and collisions are avoided.
As our collective hunger for mobile data seems to be insatiable, the available spectrum will soon be overloaded. As more and more different technologies (such as Bluetooth, Wi-Fi, IEEE 802.15.4, etc.) are operating in shared spectrum bands, cross-technology interference becomes more common. Your Wi-Fi might be disrupted when your microwave is switched on; the YouTube video on your phone has trouble loading because your neighbor is using Wi-Fi to skype. In most contexts, delays and disruptions are merely annoying, but in an industrial environment they can be dangerous.
To tackle cross-technology interference, researchers from IDLab – an imec research group at the University of Ghent and Antwerp – have developed two machine learning solutions. First, before we can manage the spectrum, we have to identify which technologies are in the air (technology recognition) and how they are using the available bandwidth (traffic recognition). In the first phase, an autoencoder approach for technology recognition is developed. It can identify over-the-flytechnologies sharing the 2.4 GHz and 5GHz bands, i.e. LTE, Wi-Fi, IEEE 802.15.4 and Bluetooth Low Energy. The model can be trained with samples, only a limited number of which need to be labeled. In the second phase, a machine learning approach is used to identify traffic patterns, e.g. discovering whether a certain technology uses the spectrum constantly or in bursts.
The next step is to develop an algorithm to decide how different co-existing technologies can use the spectrum and, in the long term, to predict future spectrum occupation and optimize network performance.
Want to know more?
Adnan Shahid (M’15, SM’17) received the B.Eng. and the M.Eng. degrees in computer engineering with wireless communication specialization from the University of Engineering and Technology, Taxila, Pakistan in 2006 and 2010, respectively, and the Ph.D degree in information and communication engineering from Sejong University, South Korea in 2015.
He is currently working as a Postdoctoral/Senior Researcher at IDLab, a core research group of imec with research activities combined with Ghent University and University of Antwerp. He is and has been involved in several ongoing and finished European research projects such as eWINE, WiSHFUL, etc and national projects such as SAMURAI, IDEAL-IOT, Cognitive Wireless Networking Management etc. He is actively involved in writing research proposals for various national and international funding agencies.
He is a senior member of IEEE and actively involved in various research activities. He is also serving as an associate editor in various journals such as IEEE Access, Journal of Networks and Computer Application (JNCA). He is the author or co-author of more than 40 plus publications in well-known journals and conferences. His research interests include resource management, self-organizing networks, small cell networks, Internet of things, 5G wireless networks, machine learning, etc.
13 May 2019