Mariya Zheleva, assistant professor in the department of computer science, explores this question.
Dr. Mariya Zheleva is an assistant professor of computer science in the College of Engineering and Applied Sciences at University at Albany, SUNY. Prior to joining the tenure-track faculty in 2016, she was a visiting assistant professor at the University. Mariya’s research interest is in the intersection of wireless networks and Information and Communication Technology for Development. She has done work on small local cellular networks, Dynamic Spectrum Access, spectrum management and sensing and network performance and characterization.
Did you know that the radio spectrum is a precious natural resource, which helps us get electronically connected through our mobile and wireless networks? We increasingly hear that spectrum supply is short, driving up the prices for wireless services. Unlike coal or oil, which when exhausted are really hard to renew, the radio spectrum is instantly renewable. As soon as one device is done using it another device can jump right on. Like time, the spectrum is wasted every moment it is not used.
However, our governance regards spectrum as a finite resource, exclusively allocating radio frequencies to technologies and operators. This creates artificial spectrum scarcity, whereby some bands, such as terrestrial television, are rarely used, while others are overloaded. Thus, we are not really in a shortage of radio spectrum, we simply need to stop thinking about it as a finite resource and begin using it more effectively. This requires our devices to become much smarter in how they pick when and where in the spectrum to operate. They need to constantly measure, characterize and make informed, autonomous decisions. This is hard because it requires extra hardware, power, and computation, which has to be introduced in a way that does not interfere with users’ regular interaction with mobile networks.
Our research at the UbiNET Lab designs new signal features and algorithms that enable lightweight sensing and characterization of spectrum use in the vicinity of a device. We discovered signal representations that help us mine transmitter’s behavior and technology even from weak signals. We feed these features in algorithms to support devices’ decision making, improve wireless security and inform next-generation spectrum policy.