Posted by Ranveer ChandraSenior Researcher, Microsoft Research
On Sunday, Microsoft Research published a new paper in partnership with The Chinese University of Hong Kong proposing a new system for improving indoor use of wireless technologies in cities.
Radio frequency spectrum, the airwaves over which wireless devices communicate, is in increasing demand throughout the world. Access to spectrum is currently regulated by strict licensing systems that limit the users and the applications of any given set of frequencies. The unfortunate result is that a significant amount of spectrum goes unused at any given time or place.
Based on measurements taken from more than 30 diverse locations in a typical city, our study found that more than 50 to 70 percent of spectrum in the TV band alone goes unused. These vacant frequencies are called TV white spaces, or just white spaces.
White spaces are the first frontier for forward-thinking regulators throughout the world, who are increasingly recognizing just how inefficient our current methods of spectrum allocation are, and moving toward new approaches. The Federal Communications Commission in the U.S. and the Finnish Communications Regulatory Authority both allow unlicensed access to white spaces (similar to regulation for Wi-Fi), and a number of other regulators are hot on their heels. In Singapore, the Infocomm Development Authority just finished a consultation on white spaces, and in the United Kingdom, Ofcom is preparing to conduct TV white spaces pilots with the aim of full roll-out by early 2014.
In order to make use of unlicensed access to white spaces, wireless devices must somehow detect what frequencies are available for use in their location using dynamic spectrum access (DSA) technologies. While spectrum sensing technology is available, it can be expensive and difficult to implement. It also has to tread a very fine line between identifying as much unused spectrum as possible and avoiding interference with other users. Therefore, most regulation on white spaces has turned to geo-location databases to direct traffic in license-exempt bands of spectrum
While they work well to limit interference and don’t pose the same cost issues as sensors, geo-location databases provide a very conservative appraisal of which frequencies are unused. Particularly in cities where the wireless environment is complex, there are often far more white spaces available for use than a geo-location database might indicate.
There is also the added complication of indoor versus outdoor use. Even though 70 percent of demand for spectrum is in indoor environments and significantly more spectrum is left unused indoors, the majority of trials and studies of white space technology have focused on outdoor applications. What’s more, the study found that white spaces are often available in bigger chunks indoors, making them easier to use for high-bandwidth communication.
In search of a more efficient method of identifying white spaces indoors, I worked with researchers from the Chinese University of Hong Kong to develop a new system called WISER (White-space Indoor Spectrum EnhanceR). WISER uses spectrum sensing technology but – unlike previous approaches – optimizes the position of a limited number of sensors. This allows us to control costs and maximize effectiveness, without losing out on accuracy.
To evaluate our new system in a real-world setting, we built and tested a prototype on one floor of a typical office building in Hong Kong. We found that WISER was able to accurately identify 30 to 50 percent more indoor white spaces than other baseline approaches.
While there are several interesting directions that could be taken in future research on WISER, this development is an important step forward toward better spectrum sensing techniques and, ultimately, greater spectrum abundance and improved wireless communication.
With a combination of technical innovation such as the development of WISER, and the regulatory progress on dynamic spectrum access that we’re seeing throughout the world, it’s a bright future for wireless technology.