CRC Accelerates Spectrum Science in Three, Six-week Sprints
"We need to know, in real time, where there are unused radio waves that could be put to work. Big data is the key to understanding that. It gives us the power to turn data into useful insights that allow us to predict where the surplus capacity will be at any given time. The research being conducted at the Big Data Analytics Centre has the potential to transform not only the telecommunications sector but all sectors of the economy."
With the ever growing demand for the radio spectrum – accommodating the radio waves that enable our smartphones, tablets, TVs and radios – the researchers at the Communications Research Centre (CRC) are applying emerging technologies, including predictive big data analytics and cloud-based supercomputing, to help ensure that appropriate spectrum resources are available to meet future demand. And they are amassing their knowledge base faster by using these powerful tools in six-week challenges.
In early 2018, the CRC adopted an agile approach to managing selected research projects. Small, focused teams of top experts rapidly advanced the CRC's expertise and insight in three key areas.
Challenge 1 – Layering geospatial and spectrum data to geo-reference spectrum use
The CRC's spectrum research is "fuelled" by large amounts of mobile network data. For the first challenge, researchers set out to demonstrate effective methods for layering different datasets to improve the CRC's ability to analyze and visualize multiple types of datasets that are defined geographically across Canada.
The team used a geographic information system (GIS) platform and cloud-based tools to manage, analyze and overlay many different layers of related information, including crowdsourced telecom data. This enabled the team to visually display mobile broadband spectrum information in geographic areas.
Challenge 2 – Fusing big data sources to identify trends in spectrum use
In the second challenge, researchers showed the relationships between spectrum data and societal data, including weather, traffic and current events. They first gathered local event information publicly available on the Internet, and then cleaned the data to eliminate gaps and inconsistencies. They automated this data "housekeeping" in the cloud, where they also merged and correlated the resulting data with spectrum data.
While the fusion of spectrum and societal data produced predictable results (e.g., that some radio channels will be busier during a snow storm) the process of establishing this relationship provided the most valuable lessons. It moves us a step closer to being able to quantify the increase in use during such events.
Challenge 3 – Using cloud-based supercomputing to optimize engineered surface design
Supercomputing is becoming increasingly accessible through the cloud. In the third challenge, the CRC applied cloud-based virtual supercomputers to optimize engineered surface design. Engineered surfaces are printed electronics. This means you can print conductors, semi-conductors, etc. directly onto flexible materials using standard printing processes. These engineered surfaces can be embedded into materials to extend wireless coverage or to control interference between different users. They capture radio frequency (RF) waves and redirect them based on their specified design, and there can be 1070 design combinations!
Using cloud-based parallel computing and their RF expertise, the team orchestrated tens of thousands of complex simulations to refine the design of engineered surfaces in a few days. They effectively condensed computation that would have taken years using a conventional state-of-the-art workstation.
While the CRC's focus is on spectrum management, its growing expertise in big data analytics and cloud computing is drawing the attention of other federal government departments looking to apply these tools and techniques to the challenges they face. Other government departments, academic institutions, companies and consortia are invited to contact the CRC to learn more about collaborating.
- Date modified: