Using Predictive Analytics for Wireless Infrastructure
5G and other innovations are coming quickly — and amid all the change, no one seems to know for certain if the existing wireless infrastructures will be able to handle future systems.
It’s clear that demands on our mobile infrastructures are rapidly intensifying, and carriers are investing in more cell sites and related technologies. But so far, nobody has developed enough ways to measure all the cells in the market, analyze them, and derive predictive insights unique to that particular market — until now.
Comparing Cheyenne and Sacramento
Aurora Insight’s proprietary sensors can extract highly detailed information about wireless networks at the cell site level. We put it to work in a study of real-world markets in Cheyenne, Wyoming and Sacramento, California.
The study compared the Cheyenne and Sacramento markets’ carrier, band, and technology information across dozens of individual sites. Aside from providing rich, granular data, the exercise also provided outstanding contextual insights about buildout, geospatial considerations, and difference across markets. These insights had never been available before.
Aside from providing rich, granular data, the exercise also provided outstanding contextual insights about buildout, geospatial considerations, and difference across markets. These insights had never been available before.
In the 60 km2 area measured in Sacramento, T-Mobile has deployed five times more LTE cells than in the 100 km2 area around Cheyenne.
In Sacramento, T-Mobile’s cell deployment is more dense downtown, while fewer cells are deployed in more residential areas of the city.
We noticed some sites farther from the city center with higher MHz-per-capita were either located in/near neighborhoods, or along major roads and highways.
As we continue to collect data in these markets (and dozens of others), we’ll have a geospatial timeline of network buildout — an incredibly powerful tool for carriers, investors, developers, and cities, especially during times of such aggressive innovation.
Predictive analytics to steer future investments
Cheyenne and Sacramento are drops in the bucket as far as predictive market analysis goes. While there’s much more data to collect and more to learn, we already know we can answer the following questions:
- How does the location of cell sites relate to their competitive value?
- Can we compare deployment strategies from other operators?
- Can we predict future cell deployments?
- How does a 5G deployment strategy differ from that of an LTE network?
- How much infrastructure do we expect to see in other areas (developed markets or developing markets) based on population?
These answers will steer investments and innovations in markets across the globe, and it will be fascinating to track the relationship between cell deployment and population density, teasing out new trends and insights as we go.