Over the past several years, Network Function Virtualization (NFV) and Software Defined Networking (SDN) have been the main disruptive technologies driving innovation in the telecommunications industry. Wireless providers couldn’t improve their networks fast enough to keep up with customer demand. Upgrading from 3G to 4G was costly, and the time it took to change out network equipment was unmanageable.

The solution was NFV, where network functionality was no longer tied to the physical infrastructure. With SDN, telcos could replace purpose built computers with commodity hardware with all the powerful features now residing in the software. This white-box model, which AT&T proved just weeks ago works across disparate networks and equipment, not only significantly reduces cost, but also allows CSPs to be more nimble and reactive to changing market demands.

But CSPs need to think bigger than just NFV and SDN. The age of big data demands that providers look beyond the box to the bigger picture of software defined architecture.

Big Data Issues in Telecommunications

Big data technologies, when combined with streaming analytics and analytics at scale, are changing the way CSPs do business by uncovering significant new insights about their infrastructure and customers.

For decades, CSPs have captured information about customer calling patterns, wireless data usage, location data, network bandwidth statistics, and even apps and webpages accessed by mobile devices. Until recently, however, much of this data was discarded because of the difficulty in mining value from it and the expense of storing it.

During a time of stagnant growth in the telecom industry, big data – especially IoT – may be a way to increase revenues. Unfortunately, it’s not as easy as you might think. Here are a few things CSPs need to keep in mind.

IoT Is Harder Than You Think

IoT is becoming one of the biggest demands on bandwidth. According to IDC, 28 billion devices will be connected by 2020, which means there’s a tremendous opportunity if CSPs can only figure out how to make money off these devices.

To do so, CSPs must ensure their infrastructure supports a wide range of IoT use cases, which may require them to develop new networking and compute capabilities. For example, intelligent transportation, which ranges from telematics and diagnostics within cars to connected vehicles to autonomous vehicles, will require a change in CSPs’ infrastructure. Providers will likely need to develop computing capabilities to support low-latency applications at the edge of the network that require compute closer to the end device.

A major auto manufacturer, whose autonomous vehicle project is currently in the proof of concept stage, is generating two terabytes of data per hour per car during testing. Imagine how that’s going to affect your network.

Through 2018, Gartner estimates 75 percent of IoT projects will take twice as long as planned and virtually all will experience cost overruns. Clearly CSPs face a steep learning curve when it comes to IoT.

Who’s Watching the Machines?

The rise of IoT has brought with it an increase in machine-to-machine communication. It is estimated that almost half of all IoT connections will be machine to machine by 2020.

Machine learning is a natural extension of machines talking to one another. Machine learning can use the data being captured on connected devices and link it to streaming technology, allowing machines to act locally, but learn globally.

CSPs are in a unique position to exploit this footprint. For example, CSPs can use machine learning for real-time deep packet inspection to optimize traffic routing and steer network quality of service. They can also use it to help identify new services to offer customers and even open up new revenue streams in areas like targeted advertising.

As more machines talk to one another, CSPs need to ask themselves who’s watching the machines? Networks are capturing huge volumes of data from a wide range of sources, including subscriber profiles, utilization, geographic positioning, and performance to name just a few. Data preservation, integrity, and privacy should be big concerns for CSPs.

Whose Data Is It Anyway?

For years, CSPs have been collecting huge amounts of consumer data. Because of how they track this data on the underlying pipes, CSPs have even more information on us than Facebook or Google. They likely know more about us than our own mothers. A scary thought.

As CSPs try to find new revenue sources by mining this data, the questions becomes who owns it? Though Google and Facebook have been capturing and selling data to advertisers for years, CSPs have been held back due to regulatory controls.

The overturn by Congress of FCC privacy regulations that had been adopted last year but had not yet gone into effect is seen as carte blanche for CSPs to monetize detailed profiles of customers’ online behavior and preferences. Though this opens up new revenue streams—including personalization, advertising, targeting, and content specific recommendation engines—CSPs need to be sensitive to consumer privacy concerns or they may find themselves losing customers.

The takeaway

If CSPs want to tap into new revenue streams, they must embrace a software defined architecture and next generation technologies that provide data convergence, stream processing, and application agility.

Though white-box computing is driving down capital expenditures, new revenue streams such as IoT must be pursued if the industry wants to grow. CSPs also need to take advantage of the operational insights gained through advanced analytics and machine learning, which will also lead to increased savings.

A software defined architecture will allow CSPs become more nimble and responsive to the ever changing demands of the telecommunications market.

 

William “Bill” Peterson is Senior Director, Industry Solutions for MapR. Prior to MapR, Peterson was the Director of Product and Solutions Marketing for CenturyLink, and before that he ran Product and Solutions Marketing for NetApp’s Analytics and Hadoop solutions. In addition to his marketing role at NetApp, Peterson was the Marketing Co-Chair for the Analytics and Big Data committee, SNIA. He has also served as a research analyst at IDC and The Hurwitz Group, covering the operating environments, content management and business intelligence markets.