AT&T today unveiled its next-generation data-sharing network, building on software-defined networking, its ECOMP orchestration platform and newer technologies such as blockchain to create a secure and trusted environment in which data from multiple parties can fuel new applications and services. (See AT&T Launches Data Sharing Platform Indigo.)

The AT&T Network 3.0 Indigo platform is intended to transform the current generation of services in the way that the World Wide Web and browsers enabled data networks to become the Internet as we know it today, Victor Nilson, senior vice president of big data at AT&T Inc. (NYSE: T), tells Light Reading in an interview. That is being done by enabling a trusted sharing of data that today is siloed by companies for privacy and other reasons, unleashing the power of big data and analytics in a new way across a broader landscape.

In a blog you can read here, Nilson cites examples of things Indigo can enable including simple environments which combine repair services with city traffic management and utility companies to more complex telemedicine environments that let doctors, hospitals, pharmacies and insurance companies share patient data in secure and appropriate ways.

Most of what Indigo does is knit together existing technologies such as SDN, identity management, access and authentication management, blockchain for auditing as well as machine learning and artificial intelligence. It creates "a manageable information sharing environment that, on the one provides the protection, privacy, security, compliance management of appropriate data usage while at the same time unlocking some level of data that can be managed and mined across entities, within a company or across companies," Nilson says.

That doesn't necessarily mean raw data is shared -- the data may be anonymized, aggregated or presented in summary form -- but it can still be used as needed in a safe way, he says.

"This unlocks the benefits of big data across entities that might not be able to share their raw data but they are wanting to get the value of the combined data," Nilson comments. "That has been an aspirational goal for a while, with technology pieces that are out there in siloes, but there's really not an integrated mechanism, so that is fundamentally a lot of what we are trying to tackle with Indigo."

The platform has been in development for about two years and is not yet commercialized, Nilson adds. There will be more detailed white papers released over the next 30 days, and some specific use cases will be coming out later this year.

AT&T expects to put this to use for its own network services and for its enterprise customers but the company also is expecting some general use by other network operators. "We think this is going to impact the entire industry," Nilson says -- thus the Network 3.0 label.

Among the immediate problems Indigo can address is enabling Internet of Things applications that use end-point devices such as sensors that are typically low-cost and not heavily secured, he points out. By combining a virtualized network environment with a range of security approaches -- authentication, analytics and more -- Indigo would enable a more intelligent approach to security.

Today, most security is based on created "trusted entities" inside protected walls of security, but most breaches occur because such an entity is compromised or there is someone pretending to be a trusted insider, Nilson notes. Indigo would enable movement away from the binary security standard of trusted/not trusted to a nuanced approach based on reputation.

Access to data would be based on specific applications and entities involved and even trusted entities could be suspect if they suddenly pop up in unexpected places or behave in unexpected ways. In those cases, virtual partitions can be established to limit access and also limit risk, he notes.

One other aspect of Indigo that is important is its ability to scale, according to AT&T. The goal is enabling an explosion of new capabilities across the network, Nilson says.

— Carol Wilson, Editor-at-Large, Light Reading