AI and the Ubiquitous 1 Gbps Access Network

ASSIA CEO, John Cioffi, recently spoke at Tor Vergata University in Rome, Italy to a distinguished audience. He spoke about how data sciences, including both analytics and optimization, can contribute to our industry’s access networks evolution to ubiquitous 1 Gbps. Since a range of fixed and wireless access-connectivity options exist—and will continue to exist—throughout the world, it is important for the industry to find ways to make informed, economical decisions to evolve our access networks. Particularly, since there is no clear efficiency choice between fixed and wireless access-connectivity methods.

The key to accomplishing this informed evolution, is to make use of the network data that is available. By applying continuous artificial intelligent to this network data, planning exercises can make the informed, economical choices needed for effective access network evolution. During his presentation, John reviewed some of these methods and suggested possible gains and strategies that could be pursued to accelerate efficient progress towards the ubiquitous 1 Gbps goal.

Audience Reaction Highlights

Here we would like to highlight four topics that significantly resonated with the audience in Rome or sparked lively discussion around the evolution to a ubiquitous 1 Gbps access network.

Increasing speeds without the expense of fiber

For the foreseeable future, it is just not financially viable or practical to provide fiber to each home to achieve ubiquitous 1 Gbps. As a result, the possibility of using network data and AI to increase speeds to consumers everywhere without the expense of fiber garnered a very positive reaction.

Effective load balancing of management data

Some methods of load balancing management data bandwidth, including those employed by open source systems, have a tendency to overwhelm the management channel bandwidth. The fact that ASSIA’s methods do not overwhelm the management network, was seen as a valuable component to achieving a ubiquitous 1 Gbps access network.

State of Italian multi-operator vectoring

While not directly in John’s prepared discussion topics, the audience was very interested to hear John’s take on why the Italian multi-operator vectoring (MOV) approach has not proceeded as once hoped. From John’s perspective, the problem is that there is only one MOV equipment supplier, Huawei, which means there is low demand and volume, resulting in high costs. John asserted that a software management approach with multi-tenant capabilities could accomplish the same goal much easier for TIM S.p.A. (Telecom Italia) if they were to proceed with their vectoring effort.

Measurement beyond sampling

Many solutions that measure network speeds use a sampling approach. While sampling is better than no measurement, it can be misleading and isn’t a reliable indicator of the Quality of Experience for all consumers. The ASSIA approach, which regularly measures all users’ speeds everywhere generated genuine interest as a much-needed improvement to assure consumer Quality of Experience.

ASSIA Innovation

ASSIA is at the forefront of developing innovative technology for service providers to make the most of their existing copper and fiber access networks. John’s presentation “Artificial intelligence in planning the ubiquitous 1 Gbps access network” and the role AI can play, is just one of many innovations ASSIA is bringing to the industry. Read more about ASSIA’s broadband, fiber, and Wi-Fi management and optimization solutions.