NTT distributes a variety of data across business and industrial boundaries, which contributes to solving social issues and creating new value. In order to use such a massive amount of data efficiently and accurately, society as a whole has begun to actively engage in Digital Transformation (DX).

Here we will introduce three network technologies that NTT Information Network Laboratory Group is conducting R&D on to further accelerate digital transformation (DX).

Three network technologies required to accelerate digital transformation (DX)

In NTT network service for businesses that operate social systems such as transportation, agriculture and fisheries, medical care, disaster prevention, manufacturing, tourism, entertainment, it is necessary for us to link various services such as ICT through multi-orchestrators and APIs, and provide the optimal configuration according to the requirements in order to meet the demands of businesses. In addition, it is necessary to not simply provide services, but also ensure a fast and stable supply.
To do so, the following three technologies required for network services in order to accelerate digital transformation (DX):

• Network configuration technology that can be used to flexibly build advanced functions according to each use scenario
• Proactive control technology that can analyze situations from various data obtained in the course of network operation, and respond autonomously　to future demand and failures
• Operational technology for flexible and efficient operation of nationwide communications infrastructure according to the situation

Initiatives for flexible network configuration technology

NTT Information　Network Laboratory Group has developed high-speed, large-capacity networks with advanced connectivity. In addition, it has made network functions as components usable for flexible network configuration, and has created mechanisms for centrally controlling those functions.
This technology separates the parts that require hardware processing from the parts controlled by software, and replaces functions as required.
This technology was used to develop FASA (Flexible Access System Architecture), which converts part of optical access equipment into software, so it can be changed to various applications including the next-generation communication standard “5G”.

Example: FASA (Flexible Access System Architecture)

FASA can convert the discreet functions of bandwidth control, multicast processing, maintenance into software, and provide them to users who need only one function or a combination thereof.
This makes it possible to provide various services quickly.
In addition, by using general-purpose hardware for a long period of time, FASA can reduce capital investment, improve service continuity, and reduce operating costs through the standardization of maintenance operations.
If this technology is used to split optical lines, it will not only be possible to deploy 5G base stations at a lower cost, but use it for various other applications as well, such as controlling robots in factories.

For APIs that realize centralized control, we are working on standardization for common use with various partners, including carriers, system vendors, standardization organizations, and the OSS community.
In particular, we are working to implement various reference models with the aim of building a common API framework, and are providing easy-to-use open APIs that enable rapid deployment of ICT resources and optimization of configurations.

Initiatives in proactive control technology capable of predicting the future and functioning autonomously

NTT Information Network Laboratory Group places great emphasis on its work in the conversion to proactive control type networks, which function to proactively detect potential risks (congestion, failures, etc.) and changes in demand, and take proactive measures in advance. Proactive control technology is the technology required for this conversion.
With the development of AI technology, improvements are being made to the accuracy of the auto-scaling function, which automatically adds resources based on demand forecasts, as well as the auto-healing function, which performs route control in response to failure prediction information and sudden failures. Proactive control networks avoid risk in advance reduce the burden on personnel by autonomously responding to future events via AI.

Case: Demonstration experiment at NTT operating companies

NTT is steadily advancing initiatives to transition to a proactive control-type network, such as establishing anomaly detection technology using deep learning, as well as technology to autonomously analyze the causal relationship between device alarms and network failure causes, and conducting demonstration experiments at NTT operating companies.
Series of operations are divided into the following phases to use AI: 1) monitoring (detection), 2) analysis, 3) control and restoration of various factors. AI-related technology required in each phase varies, but it is possible to significantly improve learning accuracy by fully exploiting the massive data resources held by telecommunications carriers.

Example: Improvement to and labor-saving in network maintenance operations

In worksites of maintenance and operation, necessary judgments and controls come down to the skills of maintenance personnel. Experienced personnel and maintenance personnel can instantly analyze and judge situations in the event of a breakdown, and take appropriate action, but it is far from easy to find employees with extensive experience. Great stress is placed on experienced employees at worksites that require 24-hour, 365-day breakdown support, and are more likely to end up in circumstances with the risk of relying on certain employees.
The work flow of maintenance personnel in maintenance and operation can be broken down into three elements: information collection, analysis and judgment, and control.
In the first step, information collection, AI analyzes user reports and alarms on behalf of maintenance personnel, then manages and aggregates incidents and countermeasures.
In the second step of analysis and judgment, personnel compare information continuously aggregated on a regular basis in the information collection step against service standards for guaranteeing service quality, and create a foundation for analysis and judgment by system-linked AI.
Furthermore, by making the AI learn the subdivided judgment process of maintenance personnel, it is able to construct an autonomous control loop group for dealing with multiple events.
With the progress of AI-based learning and the advancement of autonomous control loops, multiple tasks that were previously handled by maintenance personnel can be automated sequentially. As more progress is made, AI will be able to implement priority judgment and automatic control based on that judgment, using information matched to service standards.
Ultimately, NTT Information Network Laboratory Group aims to build autonomous control loops that enable zero-touch operation.

Operation technology that flexibly and efficiently operates nationwide communication infrastructure in accordance with circumstances

With the increase of natural disasters in recent years, networks must be more resilient and flexible than ever before.
NTT Information Network Laboratory Group is constantly researching and developing operational technologies to that end.

Case Study: Disaster response subscriber wireless system (TZ-403D)

The “Disaster Countermeasure Subscriber Wireless System (TZ-403D)” is a system that connects NTT buildings and evacuation centers to provide services such as special public telephones at evacuation centers. Using this system, it is possible to do non-line-of-sight communication between parties tens of kilometers apart. In addition, multiple terminal stations installed in evacuation centers can be accommodated in one base station, making it possible to open not only temporary telephone lines but also Internet connections and special Wi-Fi access points.
During the “Heavy Rains of July 2018 (Western Japan Heavy Rains)”, base stations were set up at the NTT building in Soka City, Okayama Prefecture, while terminal stations were installed at two evacuation shelters in the Mabi District of Kurashiki City, Okayama Prefecture, which was severely damaged by flooding. In this case study, the system overcame the problem of the mountain between the NTT building and the evacuation center, enabling uninterrupted communication with both evacuation centers for more than a month.

Realizing a new social infrastructure that accelerates digital transformation (DX)

NTT Information Network Laboratory Group will work on research and development of networks with optimal functions, such as rapid ICT deployment and configuration optimization. We will continue to accelerate digital transformation (DX) in order to realize a new era of social infrastructure with networks.