Ericsson has launched a number of new Internet of Things (IoT) network services, which include voiceover support and machine learning.
The company said it is complementing its cellular IoT software and IoT Accelerator “with a complete set of network services”. These services are intended to enable service providers to address the deployment and operation of the massive number of IoT devices being introduced to LTE networks.
Ericsson said the new technologies will be applicable for Cat-M1 and Narrow Band IoT (NB-IoT). The new services on offer will include IoT network design and optimisation, deployment, operation and management, and are supported by the recently expanded Support Services offering.
Peter Laurin, head of business area managed services at Ericsson, said: “We anticipate IoT devices will surpass mobile phones as the largest category of connected devices as early as 2018 and, according to Ericsson’s latest Mobility Report, there will be 18 billion connected IoT devices in 2022.
“This massive uptake requires a different approach to network planning, design, operations and capabilities than traditional mobile broadband networks.”
Ericsson said it is also introducing new IoT software features, including Voice over LTE (VoLTE) support for Cat-M1. According to the company, this will enable operators to explore new use cases where it might be beneficial for IoT devices to support voice services.
This could open up opportunities to expand enterprise services to areas such as security-alarm panels, remote first-aid kits, wearables, digital locks, disposable security garments, as well as other types of IoT-enabled applications and services, Ericsson said.
Addressing the need for an adapted approach to management and operation of operators’ networks, Ericsson said it is also introducing automated machine learning to its network operations centres.
The company said this will help operators to “manage delivery cost and take a proactive approach to event and incident management”. It was found in a trial that 80 per cent of all incidents were identified by machine learning without human intervention, with the root cause correctly identified in 77 per cent of cases.