5G Analytics: A Tale of Modular Internet Architectures and Proprietary Telco Systems

Subscribe To Download This Insight

4Q 2019 | IN-5611

Impending 5G networks are expected to enable new services and pave the way for personalized user configuration and new growth from industrial and enterprise verticals. A key difference between 5G and previous network generations lies in its need to tightly integrate applications at the service layer with structured/unstructured data repositories. This is expected to deliver significant value for Mobile Service Providers (MSPs) in building data-driven services, but first they ought to rethink how they run their operations and how to modify their data strategies accordingly. 5G is inherently a decentralized approach that is expected to generate new and vast volumes of data from heterogeneous and multiple vendor equipment. Ericsson’s newly launched solution Edge Gravity, for example, is a commercial offering that offers a global and decentralized 5G edge cloud platform. The organizational form and data models that would best commercialize 5G edge offerings need to factor in a proximate pool of connected data (and metadata).

Registered users can unlock up to five pieces of premium content each month.

Log in or register to unlock this Insight.

 

5G Set to Create Vast Data Volumes

NEWS


Impending 5G networks are expected to enable new services and pave the way for personalized user configuration and new growth from industrial and enterprise verticals. A key difference between 5G and previous network generations lies in its need to tightly integrate applications at the service layer with structured/unstructured data repositories. This is expected to deliver significant value for Mobile Service Providers (MSPs) in building data-driven services, but first they ought to rethink how they run their operations and how to modify their data strategies accordingly. 5G is inherently a decentralized approach that is expected to generate new and vast volumes of data from heterogeneous and multiple vendor equipment. Ericsson’s newly launched solution Edge Gravity, for example, is a commercial offering that offers a global and decentralized 5G edge cloud platform. The organizational form and data models that would best commercialize 5G edge offerings need to factor in a proximate pool of connected data (and metadata).

The expected increase of data volumes, and the utility on offer, bodes well for MSPs transitioning and refashioning their operations from app-centric to a data-centric approaches. A case in point is SK Telecom’s T Advanced Next Generation Operational Supporting System (TANGO), an Artificial Intelligence (AI)-assisted big data platform that detects network (both wireline and mobile) anomalies and optimizes the performance and quality of network operations. Going forward, telco operations will need to cater not only to existing revenue sources but also to new growth, such as industry verticals. Some use cases require the vast data volumes expected to spring into existence. For example, the volume of data is critical for timely assurance and performance for network engineers, but there are use cases that need to leverage the power of smaller subsets of data to produce meaningful results (e.g., voice service quality).

A Proliferation of APIs, Business Processes, and Interfaces

IMPACT


5G has been designed from the outset to take advantage of virtualization and cloud computing technologies. The net result of this is that physical network elements turn into software-based equivalents that are more flexible, more scalable, and adhere to requirements in near real-time. This highlights a difference in operational control and behavioral insight between the current generation of telco data repositories and a 5G data layer; specifically, how to seamlessly tie the business logic at the service layer with data repositories that come from heterogeneous network deployments (i.e., 3G, 4G, and 5G). Furthermore, the 3rd Generation Partnership Project (3GPP)’s 5G core specification focuses on services that are invoked using Application Programming Interface (API) system calls. APIs bring about a degree of simplicity and service-oriented innovation and are best understood relative to protocols that must be adapted, re-engineered, and configured. The development of a hybrid data layer that bridges both remains a challenge, however, so vendors including Ericsson and Openwave Mobility (an Enea company) have already launched solutions that aim to address that gap.

5G natively introduces increased digitization, in turn creating more business processes and interfaces. The challenge for the industry is driving a unified data strategy; in other words, having a diverse set of processes tap into a single repository for service provisioning via standard interfaces. Existing telco processes are certain to be refashioned, but process re-engineering remains a challenging endeavor because it involves a wholesale review of existing operations. This is particularly relevant for 5G as a technology that promises a high-growth ecosystem predicated on not only telco network reliability, but also Internet scalability and agility. Further growth is expected to come from a multi-dimensional 5G ecosystem supported by multi-supplier partnerships and multi-vendor environments where analytics, big data, AI, and Machine Learning (ML) become inextricably intertwined. To that end, Nokia AVA, Accedian Skylight, and Cloudera Data Platform are some cloud-native solutions that provide AI and data management capabilities that extend to edge, multi-cloud, and on-premises deployments.

Embrace a 5G-Ready, Data-Centric Model

RECOMMENDATIONS


Many MSPs are already on a journey to unlock value from analytics and their data lakes. AT&T, Telefonica, and Vodafone are some MSPs, among many others, that have an internal data strategy in place based on network, customers, and Internet of Things (IoT) data sources to enable cost efficiencies and drive innovation. The shift to a data-centric model on a wholesale basis is certainly at the forefront of their transformation agenda and successful commercialization of industrial 5G. The question that vendors should use to frame their strategy for 5G analytics solutions is how their solution is engineered for cloud environments, rather than whether or not they have a solution readily available. Existing telco operations have stricter Service Level Agreements (SLAs) than cloud providers and Software-as-a Service (SaaS) platforms, and service uptime guarantees remain stringent. Heavyweight telecom vendors, therefore, must reach a middle ground where they promote a data-centric model predicated on two strands: effectively utilizing cloud-native agility and resource efficiency and meeting specialized telco requirements and performance isolation.

Telco networks store data in isolated repositories with little (or no) integration between distinct applications. Going forward, the industry should view the convergence of Information Technology (IT) and communications (e.g., 5G) as a harbinger of a new growth. Such convergence calls for a bold collective action to capitalize on this incipient industry change that amalgamates modular Internet architectures with proprietary telco systems. Internet players such as Google, Amazon, and Alibaba are driven by commercial ends that aim to realize cost savings, business agility, and innovation. MSPs, in contrast, are driven by standards and remain anchored to an engineering-centric approach. One challenge of working in such a hybrid environment is taming the “wild” and “dark” data at MSPs’ disposal. ABI Research believes that MSPs should embrace a 5G data-centric model that controls and unifies the underlying data repositories often locked in siloed applications. Further, the industry at large should aim to set a cost so that estimates can be made regarding required data quantity.