A platform-based approach turbo-charged Reliance Jio’s move to cloud native and 5G

Who: Reliance Jio and Jio Platforms

What: Transitioning its vast and complex 4G network to cloud native and 5G

How: Jio Platforms designed, built and deployed new tech on cloud native platforms that is backwardly compatible with all BSS/OSS functions

Results: Network management reduces annual maintenance costs by 90%, provisioning more than 5,000 sites with zero-touch; fulfilment management delivers an estimated 50% rise in operational efficiency

The Indian mobile network operator Reliance Jio faced unique challenges in transitioning a 4G network of tightly integrated physical network functions to a cloud-native infrastructure. While it was paramount to encourage migration to container-based applications, it was also important to make it as pain free as possible for the operations team which manages such a large network.

Jio Platforms designed and deployed new technology capabilities using cloud native platforms built from scratch to ensure backward compatibility with all OSS/BSS functions.

The new platform required many key capabilities:

• A user interface to create adaptors on the fly – writing custom adaptors builds significant lead times into systems integration. The new OSS needs a user interface for modeling specific data about network functions, and to allocate compute engines dynamically.
• A set of APIs for northbound systems to enable use case integrations – the new OSS must expose APIs that enable integrations with the Jio BSS and customer-, employee- and partner-facing platforms.
• A single platform personality – to support multiple domains for activating services (mobility, fixed, enterprise and IoT).
• Integrations with multi-domain orchestration – the OSS should have a unified software-defined network (SDN) orchestrator to automate transport, IP and microwave networks. The orchestrator will abstract domain-specific SDN controllers to smooth implementation.
• Multi-site and distributed systems – the platform must provide deployment options across sites (for geographic redundancy) as well as for multi-access edge computing (MEC) assurance to accommodate 5G MEC deployments and enterprise 5G networks.
• Microservices and containers – the cloud network functions’ (CNFs’) lifecycle management has to be integrated with the existing Management and Network Orchestration (MANO) platform from the beginning. Being cloud native, it must also support Microsoft Azure and Google Cloud Platform from the very start too.
• Rich asset modeling – as well as the current inventory management, complex inventory constructs such as fiber routes, logical IP/MPLS links and other associations need to be modelled through the platform.
• Configurable co-relation policies – apart from fault, configuration and performance management, the new platform should support auto-correlation as probe data, alarms, counters, configurations and call detail records will be configurable by the operations team.
• Inline and faster upgrades – necessary for all new and existing network functions through the various protocols and integration capabilities required in a heterogeneous network.

The cloud native approach to NFV

The cloud native approach is transformative: It will support automated lifecycle management of network functions, instant deployment of 5G core network slices at any scale, configuration management, portability across private and public cloud stacks, and DevOps-style continuous innovation. Together they bring stunning improvements to the efficiency of network operations and customer-facing responsiveness. Jio Platforms designed and developed assurance and fulfilment platforms to provide operations support for an existing network, a network in transition and transformed 5G infrastructure. The platforms act as a backbone for business support systems (BSS) to enable Agile operations. The platforms’ designs are aligned with TM Forum’s Open Digital Architecture (ODA) The key capabilities of a modern OSS

The diagram below shows that southbound the platforms integrate with network functions deployed in hybrid infrastructure for direct provisioning and fault, configuration, accounting, performance and security management. Northbound, the platforms provide standard interfaces to integrate with the management and orchestration layer, and are integrated with the existing AI and machine learning platforms that provide insights into network data.

Cloud native network management

The Jio Network Management System (Jio NMS) facilitates the operation and management of network nodes. It provides a foundation on which to implement Operations Support System (OSS) architectures. The Jio NMS is fault tolerant by design, so stays operational 99.999% of the time.

It serves more than 110 different types of 4G and 5G nodes, more than 5,000 sites are provisioned with zero touch, and it delivers a 90% reduction in annual maintenance costs. All the 5G Nodes are pre-integrated and more than 80 4G nodes are integrated in production network, which accelerates deployment.

Jio MANO

Jio Management and Orchestration Platform (Jio MANO) manages network functions and the lifecycles of network slices. This Jio MANO platform is a microservices-based, distributed, scalable and extensible framework for rapid integrations and service implementation. It speeds time to market for new services and capabilities while supporting scalable operations that reduce IT’s complexity and costs.

The platform can scale resources on demand, support services across distributed cloud – also known as edge cloudlets – for fixed and mobile users. It monitors and takes intelligent action to manage and orchestrate its distributed resources. It supports virtualized and containerized workloads without needing open source solutions like Kubernetes.

Jio Platforms has deployed 500 cloud native pods for Reliance Jio’s 5G core network and deployed CPE capacity at 100,000 sites across four zones in India. It has also integrated more than 100 live, multi-cloud, virtualized infrastructure manager sites in the network.

Managing fulfiment

The Jio OSS Fulfillment Management System (FMS) orchestrates workflow, and is a provisioning and activation platform designed to simplify operations by replacing high-touch management with a programmable, virtualized model. All Reliance Jio’s 5G nodes are pre-integrated and more than 80 4G nodes are integrated in production for provisioning and configuration which accelerate deployment at less cost.

Reliance Jio has also seen an estimated 50% improvement in operational efficiency.

The cloud native solutions are built on the top of TM Forum’s Open APIs, which are a strong starting point for designing microservices because they define a consistent and simple-to-use resource model. The solutions also use the Forum’s Open APIs for product inventory management, service activation and configuration, service inventory and resource inventory management.

It also used business entities from the Forum’s Business Process Framework (eTOM) as well as aggregate business entities contained in the Information Framework (SID).