When it comes to the transition to Level 4 automation, it would be nice to have the luxury of building a network from scratch – for example, a 5G private network that is isolated from any other infrastructure.
In that case, it may well be a relatively easy task to reach L4 and L5 (according to the TM Forum automation scale) and fully autonomous operations. Unfortunately, that is a luxury afforded to very few.
In reality, today’s networks are a complex hybrid of multiple generations of fixed and mobile technologies – 2G, 3G, 4G, and 5G, and so on – over a mix of copper, DSL and fibre infrastructure, with various iterations of core network and RAN technologies all built up over years.
The OSS challenge is delaying the automation journey
In addition, we have the OSS mess, with dozens (and in some cases hundreds) of legacy OSS systems supporting disparate or dedicated services and capabilities. These OSS systems continue to support live services and operational processes to existing customers and will do for some time to come.
This challenge is well understood in the industry, but retiring legacy solutions is not always a realistic option due to the cost and disruption it would cause and their highly specialised nature.
While it has been relatively easy to transform access, transport, and core networks, and move to new platforms and generations of technology to deliver innovative new services, the operational systems that support them have not kept pace with this transformation. These cannot simply be retired.
While it’s true that, given sufficient time, such OSS systems will be gracefully retired and replaced, this mix we face today – allied to the growing complexity of today’s networks – creates friction for automation projects.
It’s not just distinct tasks, and network domains that need to be addressed through your automation strategy, but also the array of unique OSS, billing, customer care systems, and so on, built up over many years that must be included in order to develop cross-domain automation, and to deliver highly autonomous networks that reach L4 and L5 targets.
At the same time, every operator has a unique environment and is in a different place on the automation journey. This means it’s tempting to focus on newer domains, such as transport, core network, RAN, service optimisation, and so on – which may be provided by a single vendor and offer an easy gain.
Ignoring legacy systems in the automation journey will create ‘chasms’
However, ignoring legacy systems in the automation journey will lead to a chasm and delay the journey to fully autonomous networks. It’s one thing to move from one generation of optical transport to another, quite another to replace the multiple silos of, for example, billing platforms that are tied to individual services.
But it does not need to be such a challenge. What is required is the integration of OSS legacy systems with automation solutions that are designed to support the vast array of legacy interfaces to OSS platforms, as well as newer interfaces to ensure that any future platforms, systems, or technologies can be easily brought into cross-domain automations.
This is Stage One of the We Are CORTEX framework. We create functional, reusable building blocks that connect to interfaces to any legacy or next-generation OSS system. These functional blocks can then be interconnected to create broader automated processes.
Many of today’s automation engines support only newer interfaces, such as REST and SOAP, but few support legacy interface. CORTEX on the other hand supports all legacy interfaces, including SFTP/FTP, SSH/Telnet, SNMP, TCP Sockets, IMAP, CORBA, and many more. Essentially, the CORTEX platform is agnostic to legacy interfaces and also enables decoupling from vendor-specific solutions.
It’s clear that transforming and automating the OSS will involve many pathways and parallel processes and will be a complex task.
The CORTEX platform can both automate legacy OSS solutions — with interfaces that cannot be understood by most automation engines — and integrate them with wider automation processes, enabling hyperautomation.
To find out more, download our recent paper here.
