Insights

Editorial

Is OSS Going to Die? 

Contributed by Marcin Nowak, Senior Solutions Manager, Comarch.

Is OSS Going to Die? 

As telecommunications networks evolve, we can see hardware providers deploy ever more advanced software. But that’s only one aspect of this evolution. Another is that an increasing amount of specialized, very sophisticated telecommunication hardware is becoming cloud-native. With the new network models, we no longer have dedicated, vendor-dependent islands. Rather, we have common hardware that’s able to host anything needed at any given moment in the network.

So, in this modern telco world, is there still room for OSS?

Network maintenance tradition

Having worked as an OSS engineer for over 10 years, I may be a bit biased, but with this baggage of experience that I carry with me, I will try to answer this tricky question.

Traditionally – and historically – OSS is a system designed to maintain the network globally or nationwide. Such a network has always been a combination of multiple technologies and various suppliers. Maintaining the network was not only limited to equipment and connections, but also to the provisioning of services to the customers. However, in today’s rapidly changing telecommunications world, the functions offered by human-controlled OSS systems are not enough.

Challenges of the future

What we can see happening these days is a substantial increase in the speed of networks and operations. Anything that happens in the network today needs to happen quickly. We must deliver new services to the customers immediately, but we also need to detect, identify and solve problems affecting our network within seconds to minimize their impact on customers.

Does traditional OSS fit into this picture? Is there a place for traditional OSS in terms of human-managed NOC, manual service provisioning and recovery, or lengthy network discovery processes? I bet not! But is it still needed? Certainly.

A well-played network orchestra

A modern network has to be agile and fast. Decisions have to be smart and made without hesitation. This can be assured through smooth cooperation on all network levels.

The perfect comparison of what we want to achieve is an orchestra. Each individual musician is an expert in their field, with their skills excelling in their particular domain. But the true excellence is that which underpins the duets, quartets, and full orchestra. In order for this musical perfection to materialize, though, a conductor is needed. Their role is vital to success. This is exactly why we use the word “orchestrator”… and there is no coincidence in the term’s origins.

The network, like an orchestra, also needs an entity that will be able to give all elements the right momentum and coordinate all activities. Such a top-level unit is exactly where modern OSS needs to be. What is needed to get to this point? A high degree of automation is certainly required, and this is even better when supported by AI/ML. And why is OSS a good place to locate AI capabilities in our network? Because it contains information about the entire network, no matter which domain or vendor we are talking about. With this information, it is possible to run a full RCA, as well as solve problems in an unprecedented way, bypassing domain boundaries and limitations of network segments provided by the vendor. As a good conductor, the OSS also needs to be able to smoothly delegate tasks and control results. This can be achieved through a high degree of network API standardization. With the leading roles of ETSI, TMF, 3GPP, GSMA and many others, we can observe this in today’s market, talking about 5G SA system architectures.

Conclusion

It seems as though OSS is not going to die – at least not yet. As the network evolves towards higher speeds, automation and autonomy, software evolution must follow. There is a distinctive role reserved for the OSS as well.

We all need an orchestrator to oversee all the tiny blocks working in our complex network environments. Today’s challenge is to get to the podium on which the conductor should stand.

Modern OSS must contain a standardized integration layer to be able to communicate with all musicians and must collect the feedback from their orchestra instantly through real-time network discovery. In the event of any problems while playing a network concert, the OSS has to be able to act wisely. This can be achieved only with a solution based on modern, scalable architecture, with AI capabilities integrating all domains from TMF TAM into one solid solution.

 

Marcin Nowak is a keynote speaker at the FutureNet World virtual event on the 20/21 April – register here