MATHIEU BRIGOT NETWORK PROTECTION
If we want to improve the quality of service of our networks, we must review our risk analysis and put in place simple, inexpensive controls.
after a customer connection goes down.
helps technicians act more quickly and decisively when on site. Also, being able to access a POP or cabinet using a mobile phone or code saves a lot of key management. Today’s growing proliferation of networks means that when a fibre link is broken, it can increasingly be backed up temporarily by a wireless or satellite network, albeit with inferior performance. However, a fibre issue can bring down radio and 5G networks, or alarm systems. Numerous converging networks, from building infrastructure and traffic management systems to cloud and fintech services can’t be replaced with wireless or satellite. So protecting network infrastructure is more important than ever and requires a combination of approaches: smart processes and monitoring, introducing robust, connected housing, and avoiding human error wherever possible.
means it take longer to return to the initial messy state (3 to 6 months), but reverting to ‘spaghetti’ seems to be unavoidable. As soon as one patch cord is misplaced, others follow. A ‘bad’ patch cord prevents subsequent technicians from positioning their own patch cord correctly, but also makes cleaning up unused patch cord difficult - by removing certain patch cord, technicians inadvertently disconnect others. LACK OF REDUNDANCY Operators put risk control or mitigation systems in place. Optical loops in the Central Office prevent complete branches of the network from being cut off by simple events such civil works errors. In the rest of the network, however, this type of redundancy is lacking. When it comes to connection points on poles or in fibre cabinets, there is hardly ever any redundancy. Essentially, sites that link huge numbers of subscribers may be adequately protected, but locations affecting only a few people are hardly considered. Surprising, as errors in the central office very rarely occur whereas errors in small POPs (Points of Presence), such as a shelter or a street cabinet, occur every day. Once everything is in place in the central office, changes are practically never made, whereas changes to connection points are made very regularly. We also regularly come across see outdoor equipment which is functional but in a deplorable state. An early warning system would mean this could be brought up to standard without major expense, before subscribers disconnect. If we want to improve the quality of service of our networks, we must review our risk analysis and put in place simple, inexpensive controls. These need to be systematic in order to avoid drifts due to the frequency of interventions. On the fibre network, existing systems can be asked to test all network connections after each intervention - and not just
We can also set up systems for monitoring the condition of cabinets and install intervention traceability systems to ensure each person has a sense of responsibility and will think about the consequences of a bad installation. Above all, we can make interventions visible to everyone by communicating more and facilitating dialogue between technicians in the field. Employees could share their performance and best practices and challenge each other.
A NEW LOOK AT RISK CLASSIFICATION
A major problem at central office level will always be serious and tens of thousands of subscribers might be cut off, including institutions such as hospitals and businesses such as banks. Whereas an error at a small POP or a street cabinet might mean as few as a dozen people are inconvenienced. However, if we classify the problems in terms of criticality, considering the frequency with they occur and the probability they won’t be detected in time, this classification could be greatly changed. For example: cutting off a bank’s communication for one day every five years might actually be less of a problem than cutting off people working at home 15 times a day for three days. These numbers are not made up – they are findings from our experiments with operators using network security systems. Without robust, durable, tamper-proof POP housing, it would be impossible to maintain broadband services. Intrusion alarms and access control systems, with links to supervision centres, prevents tampering. Built-in monitoring is also essential. For operators, knowing whether someone has tried to break in or if a cabinet door simply hasn’t been closed properly, can save a huge amount of time and effort. Furthermore, having access to an up-to-date maintenance and intervention history
Mathieu Brigot - Product & Solution Manager, Nexans is an engineer in Mechanical Systems Engineering and Innovation Project Management. He has been with Nexans for three years, after previously working in installation, managing installation of copper and fibre networks in Paris. He specialised in indoor components
for FTTH networks and was involved in finding IoT-based services and usages.
www.opticalconnectionsnews.com
19
ISSUE 28 | Q1 2022
Made with FlippingBook - professional solution for displaying marketing and sales documents online