Highly effective optical networks don’t just happen over night. There are a number of habits that help network engineers stay one step ahead.
1. Plan and build with flexibility in mind
Anticipating future needs of a network can be challenging, but a network architect can plan for growth and incorporate flexibility into the build in a number of ways. The ability to support multiple OEM platforms with a common optical infrastructure allows the network engineer to design the layer 1 infrastructure, and then have the flexibility to focus independently on the higher-level features and revenue-generating functionality. For example, if the current requirements seem to call for a CWDM-based architecture, explore if a DWDM (tunable or fixed) solution makes sense to open the network up for future expansion.
2. Troubleshoot from the ground up
When troubleshooting network issues/failures, an effective strategy is to attack the problem from the\ standpoint of the 7-Layer OSI model, starting at layer 1 (physical layer, i.e. optics and copper) and working your way up the OSI mode. Normally, issues at the physical layer will manifest themselves in the layers above, so ensuring that optics are clean, coded to the correct OEM platform, and meet the physical interface requirements is the first step in debugging.
3. Encourage your team to think outside the box
It’s important to follow the basics, but also be prepared to think outside the box when necessary! An experienced engineer once told us they had a customer who had issues early in the morning, and by the time his techs arrived to troubleshoot, the problem had gone away. He replaced the optics, thinking they were the source of the problem.
Fast forward to a year later, and the customer had a repeat failure. It was a record cold January morning, which replicated the previous year’s conditions. Turns out, there was some water in an outdoor splice enclosure that was freezing overnight and creating a micro-bend in the fiber, attenuating the signal. By the time the customer came in and noticed the issue, along with the dispatched technicians, the sun had melted the ice, eliminating the source of failure. If the engineer had not thought to compare year over year weather conditions, they would not have identified the root cause.
4. Train at the right time
Training your team as target technology evolves – from network engineers, to field technicians, to sourcing specialists – is important, but timing that training is even more critical. A good example is the transition from RF to optical in the cable world. Train too early and a technician may forget much of the training when an issue finally arrives. Train too late, and the team’s not properly prepared. New optical technology also requires the investment in and training on new optical test equipment, which needs to be budgeted into the scope of a new program.
5. Be cautious of working in silos
Communication between all stakeholders – procurement, network engineering, field operations – is critical. Send daily reports from the field to key stakeholders to ensure that various team concerns – quality, availability, technical performance – are addressed versus compartmentalized.
An example of this is when a procurement team recently standardized a single reach device to address all applications (40km, 20km, 10km) without the ops team’s knowledge. The operations team was then expected to deploy with external attenuators for each specific application. Had procurement and operations communicated, they could have identified and addressed the additional complexity and time to ensure the proper attenuator was implemented. Furthermore, the additional point of failure added a cost to the solution that negated any potential savings.
6. Know the terrain
Every network buildout has its unique challenges, from network spans and environmental concerns to accessibility issues – when do you have access, and for how long and how often? Reliability is always a concern, but even more so for hard-to-access environments. Don’t let your field techs be surprised by a challenging environment.
7. Always keep a spare
Sparing for flexibility during installation is of utmost importance. Repeated truck rolls during network turn-up impacts the bottom line and customer satisfaction. A network architect should work with the technicians to ensure they always have a “reasonable” number of spares out in the field. Sparing not only applies to startup, but also looking ahead at corner cases – odd links and unexpected or multiple OEM platforms at a remote site – that can be effectively managed ahead of time if your equipment provider has that flexibility to support a wide variety of interfaces.