Fiber Attenuation and Length for Adtran and Calix
A few weeks ago, I posted an analysis of signal strength vs distance in fiber (PON) networks based on a random sampling of ONT stats in Preseem. This is what that looks like for Adtran and Calix respectively.
In the plots below, attenuation is calculated as the Tx power as measured by the OLT minus the Rx power as measured by the ONT.
Perhaps unsurprisingly, the pattern looks very similar but inverted. It would take deeper statistical analysis to know for sure, but on quick inspection, it looks like using attenuation wouldn’t change any insights drawn from this vendor comparison.
The plots above are based on a single point in time for each ONT, which got me wondering how much variation there is in attenuation over a longer period. The plots below show the standard deviation per ONT over the course of two days.
This data shows that most ONTs show little attenuation variation over time, but there are some suspicious ones. It will be interesting to investigate the ONTs with high variation, and we’ll look for ONTs with sudden changes. In other ad hoc analyses, we’ve seen instances of fieldwork resulting in sudden, negative changes, and we were able to point them out to the customer for an immediate fix.
Lastly, the data shows more attenuation variation in the Adtran networks than Calix. That also warrants further investigation, so stay tuned!
