| CPC H04B 10/03 (2013.01) [G06F 11/0709 (2013.01); G06F 11/079 (2013.01); H04B 10/079 (2013.01)] | 20 Claims |

|
1. A system for monitoring fault events at a fiber optical network, the system comprising:
one or more processors; and
one or more non-transitory computer-readable media comprising instructions that, when executed by the one or more processors, cause operations comprising:
receiving, for each interconnected component of an optical network, corresponding optical data comprising optical measurements obtained at each interconnected component, wherein the optical network is represented by a topology comprising a plurality of interconnected components;
extracting, from each corresponding optical data comprising corresponding optical measurements obtained at each interconnected component, a set of component metrics for light transmission signals transmitted or received, via fiber optic transmission lines, at a corresponding interconnected component;
modifying each corresponding optical data according to a standardized schema using the set of component metrics to obtain a modified component data structure;
generating a network event dataset by appending, for each interconnected component, (1) a corresponding modified component data structure and (2) event data comprising an indication as to whether an event has occurred and an event type if the event has occurred;
training a machine learning model using the network event dataset and the topology to identify occurrence of new events at components of the optical network;
responsive to determining a difference between a predicted network event and an actual network event, generating an adjusted standardized schema by modifying the set of component metrics; and
retraining the machine learning model using component data structures updated according to the adjusted standardized schema.
|