Fault classification
AssigneeTelefonaktiebolaget LM Ericsson (publ)
InventorsRoberto Magri, Paolo Debenedetti, Stefano Orsi, Marcello Morchio
Life of this patent
Estimating the term…
Abstract
A method of identifying a fault in an optical link comprises continually storing samples of a signal that represents a power of an optical signal received over the optical link. When a fault is detected on said optical link, at least one stored series of samples of the signal is retrieved. The method then comprises performing a classification on the retrieved series of samples, in order to classify the series of samples as resulting from one of a plurality of predefined faults.
Claims17
A method of identifying a fault in an optical link, the method comprising:
- continually storing samples of a signal, wherein the signal represents a power of an optical signal received over the optical link;
- when a fault is detected on said optical link, retrieving at least one stored series of samples of the signal, wherein retrieving the stored samples comprises retrieving samples stored before the fault was detected;
- and
- performing a classification on the retrieved series of samples, to classify the series of samples as resulting from one of a plurality of predefined faults.
A system for identifying a fault in an optical link, the system comprising:
- a memory configured for storing at least one stored series of samples of a signal, wherein the signal represents a power of an optical signal received over the optical link, and
- a processor configured to, when a fault is detected on said optical link, retrieve said at least one stored series of samples, and perform a classification on the retrieved series of samples, in order to classify the series of samples as resulting from one of a plurality of predefined faults;
- wherein the memory and processor are configured to store the samples of the signal for a predetermined time and then overwrite the stored samples with new samples.
An optical module, comprising:
- a receiver assembly, configured to convert a received optical signal into an electrical signal and to generate a signal representing a power of the received optical signal;
- an analog-digital converter configured to generate samples of the signal representing the power;
- and
- a memory circuit configured to store the samples of the signal for a predetermined time, and then overwrite the stored samples with new samples, such that the stored samples can be retrieved when a fault is detected on an optical link including the optical module.
A network node, comprising:
- a memory for storing a classification model,
- wherein the network node is configured to:
- when a fault is detected on an optical link, receive a stored series of samples of a signal representing a power of an optical signal received in an optical module forming part of the optical link;
- and
- perform a classification on the received series of samples, to classify the samples as resulting from one of a plurality of predefined faults, wherein performing the classification comprises determining a match with one of the plurality of predefined faults for one or more of:
- a gradient of the power of the optical signal against time;
- an amount of power, duration and/or number of temporary increase(s) in the power of the optical signal, following a drop in the power of the optical signal;
- an amount of variation of the power, and/or duration, of a plurality of temporary increases in the power of the optical signal, following a drop in the power of the optical signal;
- an amount of noise in the power of the optical signal against time;
- a final power of the stored series of samples of the optical signal;
- and
- a time for the power of the optical signal to drop to zero or a final power.
Dependencies are inferred from the claim text.