Identifying link fingerprint anomalies in real-time in an all-photonics network

1. A data communication network, comprising:

a network node including an optical link and a reflectometry analyzer configured to provide a plurality of reflectometry results, each reflectometry result providing a characterization of physical and operational properties of the optical link at the time of the reflectometry result; and

a processor configured to:

receive a first set of the reflectometry results;

analyze the first set of reflectometry results to define a fingerprint of the physical and operational properties of the optical link;

receive a second set of the reflectometry results;

compare the second set of reflectometry results with the fingerprint; and

determine whether or not the optical link is secure based upon the comparison of the second set of reflectometry results with the fingerprint;

wherein the processor utilizes a machine learning algorithm to define the fingerprint of the physical and operational properties of the optical link based upon the first set of reflectometry results.

2. The data communication network of claim 1 , wherein the fingerprint of the physical and operational properties of the optical link includes a first mean value of the first set of reflectometry results.

3. The data communication network of claim 2 , wherein, in determining whether or not the optical link is secure, the processor is further configured to:

determine that the optical link is secure when a second mean value of the second set of reflectometry results indicates a difference from the first mean value that typifies aging of the optical link.

4. The data communication network of claim 2 , wherein, in determining whether or not the optical link is secure, the processor is further configured to:

determine that the optical link is not secure when a second mean value of the second set of reflectometry results indicates a difference from the first mean value that typifies tampering with the optical link.

5. The data communication network of claim 1 , wherein the fingerprint of the physical and operational properties of the optical link includes a first standard deviation from the mean value of the first set of reflectometry results.

6. The data communication network of claim 5 , wherein, in determining whether or not the optical link is secure, the processor is further configured to:

determine that the optical link is secure when a second standard deviation of the second set of reflectometry results indicates a difference from the first standard deviation that typifies changes in an environment surrounding the optical link.

7. The data communication network of claim 5 , wherein, in determining whether or not the optical link is secure, the processor is further configured to:

determine that the optical link is not secure when a second standard deviation of the second set of reflectometry results indicates a difference from the first standard deviation that typifies tampering with the environment surrounding the optical link.

8. The data communication network of claim 1 , wherein the machine learning algorithm includes at least one of a linear regression algorithm, a logistic regression algorithm, a decision tree algorithm, a support vector machine algorithm, a naïve Bayes algorithm, a k-nearest neighbor algorithm, a k-means algorithm, a random forest algorithm, a dimensionality reduction algorithm, a gradient boosting algorithm, and a directed acyclic graph (DAG) algorithm.

9. The data communication network of claim 1 , wherein the reflectometry analyzer includes the processor.

10. A method, comprising:

providing, on a network node of a data communication network, an optical link and a reflectometry analyzer;

providing, by the reflectometry analyzer, a plurality of reflectometry results, each reflectometry result providing a characterization of physical and operational properties of the optical link at the time of the reflectometry result;

receiving, by a processor of the data communication network, a first set of the reflectometry results;

analyzing the first set of reflectometry results to define a fingerprint of the physical and operational properties of the optical link;

receiving a second set of the reflectometry results;

comparing the second set of reflectometry results with the fingerprint;

determining whether or not the optical link is secure based upon the comparison of the second set of reflectometry results with the fingerprint; and

utilizing a machine learning algorithm to define the fingerprint of the physical and operational properties of the optical link based upon the first set of reflectometry results.

11. The method of claim 10 , wherein the fingerprint of the physical and operational properties of the optical link includes a first mean value of the first set of reflectometry results.

12. The method of claim 11 , wherein, in determining whether or not the optical link is secure, the method further comprises:

determining that the optical link is secure when a second mean value of the second set of reflectometry results indicates a difference from the first mean value that typifies aging of the optical link.

13. The method of claim 11 , wherein, in determining whether or not the optical link is secure, the method further comprises:

determining that the optical link is not secure when a second mean value of the second set of reflectometry results indicates a difference from the first mean value that typifies tampering with the optical link.

14. The method of claim 10 , wherein the fingerprint of the physical and operational properties of the optical link includes a first standard deviation from the mean value of the first set of reflectometry results.

15. The method of claim 14 , wherein, in determining whether or not the optical link is secure, the method further comprises:

determining that the optical link is secure when a second standard deviation of the second set of reflectometry results indicates a difference from the first standard deviation that typifies changes in an environment surrounding the optical link.

16. The method of claim 14 , wherein, in determining whether or not the optical link is secure, the method further comprises:

determining that the optical link is not secure when a second standard deviation of the second set of reflectometry results indicates a difference from the first standard deviation that typifies tampering with the environment surrounding the optical link.

17. The method of claim 10 , wherein the machine learning algorithm includes at least one of a linear regression algorithm, a logistic regression algorithm, a decision tree algorithm, a support vector machine algorithm, a naïve Bayes algorithm, a k-nearest neighbor algorithm, a k-means algorithm, a random forest algorithm, a dimensionality reduction algorithm, a gradient boosting algorithm, and a directed acyclic graph (DAG) algorithm.

18. A network node of a data communication network, the network node comprising:

an optical link;

a reflectometry analyzer configured to provide a plurality of reflectometry results, each reflectometry result providing a characterization of physical and operational properties of the optical link at the time of the reflectometry result; and

a processor configured to:

receive a first set of the reflectometry results;

analyze the first set of reflectometry results to define a fingerprint of the physical and operational properties of the optical link;

receive a second set of the reflectometry results;

compare the second set of reflectometry results with the fingerprint; and

determine whether or not the optical link is secure based upon the comparison of the second set of reflectometry results with the fingerprint;

wherein the processor utilizes a machine learning algorithm to define the fingerprint of the physical and operational properties of the optical link based upon the first set of reflectometry results.