Classification of encrypted internet traffic
A method includes obtaining a first plurality of encrypted traffic flows traversing a communication network, performing a first classification, wherein a result of the first classification identifies a traffic type associated with each encrypted traffic flow of the first plurality of encrypted traffic flows, and wherein the first classification is based on a traffic pattern of the each encrypted traffic flow, performing a second classification, wherein a result of the second classification identifies a traffic type associated with each server name indication from which the first plurality of encrypted traffic flows is associated, and wherein the second classification is based on the result of the first classification, and performing a third classification identifying a traffic type associated with each encrypted traffic flow of the first plurality of encrypted traffic flows, wherein the third classification is based on a combination of the results of the first classification and the second classification.
1. A method comprising:
obtaining, by a processing system in a communication network, a first plurality of encrypted traffic flows traversing the communication network;
performing, by the processing system, a first classification, wherein a result of the first classification identifies a traffic type associated with each encrypted traffic flow of the first plurality of encrypted traffic flows, and wherein the first classification is based on a traffic pattern of the each encrypted traffic flow;
performing, by the processing system, a second classification, wherein a result of the second classification identifies a traffic type associated with each server name indication from which the first plurality of encrypted traffic flows is associated, and wherein the second classification is based on the result of the first classification, wherein the performing the second classification comprises:
identifying for each encrypted traffic flow of the first plurality of encrypted traffic flows, the server name indication;
identifying a first subset of the first plurality of encrypted traffic flows, wherein each encrypted traffic flow in the first subset is associated with a first server name indication;
identifying a most common traffic type among the first subset, based on the result of the first classification; and
assigning the most common traffic type to the first server name indication; and
performing, by the processing system, a third classification identifying a traffic type associated with each encrypted traffic flow of the first plurality of encrypted traffic flows, wherein the third classification is based on a combination of the result of the first classification and the result of the second classification.
2. The method of claim 1 , further comprising:
filtering, by the processing system, a pool of encrypted traffic flows traversing the communication network into the first plurality of encrypted traffic flows and a second plurality of encrypted traffic flows, wherein the first plurality of encrypted traffic flows comprises encrypted traffic flows for which a metric measuring one or more conditions of the communication network at least meets a predefined threshold, and wherein the second plurality of encrypted traffic flows comprises encrypted traffic flows for which the metric falls below the predefined threshold.
3. The method of claim 2 , wherein the third classification additionally identifies a traffic type associated with each encrypted traffic flow of the second plurality of encrypted traffic flows.
4. The method of claim 1 , wherein the first plurality of encrypted traffic flows comprises a plurality of hypertext transfer protocol secure flows.
5. The method of claim 1 , wherein the first plurality of encrypted traffic flows comprises a plurality of quic flows.
6. The method of claim 1 , wherein the first classification is performed using a machine learning model that is trained on third party media stores using a supervised learning approach.
7. The method of claim 1 , wherein the most common traffic type is identified by majority voting.
8. The method of claim 1 , wherein the most common traffic type is identified by bayesian classification.
9. The method of claim 1 , wherein the second classification further includes identifying a confidence in the traffic type associated with the each server name indication.
10. The method of claim 1 , wherein the third classification is performed using a bayesian classifier.
11. The method of claim 10 , wherein the bayesian classifier uses the second classification as a prior.
12. The method of claim 1 , wherein the traffic pattern comprises a number of bytes transmitted within each time slot of a plurality of time slots for the each encrypted traffic flow.
13. The method of claim 1 , wherein the processing system is part of a network element deployed in the communication network.
14. The method of claim 1 , further comprising:
performing, by the processing system, an optimization to the communication network in response to the third classification, wherein the optimization optimizes a quality of experience associated with a traffic type associated with a subset of encrypted traffic flows of the first plurality of encrypted traffic flows.
15. A device comprising:
a processing system in a communication network; and
a non-transitory computer-readable medium storing instructions which, when executed by the processing system, cause the processing system to perform operations, the operations comprising:
obtaining a first plurality of encrypted traffic flows traversing the communication network;
performing a first classification, wherein a result of the first classification identifies a traffic type associated with each encrypted traffic flow of the first plurality of encrypted traffic flows, and wherein the first classification is based on a traffic pattern of the each encrypted traffic flow;
performing a second classification, wherein a result of the second classification identifies a traffic type associated with each server name indication from which the first plurality of encrypted traffic flows is associated, and wherein the second classification is based on the result of the first classification, wherein the performing the second classification comprises:
identifying for each encrypted traffic flow of the first plurality of encrypted traffic flows, the server name indication;
identifying a first subset of the first plurality of encrypted traffic flows, wherein each encrypted traffic flow in the first subset is associated with a first server name indication;
identifying a most common traffic type among the first subset, based on the result of the first classification; and
assigning the most common traffic type to the first server name indication; and
performing a third classification identifying a traffic type associated with each encrypted traffic flow of the first plurality of encrypted traffic flows, wherein the third classification is based on a combination of the result of the first classification and the result of the second classification.
16. The device of claim 15 , wherein the operations further comprise:
filtering a pool of encrypted traffic flows traversing the communication network into the first plurality of encrypted traffic flows and a second plurality of encrypted traffic flows, wherein the first plurality of encrypted traffic flows comprises encrypted traffic flows for which a metric measuring one or more conditions of the communication network at least meets a predefined threshold, and wherein the second plurality of encrypted traffic flows comprises encrypted traffic flows for which the metric falls below the predefined threshold.
17. The device of claim 16 , wherein the third classification additionally identifies a traffic type associated with each encrypted traffic flow of the second plurality of encrypted traffic flows.
18. The device of claim 15 , wherein the third classification is performed using a bayesian classifier, and the bayesian classifier uses the second classification as a prior.
19. The device of claim 15 , wherein the most common traffic type is identified by majority voting.
20. A non-transitory computer-readable medium storing instructions which, when executed by a processing system of a communication network, cause the processing system to perform operations, the operations comprising:
obtaining a first plurality of encrypted traffic flows traversing the communication network;
performing a first classification, wherein a result of the first classification identifies a traffic type associated with each encrypted traffic flow of the first plurality of encrypted traffic flows, and wherein the first classification is based on a traffic pattern of the each encrypted traffic flow;
performing a second classification, wherein a result of the second classification identifies a traffic type associated with each server name indication from which the first plurality of encrypted traffic flows is associated, wherein the second classification is based on the result of the first classification, wherein the performing the second classification comprises:
identifying for each encrypted traffic flow of the first plurality of encrypted traffic flows, the server name indication;
identifying a first subset of the first plurality of encrypted traffic flows, wherein each encrypted traffic flow in the first subset is associated with a first server name indication;
identifying a most common traffic type among the first subset, based on the result of the first classification; and
assigning the most common traffic type to the first server name indication; and
performing a third classification identifying a traffic type associated with each encrypted traffic flow of the first plurality of encrypted traffic flows, wherein the third classification is based on a combination of the result of the first classification and the result of the second classification.