Bot detection in an edge network using transport layer security (TLS) fingerprint

Patent metadata

Field	Value
Publication number	US-12316672-B2
Application number	US-202318383351-A
Country	US
Kind code	B2
Filing date	Oct 24, 2023
Priority date	Dec 18, 2017
Publication date	May 27, 2025
Grant date	May 27, 2025

Abstract

This disclosure describes a technique to fingerprint TLS connection information to facilitate bot detection. The notion is referred to herein as “TLS fingerprinting.” Preferably, TLS fingerprinting herein comprises combining different parameters from the initial “Hello” packet send by the client. In one embodiment, the different parameters from the Hello packet that are to create the fingerprint (the “TLS signature”) are: record layer version, client version, ordered TLS extensions, ordered cipher list, ordered elliptic curve list, and ordered signature algorithms list. Preferably, the edge server persists the TLS signature for the duration of a session.

First claim

We claim:

1. A computer program product in a non-transitory computer readable medium, the computer program product comprising program code executable in and across a set of hardware processors, the program code comprising:
first program code executable in a first hardware processor and configured to execute a script and, responsive to execution of the script, to output a set of transport layer security parameters;
second program code executable in a second hardware processor and configured to receive and extract the set of transport layer security parameters, to process the extracted set of transport layer security parameters into a hash value, and to selectively output the hash value; and
third program code executable in a third hardware processor and configured to receive the hash value, based on a machine learning model, carry out a determination whether an entity that generated the set of transport layer security parameters as reflected by the hash value has a characteristic associated with a bot, and to provide the second program code an indication based on the determination;
wherein the second program code is further configured to receive and to take a given action based on the indication.

2. The computer program product as described in claim 1 wherein the set of transport layer security parameters are associated with a Transport Layer Security (TLS) connection initiated by a client entity associated with the first hardware processor.

3. The computer program product as described in claim 2 wherein the TLS connection is initiated by the client entity by a client hello message.

4. The computer program product as described in claim 3 wherein the set of transport layer security parameters comprise information derived from the client hello message, the information being at least one of: a record layer version, a client version, a set of ordered TLS extensions, an ordered cipher list, an ordered elliptic curve list, and an ordered signature algorithms list.

5. The computer program product as described in claim 1 wherein the third program code is further configured to persist the hash value for a duration of a session associated with a client request.

6. The computer program product as described in claim 1 wherein the given action is one of: denying a client request, tarpit the client request, serving alternative content in response to the client request, passing the client request forward to an origin, and serving a response to the client request.

7. The computer program product as described in claim 1 wherein the second program code is further configured to check the hash value against a data set of known bad hash values.

8. The computer program product as described in claim 7 wherein the third program code is further configured to generate the data set of known bad hash values, and to provide the data set of known bad hash values to a server entity associated with the second hardware processor.

9. The computer program product as described in claim 1 wherein the third program code is further configured to generate the machine learning model using supervised learning.

10. The computer program product as described in claim 1 wherein the first hardware processor is associated with a client entity, the second hardware processor is associated with a server entity, and the third hardware processor is associated with a bot detection service entity.

11. The computer program product as described in claim 10 wherein the server entity is one of a set of server entities of a content delivery network (CDN).

12. The computer program product as described in claim 10 wherein the second progam code of the server entity is further configured to deliver the script to the client entity.

13. The computer program product as described in claim 1 wherein the first hardware processor is located remote from the second hardware processor.

14. The computer program product as described in claim 1 wherein the second hardware processor is located remote from the third hardware processor.

15. The computer program product as described in claim 1 wherein the set of transport layer security parameters comprise a fingerprint.

Assignees

• Akamai Tech Inc

Inventors

• Senecal David
• Kahn Andrew
• Segal Ory
• Shuster Elad
• Nguyen Duc

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N20/00

  Machine learning · CPC title

• H04L63/166

  at the transport layer · CPC title

• H04L67/02

  based on web technology, e.g. hypertext transfer protocol [HTTP] · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title