Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Tor website fingerprint identification method based on attention mechanism and LSTM

A technology of fingerprint identification and attention, which is applied in the field of network security, can solve the problems of low accuracy of fingerprint identification and achieve the effect of preventing over-fitting phenomenon

Pending Publication Date: 2022-06-03
BEIJING UNIV OF POSTS & TELECOMM
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0009] Aiming at the deficiencies of the prior art, the present invention provides a Tor website fingerprint identification method based on the attention mechanism and LSTM, which solves the problem of relatively low accuracy in the existing Tor website fingerprint identification

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Tor website fingerprint identification method based on attention mechanism and LSTM
  • Tor website fingerprint identification method based on attention mechanism and LSTM
  • Tor website fingerprint identification method based on attention mechanism and LSTM

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0036] A Tor website fingerprinting method based on attention mechanism and LSTM, which specifically includes the following steps:

[0037] Step 1. Traffic capture: Use the tcpdump tool to collect access traffic, and label each access traffic to capture the access traffic of the monitored website;

[0038] Step 2: Feature screening: Data preprocessing is performed on the captured access traffic to obtain a sequence of TLS record lengths. The TLS record length is approximated to an integer multiple of 512, symbols are added according to the direction of the data packets, and finally converted into a cell representation sequence, as Input to the website fingerprinting model;

[0039]Step 3. Model construction: Build the input layer, Partition layer, LSTM layer, attention mechanism layer, fusion layer and softmax layer, build a classifier model based on attention mechanism and LSTM, and input the cell representation sequence in step 2 into In the model, the cell representation s...

Embodiment 2

[0043] This embodiment is an improvement of the previous embodiment, a method for fingerprinting Tor website based on attention mechanism and LSTM, which specifically includes the following steps:

[0044] Step 1. Traffic capture: first set the list of websites to be monitored, and obtain the addresses of the monitored websites, visit the monitored websites in sequence through Tor Browser, visit each website 1000 times, use the tcpdump tool to collect the access traffic, and analyze the Each access traffic is tagged, that is, the website category to which the access traffic belongs, and the access traffic of the monitored website is captured;

[0045] Step 2: Feature screening: perform data preprocessing on the captured access traffic, perform TCP stream reorganization on the TCP data packet sequence, record the length of each TLS record, convert it into a TLS record length sequence, and obtain a TLS record length sequence, Approximate the length of the TLS record to an intege...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a Tor website fingerprint identification method based on an attention mechanism and an LSTM. The method specifically comprises the following steps: step 1, traffic capture; step 2, feature screening; 3, constructing a model; step 4, model training; and step 5, performance evaluation: performing performance evaluation on the trained classifier model by using the test set in the step 4. The invention relates to the technical field of network security. According to the Tor website fingerprint identification method based on the attention mechanism and the LSTM, a cell sequence is extracted to serve as a model input feature based on data features of Tor access traffic, a time feature is extracted from the cell sequence by using a multi-branch LSTM network based on a mode of combining the attention mechanism and the LSTM, and then the attention mechanism is used for performing weight parameter optimization on the time feature, so that the Tor website fingerprint identification method based on the attention mechanism and the LSTM is obtained. Therefore, the influence of the key time characteristics on the website fingerprint identification result is found and highlighted, and the defect that the existing Tor website fingerprint identification accuracy is relatively low is overcome.

Description

technical field [0001] The invention relates to the technical field of network security, in particular to a Tor website fingerprint identification method based on an attention mechanism and LSTM. Background technique [0002] The Onion Router (Tor) is a communication tool that provides anonymity to Internet users. Tor encrypts the content of communication and routing information, not only forwards encrypted traffic through randomly assigned nodes, but also divides traffic data into cells with a fixed length of 512B, and uses anonymity methods such as periodic switching circuits to ensure users Privacy of browsing activity. There are already millions of users who use Tor to visit websites anonymously every day to hide their online activities. [0003] Recurrent neural network is a special type of neural network with self-connection in the field of deep learning, which is suitable for processing time series data. Although RNN performs well in processing time series relations...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): H04L9/40H04L41/16H04L67/02G06N3/04G06N3/08
CPCH04L63/1408H04L63/20H04L41/16H04L67/02G06N3/08G06N3/044G06N3/045
Inventor 千万峰
Owner BEIJING UNIV OF POSTS & TELECOMM
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com