Systems and networking research team receives Best Student Paper award at SIGMETRICS 2021

Monday, June 28, 2021

Research conducted by recent CS master’s graduate Iman Akbari has received a Best Student Paper award at SIGMETRICS 2021, the flagship conference of the Association for Computing Machinery’s special interest group for the computer systems performance evaluation community.

Iman’s award-winning research was conducted with Cheriton School of Computer Science Research Professor Mohammad A. Salahuddin, undergraduate CS student Leni Ven, Research Professor Noura Limam, and his supervisor Professor Raouf Boutaba, along with Bertrand Mathieu, Stephanie Moteau and Stephane Tuffin, their international colleagues at Orange Labs in Lannion, France.

Their paper, “A Look Behind the Curtain: Traffic Classification in an Increasingly Encrypted Web,” which was presented virtually at SIGMETRICS in June 2021 and published in the Proceedings of the ACM on Measurement and Analysis of Computing Systems, examined the use of machine learning to classify encrypted network traffic.

Classifying traffic on a computer network is a critically important task, one that is essential for network operators and managers. Traffic classification is indispensable for a range of operations and management tasks from capacity planning and performance monitoring to security provision and intrusion detection to ensuring quality of service guarantees. But as critically important as it is, classifying traffic has become increasingly difficult because of the widespread adoption of encryption. Those secure websites we visit — the ones with https in the address — are necessary to preserve our privacy online, but encryption adds to the complexity and challenge of traffic classification, for example, identifying services and applications being used.

Neural networks employing machine learning and deep learning are increasingly being used to classify network traffic. But despite their potential, a number of fundamental issues limit their promise. For example, some work in traffic classification has used service and application labels arbitrarily, limiting the ability of machine learning approaches to analyze real-world traffic. The true power of deep learning is realized when the models learn about the nature of traffic in different service and application classes and identify complex patterns to distinguish between them.

To this end, the research team used deep learning to classify services — such as video streaming, social media and web mail — and focused on new encrypted web protocols. Importantly, they looked exclusively at encrypted web traffic, and used deep learning to find complex patterns in network traffic that are characteristic of each service class. 

“We evaluated our deep learning approach using a real-world traffic dataset from a major Internet service provider and mobile network operator,” said Raouf Boutaba, Professor and Director of the Cheriton School of Computer Science. “Our team achieved an accuracy of 95 percent in classifying services using less raw traffic and a smaller number of parameters, significantly outperforming the current state-of-the-art method. We also evaluated our approach on a public dataset that has finer application-level granularity in labelling, and here we achieved an overall accuracy of 99 percent.”


To learn more about this award-winning research, please see Iman Akbari, Mohammad A. Salahuddin, Leni Ven, Noura Limam, Raouf Boutaba, Bertrand Mathieu, Stephanie Moteau, Stephane Tuffin. A Look Behind the Curtain: Traffic Classification in an Increasingly Encrypted Web. Proceedings of the ACM on Measurement and Analysis of Computing Systems, March 2021, vol 5(1): 1–26. https://doi.org/10.1145/3447382

  1. 2021 (49)
    1. September (3)
    2. August (7)
    3. July (4)
    4. June (8)
    5. May (9)
    6. April (3)
    7. March (8)
    8. February (4)
    9. January (3)
  2. 2020 (73)
    1. December (7)
    2. November (6)
    3. October (4)
    4. September (5)
    5. August (4)
    6. July (7)
    7. June (4)
    8. May (11)
    9. April (13)
    10. March (3)
    11. February (3)
    12. January (6)
  3. 2019 (90)
    1. December (3)
    2. November (8)
    3. October (6)
    4. September (9)
    5. August (7)
    6. July (8)
    7. June (10)
    8. May (5)
    9. April (11)
    10. March (6)
    11. February (10)
    12. January (7)
  4. 2018 (82)
  5. 2017 (50)
  6. 2016 (27)
  7. 2015 (41)
  8. 2014 (32)
  9. 2013 (46)
  10. 2012 (17)
  11. 2011 (20)