Computer scientists at the University of Texas at Dallas have invented a new weapon against cheating video games.
Investigators developed a fraud detection approach using the popular first-person shooter game Counter-Strike. But the mechanism can work for any online multiplayer game (MMO) that sends data traffic to a central server.
Their research was published online Reliable and secure computing IEEE operations.
Counter-Strike players work in teams to secure locations of factories against terrorists, detonating bombs and rescuing hostages. Players can earn currency in the game to buy more powerful weapons, which is the key to success. Various game software traps are available online.
“Sometimes you can tell when you’re playing against players who use cheating, but sometimes it’s not obvious,” said Md Shihabul Islam, Erik Jonsson of the UT Dallas School of Engineering and Computer Science, School of Engineering and Computer Science. research, to have fun playing Counter-Strike. “It’s not fair to other players.”
In addition to fair play, cheating can also have an economic impact on players who are unhappy when they go to play other games, Islam said.
Fraudulent events can have serious consequences for exports, a fast-growing industry with revenues of about $ 1 billion a year. The scam could lead to penalties against teams and players, including disqualification, loss of money and a ban on future participation, according to the UK Esports Integrity Commission.
Detecting scams in MMO games can be challenging because the data that goes from a player’s computer to the game server is encrypted. Previous investigations have relied on decrypted game records to detect fraud after the incident. The UT Dallas researchers ’approach eliminates the need for decrypted data and instead analyzes the encrypted data traffic to and from the server in real time.
“Players who cheat send traffic in a different way,” said Dr. Latifur Khan, the author of the research, a professor of computer science and director of the Big Data Analytics and Management Lab at UT Dallas. “We’re trying to capture those characteristics.”
“Once detected, we can give a warning and throw the player in an elegant way if they continue to cheat for a certain period of time. Our intention is to keep games like Counter-Strike fun and fair for all players.”
– Dr. Latifur Khan, Professor of Computer Science at the Erik Jonsson School of Engineering and Computer Science
For the study, 20 students from the UT Dallas-class Cyber Security Essentials for Practitioners downloaded Counter-Strike and three software traps: an aimbot that automatically targets the opponent; speed hack, which allows the player to move faster; and the wall, which makes the walls transparent so that players can easily see the opponent. The researchers created a server dedicated to the project so that student activities would not be disrupted by other online players.
The researchers looked at the server provided and the traffic in the game. The data is traveled in information packets or packets. Packages can be of different sizes, depending on the content. The researchers looked at features such as the number of incoming and outgoing packets, their size, the time they were transmitted, the direction, and the number of packets in an explosion, which is the set of continuous packets.
By controlling the data traffic of student players, the researchers identified patterns that represented fraud. They then used this information as a machine learning model, a form of artificial intelligence, to predict fraud based on game data models and features.
The researchers adapted their statistical model, based on a small set of players, to work for larger populations. Part of the scam detection mechanism is to send data traffic to a graphical processing unit, which is a parallel server, to speed up the process and remove the workload from the central processing unit of the main server.
The researchers want to extend their work to create an approach to games that do not use a client-server architecture and make the detection mechanism more secure. According to Islam, gaming companies can use the UT Dallas technique to train game software to detect fraud with their data. If fraud is detected, the system can take immediate action.
“Once detected,” Khan said, “we can give a warning and throw the player in an elegant way if they continue to cheat for a certain period of time.
“Our goal is to keep games like Counter-Strike fun and fair for all players.”
Reference: Md Shihab Islam, Bo Dong, Swarup Chandra, Latifur Khan and Bhavani M. Thuraisingham, “GCI: A GPU Based Transfer Learning Learning Approach to Cheat Detecting Cheat Computer Computer Game”, August 3, 2020, Reliable and secure computing IEEE operations.
DOI: 10.1109 / TDSC.2020.3013817
Other authors of the study include Swarup Chandra PhD’18, an engineer with a degree from Hewlett Packard Enterprise, and Bo Dong, a doctoral student in UT Dallas. Dr. Bhavani Thuraisingham, Professor of Engineering and Computer Science, Professor of Computer Science and Executive Director of the UT Dallas Institute for Cybersecurity Research and Education is the lead author of the research.
The research has been funded by the National Science Foundation, the Air Force Office of Scientific Research, the National Security Agency, IBM and Hewlett-Packard Development Co.