According to recent study, gait authentication has proven to be a practical method of preventing cybercrime on smartphones and other mobile devices in real-world tests.
Smartphone users were instructed to go about their everyday activities while motion sensors in their mobile devices recorded information on their stride patterns for a study directed by the University of Plymouth.
The findings revealed that the system was generally about 85% accurate in identifying a person’s gait, but that number increased to about 90% when they were walking regularly and quickly.
Around the world, there are more than 6.3 billion smartphone users who use their smartphones to offer a variety of services and store private and sensitive data.
Although there are authentication methods like passwords, PINs, and biometrics, studies have shown that the degree of security and usability of such methods vary greatly.
According to the researchers’ article in Computers & Security, the study shows that, given the right conditions, gait recognition could be a practical method for shielding people and their data from potential criminal activity.
In order to deliver more secure and practical solutions, academics at Plymouth’s Centre for Cyber Security, Communications and Network Research have concentrated on developing a variety of novel authentication procedures.
As smartphones have developed, security controls have had to advance significantly. This has led to a significant rise in user authentication, where users repeatedly need to authenticate both their devices and the numerous apps they contain. Gait authentication has emerged as a non-intrusive way of capturing a necessary level of personal information, but until now all tests of it have taken place in a controlled environment.Professor Nathan Clarke
This study expands on that earlier work by evaluating a multi-algorithmic gait detection system and becoming the first to use real-world data in its application.
For the study, 44 volunteers between the ages of 18 and 56 were instructed to carry a smartphone for seven to ten days.
To record the sensor data gathered by the smartphone’s gyroscope and accelerometer during various physical activities, they were instructed to put the device in a belt pouch.
During the test, each participant produced an average of 4,000 sample activities, which were divided into records of slow and fast walking, as well as climbing and descending stairs.
With individuals moving up or down stairs, the potential error rate increased to 24.52% and 27.33%, respectively, from 11.38% and 11.32% for regular and rapid walking, respectively.
This, according to the researchers, emphasizes the necessity to develop the capacity to automatically distinguish a larger range of walking activities so that a multi-algorithmic approach to identification can focus on particular walking traits.
Nathan Clarke, Professor of Cyber Security and Digital Forensics at the University of Plymouth and recently made a Fellow of the Chartered Institute of Information Security, said:
“As smartphones have developed, security controls have had to advance significantly. This has led to a significant rise in user authentication, where users repeatedly need to authenticate both their devices and the numerous apps they contain. Gait authentication has emerged as a non-intrusive way of capturing a necessary level of personal information, but until now all tests of it have taken place in a controlled environment.”
“Gait recognition alone will not be the answer to usable and convenient authentication, however it could form a critically important tool within the cyber arsenal that could contribute towards creating a stronger awareness of a user’s identity. This study demonstrates, for the first time outside of laboratory-controlled conditions, what level of performance can be achieved realistically. It is clear performance levels are impacted; however the study has also shown that for most users these issues can be overcome to an acceptable level.”