According to recent study, real-world experiments have revealed that gait authentication might be a realistic method of securing smartphones and other mobile devices from cybercrime.
A research sponsored by the University of Plymouth allowed smartphone users to go about their normal routines while motion sensors on their phones recorded data about their stride patterns.
The findings indicated that the system was around 85 percent accurate in recognizing an individual’s gait on average, with that percentage jumping to over 90 percent when they were walking regularly and quickly.
There are already over 6.3 billion smartphone users worldwide, who utilize their smartphones to deliver a variety of services as well as to store sensitive and secret information.
While authentication systems like as passwords, PINs, and biometrics are available, studies have shown that the degree of security and usefulness of such approaches varies significantly.
According to the researchers, who published their findings in Computers & Security, the study demonstrates that, within an acceptable context, gait recognition might be a feasible tool for safeguarding persons and their data from possible criminality.
Academics at Plymouth’s Centre for Cyber Security, Communications, and Network Research have been working on a variety of novel authentication processes in order to deliver more secure and useable solutions.
This research expands on previous work by evaluating a multi-algorithmic gait detection system and is the first to apply it to real-world data.
For the study, 44 volunteers ranging in age from 18 to 56 were instructed to carry a globally accessible smartphone device for seven to ten days.
They were instructed to put the smartphone in a belt pouch and collect sensor data obtained by the device’s gyroscope and accelerometer during various physical activities.
During the test, each participant created an average of 4,000 sample activities, which were divided into recordings indicating typical and rapid walking, as well as ascending and descending stairs.
This revealed a possible error rate of 11.38 percent and 11.32 percent for regular and rapid walking, respectively, with the result jumping to 24.52 percent and 27.33 percent when individuals went down and upstairs.
According to the researchers, this highlights the need to develop the capacity to automatically identify a broader range of walking activities so that a multi-algorithmic approach to identification can target particular walking features.
“As cellphones have evolved, security measures have had to increase dramatically,” said Nathan Clarke, Professor of Cyber Security and Digital Forensics at the University of Plymouth and a Fellow of the Chartered Institute of Information Security. This has resulted in a considerable increase in user authentication, with users being required to verify both their devices and the myriad applications they contain on a regular basis. Gait authentication has developed as a non-intrusive method of gathering a required amount of personal information, but all experiments of it have taken place in a controlled setting until now.
“Gait recognition alone will not be the solution to practical and easy authentication; nonetheless, it might constitute a crucially significant instrument inside the cyber arsenal, contributing to a greater knowledge of a user’s identity.” This research shows, for the first time outside of laboratory-controlled circumstances, what degree of performance is practically achievable. It is obvious that performance levels are harmed; nevertheless, the research also shown that for the majority of users, these concerns can be solved to an acceptable level.”
