Metrics-Augmented Authentication via Latency and Data Entry Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current authentication methods for accessing computing systems rely heavily on traditional credentials like UID and password, which are vulnerable to fraudulent activities, as they do not adequately verify the authenticity of clients based on their behavior patterns and environmental metrics.
Innovation Solution
The implementation of a Metrics-Augmented Authentication (MAA) system that utilizes login metrics such as latency, network metrics, and data entry patterns to authenticate clients by comparing these metrics to registered metrics associated with authentic clients, thereby determining the comporting status and performing appropriate authentication actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional credential-based authentication (UID and password) is used, then the authentication process is simple and fast, but it is vulnerable to fraudulent activities and cannot adequately verify client authenticity
Solution Approach 1:
The patent transforms the authentication approach by changing from static credential verification to dynamic metric analysis. It measures multiple login metrics (latency, network metrics, data entry patterns) and compares them against registered metrics to determine comporting status, thereby enhancing security through parameter-based behavioral analysis rather than simple password checking
Solution Approach 2:
The authentication agent serves as an intermediary between the login agent and the target system. It intercepts login requests, collects and analyzes multiple metrics, determines comporting status, and decides whether to permit or deny access. This intermediary layer adds security verification without requiring changes to the core authentication infrastructure
2Measurement precision
If multiple login metrics are collected and analyzed, then the accuracy of distinguishing authentic from fraudulent clients improves, but the authentication processing time increases
Solution Approach 1:
The system performs preliminary actions by collecting and registering login metrics during normal login operations before fraud detection is needed. The authentication agent continuously gathers metric data (latency, network metrics, data entry patterns) and stores them as baseline registered metrics, so that when authentication is required, the comparison can be performed quickly against pre-established profiles
Solution Approach 2:
The patent implements a graduated authentication approach where the system can perform partial metric collection and analysis based on risk assessment. Not all metrics need to be fully analyzed in every case - the system can perform quick initial assessments and only conduct full multi-metric analysis when fraud suspicion arises, thereby balancing accuracy with processing efficiency
3Object-affected harmful factors
If strict metrics-based authentication actions are taken (terminating connections, preventing login), then unauthorized access is reduced, but legitimate client access may be blocked
Solution Approach 1:
The system implements feedback mechanisms where authentication decisions are based on comporting status determination that compares current login metrics against registered metrics. The feedback loop allows the system to learn from authentication outcomes and adjust its behavior. Additionally, the patent mentions feedback from honeypot systems that provide information about fraudulent activities, enabling the system to refine its authentication decisions and reduce false positives
Solution Approach 2:
The patent employs preliminary anti-action through honeypot systems that are set up to detect and study fraudulent login attempts before they can access the real target system. The honeypot collects data about fraudulent behaviors and metrics, enabling the authentication agent to preemptively identify and block similar fraudulent attempts against the actual system, while legitimate users are unaffected
Data Source
AI summary
A method comprises an authentication agent receiving a communications protocol message from a login agent of a client attempting to login to a target system. The authentication agent determines a login metric associated with the protocol message. The login metric comprises a latency, network, and/or data entry metric. The authentication agent receives credentials associated with an authentic client of the target system and compares the login metric with a registered metric associated with the authentic client. Based on the login metric comporting with the registered metric, the authentication agent continues login processing or performs a non-comporting metric action. Another method comprises the authentication agent sending a training request to the login agent, receiving a training response, determining a login metric associated with the training response, and recording the login metric among registered metrics of an authentic client. A computing system can implement the methods.