IoT Web of Trust for Sensor Security
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The diverse and non-uniform nature of IoT devices poses challenges in securing data transmission paths and detecting compromised sensors, as they have varying computing capabilities, network positions, communication protocols, and lack central control, making it difficult to ensure secure transmission of sensitive information.
Innovation Solution
A security framework that uses trusted devices to vouch for other devices, monitors electronic communication patterns, applies machine learning algorithms to classify devices as trusted or compromised, and implements a silo mechanism to isolate compromised devices, including blacklisting and quarantine periods, to maintain network security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a security framework is implemented for diverse IoT devices, then network security and detection capability are improved, but device complexity and implementation difficulty increase
Solution Approach 1:
The patent introduces a web of trust as an intermediary layer between IoT devices and the security monitoring system. This web of trust acts as a mediator that translates diverse device behaviors into standardized trust metrics, enabling security assessment without requiring complex integration with each individual device's heterogeneous security mechanisms.
Solution Approach 2:
The patent replaces traditional mechanical security approaches (such as centralized authentication servers and manual security configurations) with a decentralized, algorithm-based web of trust system. Machine learning algorithms automatically analyze communication patterns and establish trust relationships, eliminating the need for complex manual security management infrastructure.
2Measurement precision
If machine learning algorithms are used to classify devices, then detection precision is improved, but computational energy consumption increases
Solution Approach 1:
The patent applies machine learning algorithms selectively rather than continuously to all devices. The system monitors communication patterns and applies classification algorithms only when suspicious activities are detected or when trust assessments need updating, reducing overall computational energy consumption while maintaining high detection precision through targeted analysis.
Solution Approach 2:
The system performs preliminary lightweight monitoring of communication patterns continuously, and only invokes full machine learning classification when anomalies are detected. This preliminary action filters out normal traffic without intensive computation, reserving high-energy ML algorithms for cases where they are most needed, thus balancing detection precision with energy efficiency.
3Difficulty of detecting and measuring
If communication patterns are monitored continuously, then detection capability is improved, but loss of time and processing overhead increase
Solution Approach 1:
The patent implements periodic sampling of communication patterns rather than continuous monitoring. The web of trust system evaluates trust metrics at regular intervals and updates classifications periodically, which reduces processing time and computational overhead while maintaining effective detection capability through systematic periodic assessment of device behaviors.
Solution Approach 2:
The system skips detailed analysis of normal, predictable communication patterns and focuses processing resources on detecting deviations and anomalies. By rushing through routine traffic with lightweight validation and applying intensive analysis only when necessary, the system maintains high detection capability while minimizing overall processing time.
Data Source
AI summary
Apparatus and methods are provided for tracking and validating behavior and communication patterns of sensors connected to an Internet-of-Things (“IoT”) network. Preferably, trusted IoT sensors monitor communication patterns exhibited by other trusted and/or untrusted sensors. An untrusted monitored sensor may be assigned a trusted status based on applying artificial intelligence and/or machine learning algorithm to monitored and/or historical communication patterns exhibited by the monitored sensor. A trusted group of sensors may continue to grow by adding other trusted sensors. If a compromised sensor is detected, a silo may be erected around the compromised sensor. The silo may include disconnecting the compromised sensor from the network. After erecting the silo, communication patterns exhibited by the compromised sensor may be continue to be monitored. After a pre-determined time period the compromised sensor may be reassigned a trusted status or purged from the trusted group and/or network.