IoS Anomaly Detection Using Sensory Payload and Packet Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solutions are inadequate for detecting cyberbullying through abusive senses in Internet of Senses (IoS) applications, which involve tactile, olfactory, or thermal interactions, as they are limited to text-based or image-based analytics.
Innovation Solution
A method involving payload-based and packet-based analyses is applied to sensory, network, and context data to detect anomalies in IoS applications, using machine learning models and rules-based algorithms to identify abusive interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If text-based or image-based analytics are used for cyberbullying detection, then detection capability for traditional cyberbullying is achieved, but detection capability for sense-based cyberbullying in IoS applications is lost
Solution Approach 1:
The detection system is designed to handle multiple types of data (sensory data, network data, context data) through a unified anomaly detection framework. The network node performs payload-based analysis on sensory and context data while simultaneously performing packet-based analysis on network data, making the system versatile across different data types while maintaining reliable detection through consistent anomaly identification mechanisms.
Solution Approach 2:
The invention extends cyberbullying detection from traditional text/image dimensions to include sensory dimensions (tactile, olfactory, thermal). By analyzing sensory data packets in addition to traditional content, the system adds a new dimension to detection capability, enabling it to detect abusive interactions in IoS applications that involve sense-based communication.
2Measurement precision
If payload-based analysis is applied to sensory data, then detection accuracy for abusive interactions is improved, but computational complexity increases
Solution Approach 1:
The analysis process is segmented into two distinct approaches: payload-based analysis applied to sensory data and context data, and packet-based analysis applied to network data. This segmentation allows the system to tackle complex detection tasks through divided, specialized analysis paths, improving detection accuracy while managing computational complexity through structured division of labor.
Solution Approach 2:
The network node serves as an intermediary that receives sensory data from the sender node and performs comprehensive analysis before transmitting results to the receiver node. This intermediary role centralizes the computationally intensive payload-based and packet-based analyses, enabling detailed detection accuracy while isolating computational complexity to a specific system component rather than distributing it across all nodes.
3Measurement precision
If packet-based analysis is applied to network data, then detection of traffic patterns is improved, but processing time increases
Solution Approach 1:
The system dynamically applies packet-based analysis to network data to detect traffic patterns associated with cyberbullying. By making the analysis adaptive to the data type (packet-based for network data, payload-based for sensory data), the system optimizes processing efficiency while maintaining precise traffic pattern detection capability.
Solution Approach 2:
The invention merges payload-based analysis and packet-based analysis into a unified anomaly detection framework at the network node. Both analysis methods operate simultaneously on different data types and their results are combined to form the final anomaly determination, enabling comprehensive traffic pattern detection while balancing processing time through parallel execution of complementary analysis techniques.
Data Source
AI summary
A technique for detecting security attacks in Internet of Senses (IoS) applications is disclosed. A method implementation of the technique is performed by a network node (300) and comprises obtaining (S201B), from a sender node (200), sensory data (101) and at least one of network data (102) or context data (103). The method further comprises determining (S301) anomaly by applying at least one of a payload-based analysis or a packet-based analysis. The payload-based analysis is applied to at least one of the sensory data (101) or to the context data (103). The packet-based analysis is applied to the network data (102). The payload-based analysis analyzes a payload of data packets, while the packet-based analysis analyzes a pattern of data packet traffic. The method further comprises transmitting (S302), to a receiver node (400), a result of the determining step (S301).