VoIP Spam Classification via Packet Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Voice over Internet Protocol (VoIP) systems are vulnerable to attacks like Stealth Denial of Service, Stealth Distributed Denial of Service, and Voice/Voice Mail Spam, which can disable user devices and entire systems, and current security measures are inadequate to address these specific threats.
Innovation Solution
A system and method for automatically classifying voice communications as spam or legitimate signals using machine learning models, such as decision trees and neural networks, to identify machine-generated speech and commercial advertisements, and executing actions like blocking or deleting such signals, with the ability to update classification models and provide trust scores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If VoIP systems are deployed to provide voice communications, then productivity and efficiency are improved, but vulnerability to attacks like Stealth DoS, Stealth DDoS, and Voice Spam increases
Solution Approach 1:
The system performs preliminary classification of voice communications before they reach user devices. By analyzing voice packets and extracting features (energy, pitch, statistical properties) in advance, the system can identify spam and DoS attempts before they impact users, enabling preventive action rather than reactive response
Solution Approach 2:
The invention introduces an intermediary classification system between the VoIP network and user devices. This intermediary component analyzes voice packets, extracts features, and classifies communications as legitimate or malicious, acting as a buffer that protects users from attacks without disrupting legitimate voice services
2Reliability
If spam detection systems are implemented to block malicious calls, then security is improved, but false blocking of legitimate calls may occur
Solution Approach 1:
The system changes parameters for analyzing voice communications by extracting multiple features (energy, pitch, statistical properties) and using machine learning models that consider multiple parameters simultaneously. This multi-parameter approach enables more accurate discrimination between spam and legitimate calls, reducing false positives while maintaining high security
3Speed
If voice communication packets are processed in real-time for classification, then response time is improved, but system complexity increases
Solution Approach 1:
The system segments the classification process into distinct stages: packet reception, feature extraction (energy, pitch, statistical properties), model analysis, and classification decision. This segmentation allows each component to be optimized independently and enables real-time processing by breaking down the complex task into manageable, sequential operations
Data Source
AI summary
The present invention provides a system, method and apparatus for automatically classifying voice communications, such as voice messages and phone calls in prerecorded voicemails (one speaker) and two-way conversations, as either spam or legitimate signals in a communications system (e.g., SIP, IMS, UMA, etc.). More specifically, the present invention classifies a voice communication session by receiving one or more voice communication packets associated with the voice communication session, extracting one or more properties from the received voice communication packets and classifying the voice communication session based on the extracted properties. The present invention can also be implemented as a computer program embodied on a computer readable medium wherein each step is performed by one or more code segments.