Classifier-Based Message Routing for Roaming SMS Security
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The reliability of visited network components, such as MSC/VLR, in telecommunications networks is questionable, especially when subscribers roam, making them vulnerable to malicious interceptions, as the home network lacks visibility and cannot independently verify the authenticity of responses, allowing malicious networks to feign location updates and intercept SMS messages intended for other networks.
Innovation Solution
A conditional message routing method is implemented in the home network's SMSC, which assesses the reliability of visited network components by using a classifier trained on previous communications data, evaluating parameters like encryption responses, message volumes, and traffic patterns to determine the reliability state and decide on message forwarding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the home network forwards messages to visited network components during roaming, then message delivery capability is improved, but security against malicious interception deteriorates
Solution Approach 1:
The system performs preliminary reliability assessment of visited network components before forwarding messages. The classifier evaluates historical communication data, encryption response correctness, and traffic pattern consistency to predict reliability in advance, allowing the home network to make informed routing decisions that balance delivery capability with security protection.
Solution Approach 2:
The reliability classifier acts as an intermediary between the home network and visited network components. It mediates the message forwarding decision by evaluating the trustworthiness of visited network components based on multiple parameters including encryption responses and traffic patterns, thereby protecting the home network from malicious interception while maintaining message delivery.
2Reliability
If the home network implements reliability assessment of visited network components, then security is improved, but system complexity deteriorates
Solution Approach 1:
The visited network components themselves provide the reliability assessment data by responding to authentication requests and communicating their own operational status. The classifier uses self-reported information from visited network components combined with historical data, reducing the need for complex external verification mechanisms while maintaining reliable security assessment.
Solution Approach 2:
The patent replaces complex manual or rule-based security verification mechanisms with a machine learning classifier. The classifier automatically evaluates multiple parameters including encryption responses, message volumes, and traffic patterns to determine reliability, substituting mechanical verification processes with intelligent automated assessment that scales more efficiently.
3Measurement precision
If the classifier uses multiple parameters for reliability assessment, then detection accuracy is improved, but computational complexity deteriorates
Solution Approach 1:
The classifier implements a tiered assessment approach where it evaluates multiple parameters including encryption responses, message volumes, and traffic patterns, but focuses computational resources on the most discriminative features. By assessing more parameters than strictly necessary (excessive action), the system achieves high detection accuracy while the classifier's learning process identifies the optimal subset of features for decision-making.
Data Source
AI summary
A computer implemented method of selective transmission of a communications message from a home telecommunications network to a visited telecommunications network for delivery to a subscriber device roaming via the visited network includes receiving the message at the home network; identifying a communications component of the visited network to which the messages is to be routed; defining an input vector for a classifier, the input vector encoding parameters of the communications component; executing the classifier to classify the communications component to determine a reliability state of the communications component; and forwarding the message to the communications component of the visited network for delivery to the subscriber device in dependence on the classified reliability state.