User localization through SMS serves as a low-cost, scalable alternative to conventional location tracking
Background:
User localization is traditionally achieved through GPS, network triangulation, or internet-assisted location sharing. GPS-based tracking depends on device permissions, while network-based methods require extensive infrastructure and user cooperation. Additionally, in low-connectivity environments or for security-sensitive applications, conventional techniques fail to provide efficient, non-intrusive, and scalable localization. Existing solutions often rely on explicit user consent or internet connectivity, limiting their feasibility in many real-world scenarios. Given the widespread reliance on SMS for communication, leveraging SMS signaling behavior presents a novel, underutilized opportunity for accurate and stealthy localization.
Technical Overview:
Northeastern researchers (in collaboration with researchers from NYU Abu Dhabi and TU Dortmund) have developed an innovative method for user localization by analyzing SMS delivery report timing variations. When an SMS is sent, it triggers an automatic delivery report from the recipient's device, containing valuable timing metrics that reflect network characteristics and user location. This system collects and processes multiple SMS timing data points, constructing unique timing fingerprints for various locations. A deep learning model is then trained on extensive datasets, allowing it to accurately predict user locations without requiring GPS, active internet access, or explicit consent. This method is adaptable to global mobile networks and ensures high-accuracy localization, even in challenging environments. By utilizing existing SMS infrastructure, this innovation bypasses the need for dedicated tracking mechanisms, offering a low-cost, scalable alternative to conventional location tracking.
Benefits:
Application:
Opportunity: