[Paper Review] A note on blind contact tracing at scale with applications to the COVID-19 pandemic

Jack K. Fitzsimons, Atul Mantri+3 more|arXiv (Cornell University)|Apr 10, 2020

Privacy-Preserving Technologies in Data11 references4 citations

TL;DR

This paper proposes a cryptographically secure, scalable framework for blind contact tracing using secure two-party computation (2PC) to identify close contacts between individuals based on location data without exposing private location histories. By adapting an equality-testing protocol from Ishai et al. and applying tessellation-based proximity checks, the method enables private, efficient, and privacy-preserving contact tracing at national scale using existing cloud infrastructure, with demonstrated feasibility for countries like Ireland, Poland, and Japan.

ABSTRACT

The current COVID-19 pandemic highlights the utility of contact tracing, when combined with case isolation and social distancing, as an important tool for mitigating the spread of a disease [1]. Contact tracing provides a mechanism of identifying individuals with a high likelihood of previous exposure to a contagious disease, allowing additional precautions to be put in place to prevent continued transmission. Here we consider a cryptographic approach to contact tracing based on secure two-party computation (2PC). We begin by considering the problem of comparing a set of location histories held by two parties to determine whether they have come within some threshold distance while at the same time maintaining the privacy of the location histories. We propose a solution to this problem using pre-shared keys, adapted from an equality testing protocol due to Ishai et al [2]. We discuss how this protocol can be used to maintain privacy within practical contact tracing scenarios, including both app-based approaches and approaches which leverage location history held by telecoms and internet service providers. We examine the efficiency of this approach and show that existing infrastructure is sufficient to support anonymised contact tracing at a national level.

Motivation & Objective

• To address privacy concerns in large-scale contact tracing during the COVID-19 pandemic by enabling secure comparison of location data without revealing sensitive information.
• To develop a scalable, cryptographically secure solution that works across diverse data sources, including GPS, cell-tower triangulation, and Bluetooth, without requiring smartphone app installations.
• To demonstrate that the proposed protocol can be efficiently implemented using existing cloud computing resources, supporting national-level deployment.
• To enable collaboration between governments and telecom/IT providers by allowing private comparison of location databases while preserving data confidentiality.
• To provide a practical, information-theoretically secure framework for proximity testing that hides both location histories and exact timestamps of matches.

Proposed method

• The protocol reduces private proximity testing to a series of secure equality tests using pre-shared keys, adapted from Ishai et al.'s perfectly secure 2PC equality protocol.
• It employs tessellation of the Earth's surface into grid cells to discretize location data, enabling efficient and private comparison of proximity within a threshold distance.
• The system uses time quantization into 20-minute intervals to obscure exact matching timestamps and prevent re-identification.
• Location coordinates are scaled to kilometers using latitude and longitude conversion constants based on the WGS84 ellipsoid model for accurate distance computation.
• The protocol is designed to run in parallel across cloud containers, leveraging virtual private clouds and firewalls to isolate sensitive data.
• Implementation is supported by a lightweight Python library and a key generation service available to the public upon request.

Experimental results

Research questions

• RQ1 Can secure two-party computation be used to detect close contacts between individuals without revealing their private location histories?
• RQ2 Is it feasible to implement private proximity testing at a national scale using existing cloud infrastructure and cryptographic techniques?
• RQ3 How can multiple geolocation sources (GPS, cell-tower, Bluetooth) be securely integrated into a unified contact tracing framework?
• RQ4 What is the communication and storage overhead of such a system when applied to populations of varying sizes?
• RQ5 Can the protocol preserve privacy not only of location data but also of matching timestamps and data access patterns?

Key findings

• The protocol achieves information-theoretic security with optimal communication complexity, ensuring no leakage of private data during proximity checks.
• For a small nation like Ireland (population 4.8M), the system requires approximately 610 TB of keys and 850 TB of communication overhead for a two-week tracing window.
• For a large nation like Japan (population 133.9M), the required key size is about 11,000 TB and communication overhead is 16,000 TB, both within feasible limits of modern cloud infrastructure.
• The system remains scalable and efficient when parallelized across cloud instances, with communication rates of up to 100 Gb/s available on platforms like Oracle Cloud and AWS.
• The approach supports historical data from diverse sources such as telecom providers and GPS logs, enabling contact tracing without requiring real-time app-based data collection.
• A lightweight Python library and key generation service are publicly available for deployment and community use, facilitating real-world adoption.