Privacy-Preserving Proximity Detection through Haversine Distance and Geo-Hash

Abstract

Proximity implies computing suitable distance between nearby people and places. Proximity detection is very crucial to many multimedia applications. For instance, fatal incidents occur near forests if animals like elephants walk into nearby crowded regions, resulting in loss of life and property. During the COVID-19 pandemic, the infection spread globally due to a lack of proper proximity detection strategies. Many solutions were proposed based on Bluetooth, WiFi, wearable sensors, ultrasound, and GPS co-location to detect the proximity. However, privacy-preserving solutions were limited, which hindered the privacy-aware society. In this paper, we propose two privacy-preserving proximity detection frameworks called Privacy-Preserving Proximity through Haversine Distance (P3-HD) and Privacy-Preserving Proximity through Geohash (P3-GH). With the help of available GPS data and the traditional Haversine formula, P3-HD computes the distance to detect if the user is in the proximity zone. The other framework, P3- GH detects the proximity on the top of the encrypted Geohashes. These proposed frameworks secure the data of the authentic user from any unauthorized access. GPS data may introduce anomalies that lead to false results. We have proposed a secure framework to detect the anomalies called SADHE, a secure method for detecting anomalies in GPS trajectory without compromising users� location. We compare various primacy and deficiency against potential threats, specifically replay attacks, poison attacks, frequency analysis attacks, and man-in-the-middle attacks, and validate the robustness of the proposed frameworks.