Abstract:
User identification using the spatio-temporal geometries or trajectories has always been the interest of many researchers and is a very specific topic in fields related to data. Nowadays large spatio-temporal data are collected using different techniques including the smartphones GPS. However, one of the major concerned issue regarding the popularity of GPS-based devices and systems is large scalability of the personal location information (that is often highly dimensional) generated by them and the sharing of that massive data with applications or identity systems. Traditional user verification techniques usually separately consider spatial and temporal approaches. Although there have also some work that has been done to integrate both the spatial and temporal information for user identity prediction but most of them suffer from the overfitting problem because of the large number of spatio-temporal trajectory patterns. Blockchain technology recently introduced in several areas after the successful working in the domain of crypto-currencies. Advent of blockchain can help resolving the concern of large scalability of mobility data by its reliable storage capacity, immutability and decentralized trustless data processing features. We consider a spatio-temporal Blockchain that registers both time and location attributes of the users. In this research, we propose a novel approach to uniquely identify an individual by using its spatio-temporal fingerprints which are stored in blockchain. Fingerprint’s defines the spatial data points or the spatial trajectory of that respective individual. Proving this research, we developed a prototype blockchain system using the Hyperedge Fabric in which user spatio-temporal fingerprints are recorded on the basis of user mobile GPS data. And then we verify the user using the previously recorded data in blockchain and user-given location data to the system. Furthermore, we generate a new private key for the individual after the guaranteed verification steps satisfying our set threshold.