Encrypted Cartographic Masking for Geospatial Privacy Protection in Location Based Services
ISBN 978-85-88783-11-9
Authors
1Clarke, K.
1UC SANTA BARBARA Email: kclarke@geog.ucsb.edu
Abstract
A new method of cartographic masking is presented with the goal of ensuring effective undecipherable, but reversible, encryption of sensitive geographic point locations. The method works by converting a coordinate set into one meter resolution grid references using the Military Grid Referencing System (MGRS) using open source software, and is global in application. It computes a set of spatial descriptors for the point distribution (nearest neighbor statistic, standard distance, etc.) and uses brute force computing methods to select an encryption schema that (a) ensures that no point lies at its true location and (b) the encrypted point distribution’s spatial description parameters are as close as possible to the actual values. An important input parameter is the root mean square distance of the desired encryption error, in increments of powers of tens of meters within a 100km MGRS coded square. An encrypted point data set can be separated from its key, and still be useful in aggregation, measurement, mapping and analysis. The key is encoded into a graphic file in GIF format. When this key and the inverse software are available, assumedly to a pre-selected set of permitted users, perfect decryption is possible. Having only one data file makes a variety of uses possible, since it can be widely distributed without risk of violating geoprivacy. Applications may be in health geography, protection of sensitive species distributions or archaeological records, or to protect personal privacy when using location based services. Future work will examine concurrent encryption of the coordinate points’ elevations and time stamps.
Keywords
point distribution; coordinate; encryption