An end-to-end encrypted domain proximity recommendation system using secret sharing homomorphic cryptography

Location privacy preservation is where a person´s location is revealed to other entities, such as a service provider or the person´s friends, only if this release is strictly necessary and authorized by the person. This is especially important for location-based services. Other current systems use only a 2D geometric model. We develop 3D geometric location privacy for a service that alerts people of nearby friends. Using a robust encryption algorithm, our location privacy scheme guarantees that users can protect their exact location but still be alerted if and only if the service or friend is nearby and to then determine whether they are getting closer. This is in contrast to other non-secure systems, systems that lack secret sharing, and systems that use location cloaking. In our system, such proximity information can be reconstructed only when a sufficient number of shared keys are combined together; individual shared keys are of no use on their own. The proposed ring homomorphism cryptography combines secret keys from each user to compute relative distances from the encrypted user´s location end. Our secret sharing scheme doesn´t allow anyone to deceive, mislead, or defraud others of their rights, or to gain an unfair advantage. This relative distance is computed entirely in the encryption domain and is based on the philosophy that everyone has the same right to privacy. We also propose a novel protocol to provide personal anonymity for users of the system. Experiments show that the proposed scheme offers secure, accurate, fast, and anonymous privacy-preserving proximity information. This new approach can potentially be applied to various location-based computing environments.