Supporting privacy of computations in mobile big data systems

摘要: Cloud computing systems enable clients to rent and share resources of third party platforms, have gained widespread use in recent years. Numerous varieties mobile, small-scale devices such as smartphones, red e-health devices, etc., across users, are connected one another through the massive internetwork vastly powerful servers on cloud. While mobile store “private information” users location, payment, health data, they may also contribute “semi-public (which include crowdsourced data transit, traffic, nearby points interests, etc.) for analytics. In a scenario, device seek obtain result computation, which depend its private inputs, from other and/or any “public inputs” Internet. We demonstrate new method delegating real-world computations resource-constrained using an encrypted program known garbled circuit. Using version client’s server cloud executes circuit returns resulting outputs. Our system assures privacy input output enables client verify that evaluator actually performed computation. analyze complexity our system. measure time taken construct well evaluate it varying number servers. we practical, preserving search application locates nearest point interest feasibility.