Extremely Efficient and Privacy-Preserving MAX/MIN Protocol Based on Multiparty Computation in Big Data

In the era of the Internet of Things, numerous devices generate a tremendous amount of data and the cloud-based big data analytics efficiently extracts useful information from it. However, using the outsourced cloud comes with privacy concerns, such as data being exposed. Multiparty computation is a powerful scheme that provides data privacy. The target of this work is to propose a privacy-preserving protocol based on multiparty computation that is extremely efficient in finding the maximum/minimum values in big data. Since existing maximum/minimum protocols call as many relatively expensive comparison protocols as the volume of data, the execution time for big data rapidly increases in proportion to the volume of data. The existing works were limited, as they could only show experiment results for a small dataset. However, since the proposed protocol calls more efficient equality protocols relative to the data length-which is much smaller than the volume of big data-it is very efficient for large amounts of data. For a million data that have not been attempted in existing works, the secure version of our privacy-preserving maximum/minimum protocol took 76.5 seconds and the efficient version took 9.73 seconds. Moreover, since the proposed protocol allows parallel computation, its execution time can be greatly reduced when it is executed in the cloud that enables numerous parallel computations.