Abstract-Cloud computing becomes essential in these days for the enterprises. Most of the large companies are moving their services and data to the cloud servers which offer flexibility and efficiency. Data owner (DO) hires a cloud service provider (CSP) to store its data and carry out the related computation. The query owner (QO) sends a request which is crucial for its future plans to the CSP. The CSP computes all necessary calculations and returns the result back to the QO. Neither the data nor query owners want to reveal their private data to anyone. k-Nearest Neighbour (k-NN) interpolation is one of the essential algorithms to produce a prediction value for an unmeasured location. Simply, it finds k number of nearest neighbours around the query point to produce an output. Oblivious RAM (ORAM) has been used to protect the privacy in cloud computing. In our work, we will perform the k-NN method using the kd-tree and ORAM without revealing both the data-owner's and query owner's confidential data to each other or to third parties. The proposed solution will be analysed to ensure that it provides accurate and reliable predictions while preserving the privacy of all parties.

One of the most emerging computer technologies of this decade is cloud computing that allows data owners to outsource their storage and computing requirements. It enables data owners to avoid the costs of building and maintaining a private storage infrastructure. While outsourcing data to cloud promises significant benefits, it possesses substantial security and privacy concerns, especially when data stored in the cloud is sensitive and confidential, like a business plan. Encrypting the data before outsourcing can ensure privacy. However, it will be very difficult to process the cipher text created by the traditional encryption method. Considering this fact, we propose an efficient protocol that allows a query owner to retrieve the interpolation of the top k records from two different databases that are closest to a query point. Note that the databases are stored in two different cloud service providers in encrypted form. We also show that the proposed protocol ensures the privacy and the security of the data and the query point.