Exact Multi-Task Aggregation with Confidential Queries and Designated Recovery

In the field of IoT healthcare, multiple data users may request aggregate statistics from the same set of data owners simultaneously. Most existing privacy-preserving aggregation schemes are designed for a single, fixed task and often expose the task definition or participation decisions for individual tasks, while offering limited support for lightweight result verification. This paper therefore proposes a verifiable, privacy-preserving, multi-task data aggregation scheme that is suitable for IoT healthcare. Each task uses symmetric encryption with keys distributed to authorised data owners via multicast encapsulation based on the Chinese Remainder Theorem. This enables the cloud to distribute tasks without knowing their content. To support precise multi-task aggregation, each data owner uses the Chinese Remainder Theorem to package slot-grouped data and authentication tags into a plaintext message and upload a ciphertext. Participation decisions for each task are protected through additive sharing between two non-colluding aggregator entities, rendering non-participation indistinguishable from zero contribution. Threshold-based validity checks suppress the output of tasks with insufficient contributors, while lightweight homomorphic authentication tags allow designated data users to verify the integrity of the reconstructed results. Security analysis shows that under the honest-but-curious model, the scheme ensures data confidentiality, decision privacy, task privacy, result confidentiality and receiver verifiability. Experimental results indicate that the scheme is suitable for medical IoT scenarios, introducing only moderate overhead while maintaining low costs at the aggregation end.