A distributed protocol for privacy preserving aggregation with non-permanent participants.

Recent advances in techniques that combine and analyze data collected from multiple partners led to many new promising distributed collaborative applications. Such collaborative computations could occur between trusted partners, between partially trusted partners, or between competitors. Therefore preserving privacy is an important issue in this context. This paper presents a distributed protocol for privacy-preserving aggregation to enable computing a class of aggregation functions that can be expressed as Abelian group. The proposed protocol is based on an overlay structure that enables secret sharing without the need of any central authority or heavyweight cryptography. It preserves data privacy such that participant data is only known to their owner with a given probability. The aggregation result is computed by participants themselves without interacting with a specific aggregator. The aggregation result is accurate when there is no data loss. A strategy to handle the problem of