Privacy in Distributed Monitoring

In many emerging applications, such as real-time traffic monitoring, financial analysis, sensor network monitoring an important task is the continuos monitoring of stream data. In these contexts where large amount of data arrive continually the data processing requires to access often valuable personal information. As a consequence, the entire monitoring process could put at risk the privacy of people represented in the stream data. In this paper, we study the privacy issues in distributed systems during the monitoring of thresholds functions, where several nodes contribute with their data to the monitoring of a specific event. We provide a privacy-preserving framework suitable to find an acceptable trade-off among privacy protection, data quality and system performance. Using real-life data from GPS devices of private cars, we demonstrate the effectiveness of our approach in a case study consisting of the monitoring of customers mobility behaviors; in other words, we show how techniques for efficient communication can be used while preserving the individual privacy of the actors who are participating to the collective analysis.