Privacy protection publishing method of converged data stream in Internet of things

Abstract

The invention discloses a privacy protection publishing method of a converged data stream in an Internet of things. According to the method, firstly Laplacian noises are added to the data stream for privacy protection on the basis of a result of dimension partitioning, then dynamic grouping of the data stream is carried out on the basis of system threshold values of self-adaptive updating, and finally, noise smoothing processing is carried out according to noise data and a dynamic grouping result to obtain a data stream which can be directly published and has privacy protection. Compared withgeneral privacy protection methods, the method of the invention improves utility of publishing data through learning the dimensional correlation and temporal correlation of the multidimensional data stream, designs a self-adaptive threshold updating strategy and a dynamic grouping strategy on the basis of feedback errors, and guarantees practicality of the method in practice. The method of the invention realizes self-adaptive real-time publishing of the multidimensional data stream, a whole process is simple and easy to realize without the need for complicated encryption and decryption operations, calculation overheads are low, and a use value is high.

Description

technical field [0001] The invention belongs to the field of privacy protection, and in particular relates to a privacy protection release method for converging data streams in the Internet of Things. Background technique [0002] With the advent of the Internet of Things and the era of big data, the new mobile perception model based on mobile smart devices acquires and publishes massive data streams in the physical world through advanced comprehensive perception technology, thus greatly promoting the development of application services based on data perception It also greatly improves and facilitates people's daily life, such as traffic flow monitoring, disease monitoring and prevention, service recommendation, etc. However, the release of a large amount of data streams seriously exposes users' private information. Since sensors mostly interact with people, the data streams from sensors inherently contain a large amount of user-sensitive information. For example, health m...

AI Technical Summary

Problems solved by technology

At present, the existing mechanism is difficult to adaptively learn the dynamic characteristics of the data flow, and needs to rely on the parameters defined in advance, resulting in excessive noise addition and destroying the seat utility of the original data flow.

Images