Edge computing-oriented federated learning indoor positioning privacy protection method

Abstract

The invention provides an edge computing-oriented federated learning indoor positioning privacy protection method. The method is based on federated learning and differential privacy protection technologies. According to the method, credible federation training of an indoor positioning model is carried out in an edge computing environment; in the training process, all participating users do not share training data, distributed training and credible aggregation of the indoor positioning model are carried out only by sharing positioning model parameters, meanwhile, the model parameters are updated in an end-cloud cooperation iteration mode, so that the indoor positioning model is continuously optimized, and privacy protection and cooperative benefits of multi-user positioning model training are achieved. Experimental results show that compared with a traditional centralized model training method and a federated learning-based model training method, the method provided by the invention notonly can provide proven privacy protection, but also ensures the positioning effect of the model under the condition of increasing extremely little calculation overhead.

Description

technical field [0001] The invention relates to the service field of indoor positioning, and relates to protecting the data privacy of the user when the user obtains the service by using the location. Background technique [0002] In the traditional cloud-centric computing method, the data collected by the mobile device will be all uploaded and stored in the cloud server for centralized computing and processing. However, with the rapid development of technologies and fields such as the Internet of Things, crowdsensing, and social networks, ubiquitous mobile devices and sensors continue to generate massive amounts of data, and hundreds of millions of users have a huge amount of interactions while enjoying Internet services. The explosive growth of data on the edge side has caused cloud computing to consume a lot of computing and storage resources when processing this data, and its capabilities will be stretched. Edge computing can migrate the pressure of cloud computing and ...

AI Technical Summary

Problems solved by technology

Secondly, the cloud server aggregates the model parameters of each edge device node for differential privacy protection, confuses the contribution of each edge node to the global model, and realizes the protection of the edge node model; patent CN110267197A proposes a WIFI fingerprint-based indoor positioning A lightweight privacy protection system and method. While providing services to users, the method uses encryption algorithms such as Paillier to perform encryption processing when data is uploaded and processed to achieve the effect of protecting user privacy; however, the method uses public key cryptography system, the calculation overhead is very large, and it is not suitable for the edge computing environment; the patent CN107222851A proposes a differential private protection method, while providing location services for users, the server uses the AP sequence uploaded by the client to construct a fingerprint data set, using differential privacy protection The technology performs disturbance processing in the clustering process to protect the user's location privacy and the server's data privacy

However, this method is based on the cloud server architecture, which is difficult to adapt to the exponential growth of the current terminal device data volume and the privacy of the model training process.

Images