Research On Privacy Preserving In Internet Of Things Based On Fog Computing

With the development of the Internet of Things,more and more devices will be connected to the Internet of Things.Massive data will also be generated,which possess massive computing resources,storage space and communication bandwidth.At the same time,faster and faster real-time response is required.Traditional cloud computing frameworks have gradually been unable to meet these requirements.As an extension of cloud computing,fog computing was introduced.Fog computing is deployed at the edge of the network,which is close to Io T devices and users.It can provide real-time services without submitting some tasks to the cloud,which can effectively reduce the pressure on the cloud and achieve lower communication consumption.For most of the existing Io T applications,in order to realize real-time monitoring and intelligent control,the cloud control center needs to continuously collect data from the Io T devices to perform operations such as calculation and analysis,the collected data often contains some private information.How to perform secure data aggregation without revealing privacy has become a challenge.However,due to some problems such as faulttolerant,only supporting sum aggregation and differential privacy issues,the existing privacy-preserving data aggregation schemes are not yet well applied in actual scenarios.In view of the security and privacy issues faced by data aggregation in the Io T applications,the main work is as follows:(1)An efficient privacy protection data aggregation scheme in a smart grid based on fog computing is propoesd.First,a three-tier network architecture of cloud-fog-smart meter is proposed based on fog computing.By using fog computing,the operation of aggregation will be carried out in the fog node in advance,which effectively reduces the computing pressure of the cloud control center and communication consumption.Then the electricity data counted by smart meters is encrypted by using additive El Gamal homomorphic encryption,which is more efficient than the Paillier algorithm which used in most existing schemes.At the same time,using the message authentication code(MAC)to ensure the integrity and authentication of the data,which can successfully resist false data injection attacks and replay attacks.Moreover,this scheme is fault-tolerant.Even if some smart meters fail to send electricity data to the fog node,the fog node and the control center can successfully aggregate and count the remaining electricity data submitted for normal operation.Then,by adding noise to the data to realize the differential privacy,whicg can successfully resist the adversary’s differential attack.Finally,simulation experiments show that compared with the existing aggregation scheme,this scheme can meet the security and privacy requirements,and reduce the computational consumption and communication overhead.(2)A data aggregation scheme that supports multiple functions of privacy protection in the Industrial Internet of Things based on fog computing is proposed.Most of the existing privacy protection data aggregation schemes can only calculate the sum,while some common aggregation functions such as variance and inner product cannot be realized.In this scheme,multiple aggregation functions are realized.First,fog computing is introduced and a layered network framework is proposed.Then the use of BGV homomorphic encryption allows the fog node to calculate multiple aggregate functions without leaking privacy,such as mean,variance and one-way ANOVA.At the same time,the message authentication code is also used to ensure the security of the aggregation scheme.Finally,the simulation experiment shows that this scheme has a smaller computational cost.