Research On Data Aggregation Method For Detecting False Injection In Edge Computing

An essential feature of edge devices is that the scale of data generated is huge,and the communication capabilities and computing capabilities of edge devices are limited.Therefore,to improve transmission efficiency in edge computing,data aggregation methods are generally used.However,due to the harsh environment of edge devices,they are more vulnerable to malicious attacks,such as Fake Data Injection(Fake Data Injection,FDI)attacks.FDI attacks will maliciously tamper with data,resulting in a waste of communication resources and affecting aggregation results' accuracy.Cause the control center to make wrong decisions.Therefore,it is of great significance to ensure data privacy and aggregation accuracy in edge scenes.Based on this problem,this thesis focuses on the privacy protection of data in edge scenarios and the aggregation method that can detect false injections.The specific content is as follows:(1)A false data detection method suitable for edge environments is proposed.This method mainly uses a joint signature algorithm based on shamir sharing and a dynamic reputation value update algorithm.First,the data verification in this method uses a common signature algorithm based on shamir sharing.The attacker compromises some devices for false data injection and will be intercepted by the edge server.Only by compromising multiple deployed devices can the correct signature authentication be constructed.Secondly,to improve the filtering performance,this paper proposes a dynamically updated reputation value algorithm.The edge server will record all devices.The reputation value update depends on the device's historical reputation value and the current and historical data.Based on this update algorithm's characteristics,the deviation value isolates malicious devices based on the rapid decline in reputation value when devices are frequently attacked.Then,the homomorphic encryption algorithm is used to aggregate the data to ensure privacy of the data.When the data is generated,the data is encrypted so that the data always exists in the form of ciphertext during the entire transmission process.The aggregation scheme proposed in this paper can protect data privacy and resist false data injection attacks,especially FDI attacks launched by multiple compromised devices.It also can detect and isolate malicious devices and tolerate device faults.Finally,the scheme's storage performance,filtering performance,and aggregation accuracy are verified through simulation experiments and compared with similar schemes.The experiment shows that this scheme has better filtering performance against the false injection attacks launched by compromising nodes,Compared with the existing privacy protection type aggregation scheme,this scheme has higher aggregation accuracy.(2)Apply the scheme mentioned above to smart transportation scenarios,design and implement a traffic flow analysis system that can detect false data in smart transportation.First,the system monitors the traffic area in the city and can be obtained through data aggregation.Real-time traffic flows in each area,and analyze traffic congestion through traffic flow information.Secondly,the system can detect and filter false data to ensure the availability of collected data.The system performs statistical analysis on false injection attacks in the monitoring area and evaluates the equipment's security status according to the equipment's attack situation.When the equipment is frequently attacked,Give an early warning so that the traffic center can respond.Finally,the system maintains the reputation value of each device.Analyzing the device's reputation value change curve can isolate malicious devices that frequently launch FDI attacks and devices that have errors(offline).