Study On Privacy-preserving Data Aggregation Protocols Based On Paillier Homomorphic Encryption In IoT

As the development of the information technology,the Internet of Things(IoT)has been widely used to upgrade of traditional industrial and transformation of social lifestyles.Edge computing-enhanced IoT data aggregation can make full use of the computing resource of edge equipment and save communication bandwidth,so it has been one of the most fascinating areas of IoT.As IoT devices are usually deployed in complex areas,they are vulnerable to data eavesdropping attacks,resulting in data leakage.Therefore,cryptographic technology should be used to achieve security of data transmission.Meanwhile,since IoT has resource-constrained devices,it is particularly necessary to improve the encryption and aggregation efficiency.In this thesis,we focus on the research of efficient and secure edge computing-enhanced IoT data aggregation scheme.On the one hand,we propose two IoT data aggregation schemes based on the Paillier cryptosystem.On the other hand,we propose a modification of the Paillier algorithm to improve the performance.The details are described as follows.1)Propose a privacy-preserving multi-dimensional data aggregation scheme for IoT system.According to the characteristics of the generation of the IoT data with little change in each time interval,the scheme uses the precomputed look-up table to improve the computing efficiency.Therefore,a part of modular exponent operation of the Paillier encryption can be implemented by several modular multiplications.The scheme merges multi-dimensional data into a large integer,to achieve batch encryption and decryption.Meanwhile,to avoid data overflow problems in the sum operation,a few zeros are concatenated between two adjacent dimensional data.To improve the security,the scheme adds random number at the high memory address of the merged data.When the number of data dimension is 8,the performance evaluation shows that the scheme gets approximate 50%encryption performance improvement while the same communication load compared with the existing dimensional scheme EMSA and reduces 7/8 of ciphertext communication load compared with the existing dimensional scheme MMDA.2)Propose a privacy-preserving statistical aggregation scheme for IoT system.The scheme supports multiple statistical aggregation functions,including arithmetic mean,quadratic mean,weighted mean and variance.In the scheme,the mean values can be calculated by the edge device and control center cooperatively.Firstly,the edge device computes the mean value in ciphertext since it only has the ciphertext of the aggregated data.Secondly,after receiving the mean in ciphertext,control center calculates the correct mean by using the modified extended Euclidean algorithm to process the decrypted mean.As a result,the scheme avoids revealing the total number of IoT device to control center during the whole process.The scheme utilizes three modified Montgomery exponentiation algorithms to improve the aggregation efficiency in the edge device.The performance evaluation shows that,compared with the best existing statistical aggregation scheme,the scheme gets 62.5%aggregation performance improvement for 1024 bits modulus,and 50%and 33%decrease of communication overload on arithmetic mean and variance statistics respectively.3)Propose an efficient modification of the Paillier cryptosystem.We develop practical applications of Hensel lifting in the Paillier cryptosystem to accelerate the decryption computation.For the modular exponentiation of decryption process,the proposed algorithm calculates the M mod p and then lifts it to M mod pk.The simulation shows that,compared with the traditional CRT decryption process,the proposed algorithm gets 29.9%efficiency improvement while using short parameters.