The Study Of Data Collection Privacy Protection In Smart Cities

Smart cities acquire and transmit information through the ubiquitous Internet of Things,process massive amounts of real-time data through cloud computing,and feedback the processing results to the control system to achieve intelligent and automated control,ultimately bringing the city to a “smart” state.The development of technologies such as big data,cloud computing,and Internet of Things provides effective technical means for data exchange,sharing,analysis,and contributes to the security,science,and effcient operation of smart cities.Smart cities use advanced information technology and communication technology to achieve efficient operation of urban intelligence.In the course of operation,massive multi-source heterogeneous data will be generated.The amount of data is huge and the growth rate is fast.Usually,these data contain a large amount of sensitive information of individuals or organizations,so in the process of data collection,there is a serious risk of privacy leakage.The collected data may threaten national,social and personal security.Due to the dynamic nature,rapid growth and strong real-time performance of urban data,the traditional data collection privacy protection method is no longer applicable.Based on the characteristics of urban big data,this paper proposes three privacy protection algorithms for urban big data collection based on MapReduce model.First,the k-anonymity algorithm based on the MapReduce model uses the MapReduce model to process a large number of data sets,and divides the equivalence classes in the process,so that the number of tuples in each equivalence class is k and completely is indistinguished from other k-1 records.Second,based on the MapReduce model's single sensitive(?,k)-anonymous algorithm,since the k-anonymous model does not impose any constraints on sensitiveattributes,a single sensitive(?,k)-anonymous model is used to ensure that the published data meets k-anonymity and in an equivalence class,the percentage of data records associated with any sensitive attribute value is not higher ?.Third,based on the MapReduce model's multi-sensitive(?,k)-anonymous algorithm,the anonymous constraint of the single sensitive(?,k)-anonymous model faces a specified sensitive value,so the model is not applicable to the privacy protection research of multiple sensitive attributes.In order to extend the anonymous constraint to all the values of sensitive attributes,a multi-sensitive(?,k)-anonymous model based on the MapReduce is proposed to protect each sensitive attribute in the data set.The experimental results and theoretical analysis show that these three algorithms can not only effectively protect data privacy but also have the characteristics of small information loss and low execution time.