A Kind of Efficient Data Archiving Method for Historical Sensor Data

Nowadays there are more than 10 billion sensors in the Internet of Things. These sensors will produce a mass of historical data as time goes on. And these data are useful in data mining and may create great value for human. However, historical sensor data are hard to archive because they have the characteristics of mass, complexity, isomerism and dynamic nature. Therefore, how to archive historical sensor data to storage efficiently is a meaningful issue deserved to study. In this paper, an efficient data archiving method is proposed. It mainly focuses on three aspects: memory allocation strategy, data compression technique and data storage structure. This method not only considers the efficiency of each aspect but also the combination of each aspect in order to ensure the high efficiency of whole data archiving method.