Query Optimization for Room Temperature Sensor using Principal Component Analysis

In the Internet of Things (IoT), a lot of data is sent to the cloud to represent what happens to the device. To transmit a lot of data, it takes a lot of energy to do that activity. Therefore that, we need a method to save energy, namely by using query optimization. This method is a process to analyze a query to determine what resources are required by the query and to identify if the use of these resources can be optimized by reducing the amount of data without significantly changing the output. In this paper, the authors propose a query optimization by using the Principal Component Analysis (PCA) method for save energy on the device. First, the temperature sensors collect data from a room. Then the data collected from the temperature sensor is used to generate a matrix for PCA. After the data is made into a matrix, a covariance matrix is decomposed to obtain the eigenvalues and eigenvectors. After getting the eigenvalues and eigenvectors, a new variable is determined, namely the Principal Component (PC) value to determine how many data will be reduced. The experimental results show that the NMSE value is not too large despite having 3 dimensionality reduced. Therefore, it is expected for PCA to be used for query optimization effectively.