IJCATR Volume 9 Issue 12

Optimization of Missing Value Data Imputation Automatic Dependent Surveillance Broadcasting (ADS-B) Based on K-Nearest Neighbor and Genetic Algorithm

Didik Hariyanto , Sholeh Hadi Pramono , Erni Yudaningtyas
10.7753/IJCATR0912.1003
keywords : ADS-B, Missing Value, Imputation, K-Nearest Neighbor, Genetic Algotirhm

PDF
The flight navigation equipments technology use still conventional, namely using radar, now slowly starting to switch to Automatic Dependent Surveillance-Broadcast (ADS-B [6]. In this study, using RTL-SDR to detect aircraft and carry out tests through the Monte Carlo alltitude method, latitude, and longitude only [3]. However, in this system there is a problem regarding the missing value in the preprocessed data results / ADS-B flow data. In handling missing values, the KNN method is the most popular, but the weakness in the KNN method, can reduce the performance[9]. So a Genetic Algorithm (GA) is proposed to optimize the k value in the KNN method. The results of this study obtained a better MSE value in the imputation process. Altitude k = 3, with MSE 128668.96, Speed k = 6, with the MSE value = 457.5201, while the k value in the Heading variable k = 61 with MSE = 752.1429. For Lattitude and Longitude, the value of k = 3, MSE 9.16E-05 and k = 2 and MSE 1.68E-05.
@artical{d9122020ijcatr09121003,
Title = "Optimization of Missing Value Data Imputation Automatic Dependent Surveillance Broadcasting (ADS-B) Based on K-Nearest Neighbor and Genetic Algorithm",
Journal ="International Journal of Computer Applications Technology and Research(IJCATR)",
Volume = "9",
Issue ="12",
Pages ="327 - 331",
Year = "2020",
Authors ="Didik Hariyanto , Sholeh Hadi Pramono , Erni Yudaningtyas "}
  • ADSB data is very important for the development of the world of aviation for human safety
  • This study proposes the use of SDR technology for aviation
  • The results of the implementation of SDR in this study require an imputation method to correct missing values
  • Performance in the Imputation method using MSE.