Enhancing Wireless Data Transmission: A GAN-based Approach for Time Series Data Restoration

Recent technological advancements have led to a significant increase in the use of wireless communication for data transmission, particularly in sensor networks and device-to-device communications, necessitating not only rapid but also reliable data exchange. In conventional wireless communication, if a packet is lost in transmission due to retransmission, it is retransmitted. In this case, the subsequent packets will wait for them so that the order of the packets is not reversed, which is called Head of Line (HOL) Blocking. This problem can lead to overall transmission delays and losses, and is one of the reasons for inconsistent data transfer times. In this paper, we propose a technique to significantly reduce the retransmission rate of the sender and data loss by retransmission using Generative Adversarial Net (GAN). To evaluate the performance of this study, we applied our technique to a portion of SolarCube solar data assuming that it was corrupted, such as NaN processing, value change, and sign change, and the average recovery rate was 92.36%, demonstrating that it is possible to detect and recover losses in the data transmission process. This research is expected to have applications in data loss detection and recovery during transmission of time series data.