Throughput Prediction for Multimedia IoT in Wireless Network
Rosa Eliviani, Yoanes Bandung · 2022
Wireless networks have been overgrown in recent years. In improving the quality of wireless networks, many studies have been conducted on many ways to improve the wireless network’s QoS (Quality of Service). The things that become the measurement of QoS are throughput, delay, jitter, and error rate. Predicting the throughput is an excellent way to improve the QoS. The result of throughput prediction helps to select the bitrate automatically and minimizes rebuffering. In this paper, we conduct a comparative study of throughput predictions for multimedia IoT in wireless networks. Then we experiment by collecting datasets, processing the data, and then predicting throughput data. We collect data from transmission data in the form of images from the ESP32-Cam to the Server. The throughput prediction uses Autoregressive (AR), Moving Average (MA), Autoregressive Integrated Moving Average, and (ARIMA). The result of this paper is AR algorithm has better performance in predicting throughput.