Investigating the Radiation Patterns of 2.4 GHz Wi-Fi GPON ONT used by Telkom Indonesia: A Comparative Study using Hidden Markov Model

To ensure fast and reliable internet service via Gigabit Passive Optical Network (GPON) Optical Network Terminal (ONT), users should select a GPON ONT model that best meets their specific requirements. Therefore, a study was designed to compare the radiation patterns of three types of GPON ONT, i.e. Dual Band Premium, Dual Band, and Single Band, using an ESP8266 device as a sensor to capture the Received Signal Strength Indicator (RSSI) value with Python language processing, then tested in an Anechoic Chamber. Besides its affordable price, ESP8266 has good performance which resulted in more accurate data capture. The captured data will be processed to form a radiation pattern and find the comparison pattern. In the process of forming radiation patterns, a Kalman Filter is used for smoothing data processing to reduce noise and increase the accuracy of the data obtained. In addition, the use of a Hidden Markov Model (HMM) is applied to determine the quality of radiation emission produced by each type of ONTs, as HMM can identify patterns in sequential data, thus providing a more in-depth and accurate analysis of signal variations. The premium type is used as a reference to compare the quality of Dual Band and Single Band type respectively. The results indicate that the GPON ONT Single Band model outperforms the Dual Band model in quality score, based on RSSI measurements using the ESP8266 device. The radiation pattern reveals that the Single Band model has a more uniform distribution of RSSI values, whereas Dual Band models exhibit higher RSSI values at specific angles only. Thus, users can select the Wi-Fi GPON ONT type that best aligns with their particular needs and environmental conditions