Traffic Source Modelling in Interactive Virtual Reality

Virtual Reality (VR) stands as a pivotal technology merging computer science and human- computer interaction, rooted in immersive experiences. However, the analysis of large sets of traffic measurements in its development is challenged by high variability and limited datasets. This study aims to establish valuable data sets in VR development and analyze the data traf- fic to see the influence of nature of players and games during the gameplay. Sixty participants aged 18 to 35, having passed standard visual acuity and color blindness screenings, with diverse gaming backgrounds, were randomly divided into two groups. Each group, consisting of 30 individuals, engaged in VR experiences via Oculus Meta 3. Group 1 played Cooking Simulator and Beat Saber, while Group 2 experienced Forklift Simulator and Medal of Honor. Utilizing box plots, the data rate distribution reveals varying ranges among traces, with lowest data rates observed in two traces of Cooking Simulator and Beat Saber. Total Variation Distance (TVD) displayed the similarity between probability distributions, indicating higher TVD values for lower data rates, suggestive of concentrated data around a central point. Conversely, lower TVD values correlated with less centralized points in lower data rates. Furthermore, the detec- tion of gaps and small frames led to a left-skewed frame distribution, while the highest frame size reached is not more than around 320kb in all games. Despite each game having certain ways to guide the players to play in a specific way, TVD values show that there is a difference when same player playing different games and different players played same games.