Will Edge Computing Enable Location-based Extended/Mixed Reality Mobile Gaming? Demystifying Trade-off of Execution Time vs. Energy Consumption

Abstract

The trailblazing development in mobile and wearable-based gaming dictates both the support of new technology enablers to allow for current demand and the development of modern computational offloading strategies to decrease the energy of handheld devices and maintain the energy emissions caused both by computation and transmission of data. Modern cellular networks already provide some support for proximity-based gaming, e.g., Ingress, PokemonGo, and The Witcher: Monster Slayer, among others. However, the demand of users is pushing the boundaries toward full-immersive Extended and Mixed Reality (XR/MR) experiences. Thus, computational offloading to the wireless network Edge becomes inevitable to keep the immersion high. This paper aims to analyze the impact of computational offloading (and, thus, execution time) on energy consumption. Computationally demanding games are analyzed for cases run locally, sent to a conventional remote server (cloud), offloaded to the user-owned more energy-independent device, or to the network edge. The results show that Edge computing operates the most efficiently regarding the trade-off between energy spent for execution vs. data transmission. It is also noted that distance to the edge node remains one of the critical factors affecting energy consumption.