Communication and Computing Offloading in 6G Hotspot Scenarios

Technological progress has increased mobile device usage, boosting resource demands. Consequently, hotspots arising from heightened activities, like concerts in urban areas during peak times, lead to serious performance degradation. Often, these areas rely on a single ground-based base station, making advanced task offloading strategies essential. The advent of sixth generation (6G) Non-Terrestrial Networks (NTN) introduces diverse offloading options, including Unmanned Aerial Vehicles (UAVs) and High-Altitude Platform Stations (HAPS), aimed at addressing these challenges. Assuming that User Equipment (UE) within hotspots probabilistically offloads tasks to alleviate congestion, this study investigates the optimal offloading probabilities to minimize average task completion delays. We employ various queueing models (M/M/1, M/H2/1, and M/D/∞) within a five-tier network architecture to model different resource types. A Sub-Gradient Search (SGS) algorithm is then used to optimize the offloading probabilities. Our results show that the SGS algorithm reduces delays by 76% compared to a traditional approach, particularly favoring task completion within small cells during low system loads, and shifting to HAPS and UAVs when the load increases.