INT8-quantized YOLOv8n on the Jetson Orin NX reaches about 66 FPS, making it the fastest tested configuration for drone-side detection, while the Raspberry Pi 5 stays below real-time rates.
Placement and Allocation of Communications Resources in Slicing-aware Flying Networks
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
Network slicing emerged in 5G networks as a key component to enable the use of multiple services with different performance requirements on top of a shared physical network infrastructure. A major challenge lies on ensuring wireless coverage and enough communications resources to meet the target Quality of Service (QoS) levels demanded by these services, including throughput and delay guarantees. The challenge is exacerbated in temporary events, such as disaster management scenarios and outdoor festivities, where the existing wireless infrastructures may collapse, fail to provide sufficient wireless coverage, or lack the required communications resources. Flying networks, composed of Unmanned Aerial Vehicles (UAVs), emerged as a solution to provide on-demand wireless coverage and communications resources anywhere, anytime. However, existing solutions mostly rely on best-effort networks. The main contribution of this paper is SLICER, an algorithm enabling the placement and allocation of communications resources in slicing-aware flying networks. The evaluation carried out by means of ns-3 simulations shows SLICER can meet the targeted QoS levels, while using the minimum amount of communications resources.
fields
cs.DC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
A Performance Analysis of You Only Look Once Models for Deployment on Constrained Computational Edge Devices in Drone Applications
INT8-quantized YOLOv8n on the Jetson Orin NX reaches about 66 FPS, making it the fastest tested configuration for drone-side detection, while the Raspberry Pi 5 stays below real-time rates.