Advanced Adaptive Scheduling for Autonomous Driving in Beyond-5G/6G Networks

Intelligent Congestion Control in the Internet of Vehicles Employing Network Slicing in 5G Architecture

The Internet of Vehicles (IoV) is a decentralized network architecture that enables autonomous driving, real-time applications, infotainment services, and seamless vehicle-to-everything communication. While infotainment systems enhance the user experience by providing entertainment and navigation features, their high data demands can cause network congestion, potentially delaying mission-critical messages and compromising safety and reliability. Further, the increasing volume of connected devices and data traffic exacerbates these challenges, resulting in high latency, low throughput, and reduced network efficiency. To address this, we propose a novel lightweight adaptive network slicing strategy with four dedicated slices and priority-weighted dynamic bandwidth allocation for IoV networks to mitigate congestion and ensure the availability of the required bandwidth for critical communications. The proposed mechanism defines four distinct network slices, allocating resources to balance infotainment and mission-critical needs. Simulation results demonstrate that our approach achieves ultra-low latency (<5 ms), near-zero packet loss (<0.5%), and high throughput (53 Mbps for infotainment), significantly outperforming existing methods, and ensuring reliable communication for safety-critical tasks while improving spectrum utilization. Findings validate implementing network slicing in IoV environments, paving the way for efficient, congestion-free, and high-performance vehicular networks.

Proximal policy optimization with self adaptive penalty function for vehicular resource allocation

The results demonstrate that the proposed PP-SAPF is suitable for real-time deployment in intelligent transportation systems (ITS) and autonomous vehicles where low latency, reliable connectivity, and adaptive resource management is significant.

A multi-criteria association policy for scalable auxiliary drone networks in large-scale events

Unmanned Aerial Vehicles (UAVs) have become a flexible alternative for providing temporary communication support in scenarios where fixed infrastructure may be overloaded or unavailable. This paper investigates UAV-based auxiliary networks for large-scale events, where sudden user density spikes can degrade the performance of conventional cellular systems. A key challenge in this context is maintaining stable UAV-user associations under high user density and limited UAV resources. Building on our previous Received Signal Strength Indicator-driven Association Policy (RDAP), this work introduces the Enhanced RDAP (e-RDAP), a multi-criteria association policy that combines RSSI, Packet Delivery Ratio (PDR), and communication delay with an adaptive deployment strategy. The proposed policy is evaluated using Network Simulator 3 (NS-3) under static and dynamic mobility scenarios, including realistic human mobility patterns. Results show that e-RDAP improves the reliability-efficiency trade-off compared with RDAP. In the highest-density dynamic scenario, e-RDAP achieves a $$3.8\%$$ 3.8 % improvement in PDR and an $$11.8\%$$ 11.8 % reduction in latency compared with RDAP. It also reduces reassociations by $$17.3\%$$ 17.3 % and controls overhead by $$27\%$$ 27 % compared with RDAP. Compared with the UAV-specific MULB-LA benchmark, e-RDAP also provides higher PDR, lower delay, fewer reassociations, and lower signaling overhead, indicating that QoS-aware multi-criteria association provides additional gains beyond load-aware association alone.

Multi-Agent Autonomous 6G Service Control With Intelligent Reconfiguration

Future 6G services will require strict performance guarantees, especially in terms of delay, end-to-end (e2e) across multiple network domains including packet and radio segments. While deterministic transport and slice-based capacity allocation can improve segment-level performance, ensuring e2e Network Service (NS) performance remains challenging as it requires making decisions Near–Real-Time (Near-RT) on a per-service basis, which does not fit well within the typical centralized control and orchestration hierarchy. Multi-agent systems (MAS), where a number of distributed agents collaborate, has demonstrated its capabilities for such Near-RT control. Agents equipped with Deep Reinforcement Learning (DRL) engines autonomously made traffic routing decisions based on e2e telemetry measurements. In this paper, we extend such MAS solutions for NS traffic routing focused on covering several issues that appear under frequent NS reconfiguration, e.g., caused by end device mobility. In addition, we define a lifecycle for NS operation that includes the initial MAS deployment, model reconfiguration during operation, and NS reconfiguration. The proposed lifecycle requires the definition of DRL training and validation procedures to produce models ready to be deployed with guaranteed performance under certain network conditions. In addition, model selection algorithms are defined for the lifecycle scenarios. In case of NS reconfiguration, a procedure for probe testing the actual network conditions is proposed to improve model selection. Evaluation across a meaningful set of network and traffic scenarios shows that the MAS is able to maintain e2e delay guarantees under all the lifecycle scenarios.

Hierarchical-Projected Scheduling for Mission-Critical Vehicular Edge Computing

Intelligent Placement of 5G Network Functions on Edge-Based Infrastructures

A constrained optimization model that supports different management goals through alternative objective functions (latency-aware or power-aware) while enforcing operational constraints, including node capacities, slice-specific latency bounds, and explicit limits on VNF migrations/relocations between scheduling periods is proposed.