What type of study is this?

September 10, 2025Open Access

Adaptive QoS Policies in Smart City Mobile Networks

CMC Mahiba HAHassan AliUniversity of Al-Qadisiyah MAMohan Bangalore Anjaneyalu

Key Points

The proposed framework achieves a 46% reduction in latency, improving mobile user experiences in smart cities.
Results indicate a 25% increase in throughput, demonstrating the effectiveness of the adaptive QoS policy engine in real-time applications.
Integration of machine learning within the framework enables dynamic traffic classification and predictive resource allocation.
The study suggests that adaptive QoS policies are essential for meeting the demands of increasingly complex mobile networks in urban settings.

Abstract

The increased intricacy related to the architecture of smart cities necessitates sophisticated and intelligent adaptation of Quality of Service (QoS) techniques to satisfy the distinct needs of different mobile users and applications. The coexistence of diverse latency-sensitive, bandwidth-intensive, and mission-critical services in smart cities imposes dynamically changing QoS requirements that cannot be met with traditional static or best-effort policies. This paper proposes an Adaptive QoS Policy Framework based on Software-Defined Networking (SDN), Mobile Edge Computing (MEC), and Machine Learning (ML) based prediction models for context-aware specific service QOS delivery in real-time. This framework is centered on the Adaptive QoS Policy Engine (AQPE), which executes dynamic traffic classification alongside predictive resource allocation and policy change employing reinforcement learning. The system was simulated in NS-3 with an OpenFlow SDN controller integrated with MEC modules developed in Python. It was tested against industry standards benchmarks of latency, jitter, overall throughput, and packet delivery ratio (PDR). The results suggest the model achieves, on average, up to 46% reduction in latency, 25% improvement in throughput, and over 10% increase in PDR compared to baseline approaches. The results show the value of employing learning intelligence and edge computing within mobile network architectures intended for smart cities. As noted earlier, the adaptive metropolitan approach provides an extendable, controllable, real-time, and service-aware QoS configuration policy fulfilling the requirements for next-generation urban communication systems.

Ask AI

Helpful

Bookmark

View Full Paper