Smart City Gnosys
Smart city article details
| Title | Initializing Sdn-Nfv Integration For Smart City As Environment To Dqn Agent In Qos Optimization |
|---|---|
| ID_Doc | 31576 |
| Authors | Song I.; Tam P.; Ryoo I.; Kim S. |
| Year | 2024 |
| Published | Lecture Notes in Electrical Engineering, 1190 LNEE |
| DOI | http://dx.doi.org/10.1007/978-981-97-2447-5_3 |
| Abstract | Software-defined networking (SDN) and network functions virtualization (NFV) are two enabling paradigms at the forefront of reshaping the management of smart city core networking. The emergence of these paradigms presents a significant opportunity to revolutionize the way cities handle the core network infrastructure and services. However, current challenges are in ensuring that the Quality-of-Service (QoS) requirements for smart city applications, which consistently met to guarantee reliable and responsive connections. This paper introduces a novel approach based on deep Q-networks (DQN) to address the intricacies of QoS provisioning in the core domain of SDN and NFV. We initialize SDN-NFV systems into observable environment to DQN agent. By leveraging the epsilon-greedy algorithm, our method iteratively optimizes both online and target networks, which leads to the establishment of an adaptive control policy. Policy orchestration plays an important role in dynamically adapting network resources to meet the evolving demands of smart city applications. A key highlight of our approach is the novel reward mechanism, which is based on the efficiency of QoS metrics. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024. |
| Author Keywords | Deep Q-Networks; Environment Initialization; Quality-of-Service; Smart City; State Observability |
