TY - CHAP
T1 - Intelligent Resource Scheduling
AU - Yao, Haipeng
AU - Guizani, Mohsen
N1 - Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2023
Y1 - 2023
N2 - The continued growth in the number and applications of Internet of Things (IoT) connected devices makes it more challenging to meet multi-dimensional QoS within the same IoT network. In this chapter, we first design a network slicing architecture over the SDN-based long-range wide area network. The SDN controller can dynamically split the network into multiple virtual networks according to different business requirements. Then, a Continuous-Decision virtual network embedding scheme relying on Reinforcement Learning (CDRL) is proposed, two traditional heuristic embedding algorithms as well as the classic reinforcement learning aided embedding algorithm are used for benchmarking our proposed CDRL algorithm. Finally, we propose a hybrid intelligent control architecture, which adopts the centralized training and distributed execution paradigm. A centralized critic is introduced to ease the training process of the distributed network nodes. Besides, considering the competitive behavior of users, we formulate the resource allocation problem as a multi-user competition game model. Based on this, we proposed a multi-agent reinforcement learning-based SFCs deployment algorithm.
AB - The continued growth in the number and applications of Internet of Things (IoT) connected devices makes it more challenging to meet multi-dimensional QoS within the same IoT network. In this chapter, we first design a network slicing architecture over the SDN-based long-range wide area network. The SDN controller can dynamically split the network into multiple virtual networks according to different business requirements. Then, a Continuous-Decision virtual network embedding scheme relying on Reinforcement Learning (CDRL) is proposed, two traditional heuristic embedding algorithms as well as the classic reinforcement learning aided embedding algorithm are used for benchmarking our proposed CDRL algorithm. Finally, we propose a hybrid intelligent control architecture, which adopts the centralized training and distributed execution paradigm. A centralized critic is introduced to ease the training process of the distributed network nodes. Besides, considering the competitive behavior of users, we formulate the resource allocation problem as a multi-user competition game model. Based on this, we proposed a multi-agent reinforcement learning-based SFCs deployment algorithm.
KW - Hybrid intelligent control
KW - Long-range wide area network
KW - Network slicing
KW - Service function chain
UR - https://www.scopus.com/pages/publications/85162137597
U2 - 10.1007/978-3-031-26987-5_5
DO - 10.1007/978-3-031-26987-5_5
M3 - Chapter
AN - SCOPUS:85162137597
T3 - Wireless Networks (United Kingdom)
SP - 211
EP - 269
BT - Wireless Networks (United Kingdom)
PB - Springer Nature
ER -