On Dynamic Service Function Chain Reconfiguration in IoT Networks

Liu, Yicen; Lu, Yannan; Li, Xi; Yao, Zhigang; Zhao, Donghao

doi:10.1109/jiot.2020.2991753

Cited by 29 publications

(18 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The simulation results show that the proposed LARA algorithm has better performance in end-to-end delay and reliability assurance. We are still exploring how to combine our proposed LARA algorithm with some machine-learningbased approaches (e.g., GNN [28,29], MLP [30]), which can forecast the future trends of link/node load and traffic variations ahead of time, and thus improve the efficacy of the SSC resource allocation algorithm.…”

Section: Discussionmentioning

confidence: 99%

Reliability Assurance Dynamic SSC Placement Using Reinforcement Learning

Jiang

Zhang

et al. 2022

Information

View full text Add to dashboard Cite

Software-defined networking (SDN) and network function virtualization (NFV) make a network programmable, resulting in a more flexible and agile network. An important and promising application for these two technologies is network security, where they can dynamically chain virtual security functions (VSFs), such as firewalls, intrusion detection systems, and intrusion prevention systems, and thus inspect, monitor, or filter traffic flows in cloud data center networks. In view of the strict delay constraints of security services and the high failure probability of VSFs, we propose the use of a security service chain (SSC) orchestration algorithm that is latency aware with reliability assurance (LARA). This algorithm includes an SSC orchestration module and VSF backup module. We first use a reinforcement learning (RL) based Q-learning algorithm to achieve efficient SSC orchestration and try to reduce the end-to-end delay of services. Then, we measure the importance of the physical nodes carrying the VSF instance and backup VSF according to the node importance of VSF. Extensive simulation results indicate that the LARA algorithm is more effective in reducing delay and ensuring reliability compared with other algorithms.

show abstract

Section: Discussionmentioning

confidence: 99%

Reliability Assurance Dynamic SSC Placement Using Reinforcement Learning

Jiang

Zhang

et al. 2022

Information

View full text Add to dashboard Cite

show abstract

“…Firstly, the proposed PSO algorithm mainly depends on the online information of substrate resource utilization and security service demands, which still remains no guarantee that the results obtained by the proposed PSO algorithm are the most accurate in the dynamics. We are still exploring how to combine our proposed PSO algorithm with some machinelearning-based approaches (e.g., GNN [7], MLP [47,48], LSTM [49]), which can forecast the future trends of link/node load and traffic variations ahead of time, and thus improve the efficacy of the SSC resource allocation algorithm. Second, we have conducted evaluations on the simulators.…”

Section: Discussionmentioning

confidence: 99%

“…Driven by the diversity of heterogeneous network security services, the network framework design has been transiting from the monolithic pattern to the softwaried paradigm, which is mainly supported by the network function virtualization (NFV) and software-defined network (SDN) technologies [6][7][8][9]. These technologies present themselves as revolution pivotal network architectural design concepts that leverage the virtualization and cloud infrastructure elasticity for the purpose of supporting this quantum leap of the existing packet core which, in turn, leads to the remarkable improvement of provisioned security network services.…”

Section: Introductionmentioning

confidence: 99%

A Novel Method for Resource Efficient Security Service Chain Embedding Oriented to Cloud Datacenter Networks

Qiao

Liu

et al. 2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

An important and promising application for the network function virtualization (NFV) technology is in network security, where can dynamically and flexibly accomplish the chaining virtualized security network functions (VSNFs), e.g., network address translation, antispam and packet filter firewall, etc., and thus inspect, monitor or filter traffic flows in the cloud datacenter networks. However, the traffic flows addressed by the VSNFs mainly depend on the security service requirements from mobile users, such as the network security level, end-to-end latency, and security resource, etc. Considering the dynamic nature of cloud datacenter networks, determining the embedding of VSNFs and routing security service paths that optimizes the security resource utilization is a challenging problem, particularly without violating the end-to-end delay constraints and security service requirements. This can also be called security service chain dynamic embedding problem (SSC-DMP). In this paper, we present an NFV-enabled framework for a system that achieves the SSC dynamic embedding in the cloud datacenter networks. We first formulate an integer linear programming (ILP) model to solve the SSC-DMP exactly in small-scale network topology. Then, in order to reduce the time complexity when applying the large-scale network topology, we propose an efficient SSC dynamic embedding solution that is based on the particle swarm optimization. Extensive simulation results show that the proposed algorithm could significantly outperform the current benchmarks at least 35.2% and 23.1% in terms of resource consumption and end-to-end delay, respectively.

show abstract

“…The graph neural network was first applied to the SDN /NFV field by the author of [16], who constructed a supervised learning method for service function chain traffic prediction using GNN. The author of [17] also used GNN to predict the resource demand of NFV, which approximated the performance of integer linear programming model in polynomial time and improved the reconfiguration overhead and resource utilization of service provision, respectively. The author of [18] used the graph neural network to predict various network performance indexes under a given routing strategy in the SDN network and achieved good results.…”

Section: Related Workmentioning

confidence: 99%

Security Service Function Chain Based on Graph Neural Network

Wang

Zhang

et al. 2022

Information

View full text Add to dashboard Cite

With the rapid development and wide application of cloud computing, security protection in cloud environment has become an urgent problem to be solved. However, traditional security service equipment is closely coupled with the network topology, so it is difficult to upgrade and expand the security service, which cannot change with the change of network application security requirements. Building a security service function chain (SSFC) makes the deployment of security service functions more dynamic and scalable. Based on a software defined network (SDN) and network function virtualization (NFV) environment, this paper proposes a solution to the particularity optimization algorithm of network topology feature extraction using graph neural network. The experimental results show that, compared with the shortest path, greedy algorithm and hybrid bee colony algorithm, the average success rate of the graph neural network algorithm in the construction of the security service function chain is more than 90%, far more than other algorithms, and far less than other algorithms in construction time. It effectively reduces the end-to-end delay and increases the network throughput.

show abstract

On Dynamic Service Function Chain Reconfiguration in IoT Networks

Cited by 29 publications

References 29 publications

Reliability Assurance Dynamic SSC Placement Using Reinforcement Learning

Reliability Assurance Dynamic SSC Placement Using Reinforcement Learning

A Novel Method for Resource Efficient Security Service Chain Embedding Oriented to Cloud Datacenter Networks

Security Service Function Chain Based on Graph Neural Network

Contact Info

Product

Resources

About