Scalable Orchestration of Service Function Chains in NFV-Enabled Networks: A Federated Reinforcement Learning Approach
48
Citation
39
Reference
10
Related Paper
Citation Trend
Abstract:
Network function virtualization (NFV) is critical to the scalability and flexibility of various network services in the form of service function chains (SFCs), which refer to a set of Virtual Network Functions (VNFs) chained in a specific order. However, the NFV performance is hard to fulfill the ever-increasing requirements of network services mainly due to the static orchestrations of SFCs. To tackle this issue, a novel Scalable SFC Orchestration (SSCO) scheme is proposed in this paper for NFV-enabled networks via federated reinforcement learning. SSCO has three remarkable characteristics distinguishing from the previous work: (1) A federated-learning-based framework is designed to train a global learning model, with time-variant local model explorations, for scalable SFC orchestration, while avoiding data sharing among stakeholders; (2) SSCO allows for parameter update among local clients and the cloud server just at the first and last epochs of each episode to ensure that distributed clients can make model optimization at a low communication cost; (3) SSCO introduces an efficient deep reinforcement learning (DRL) approach, with the local learning knowledge of available resources and instantiation cost, to map VNFs into networks flexibly. Furthermore, a loss-weight-based mechanism is proposed to generate and exploit reference samples in replay buffers for future training, avoiding the strong relevance of samples. Simulation results obtained from different working scenarios demonstrate that SSCO can significantly reduce placement errors and improve resource utilization ratio to place time-variant VNFs compared with the state-of-the-art mechanisms. Furthermore, the results show that the proposed approach can achieve desirable scalability.Keywords:
Orchestration
Virtual network
Network service
Network function virtualization (NFV) is a technique to facilitate service deployment by decoupling network functions from dedicated hardware and moving them to software, namely virtual network functions (VNFs). Network services are carried out by service function chains (SFCs) comprising multiple VNFs. Given SFCs and physical machines (PMs), the VNF deployment problem asks how to assign each SFC's VNFs to PMs, such that the service ratio is maximized. Many methods use as few PMs as possible for deploying VNFs, but they lead to imbalanced loads of PMs in a cloud network. This paper proposes a load-balanced VNF deployment (LBVD) scheme, which scores each PM based on the expected ratios of residual resources of that PM and its neighbors. Then, VNFs are deployed on PMs with high scores. If a PM is busy, LBVD lets its VNF migrate to another PM for load sharing. Through simulations, we show that LBVD can balance the loads of PMs and save both deployment and migration costs.
Virtual network
Network Functions Virtualization
Network service
Cite
Citations (4)
As the cloud-based datacenter with SDN (Software Defined Network) enabled has been deployed as a real service platform, NFV (Network Function Virtualization) attracts great attention from many researchers. NFV virtualizes network equipment to implement elastic network service in the cloud. Specifically, in this paper, we propose screen acceleration VNF (Virtual Network Function) scheme that makes VDI (Virtual Desktop Infrastructure) service dynamically operational in the cloud and NFV platform.
Network Functions Virtualization
Virtual network
Network service
Network Virtualization
Cite
Citations (0)
The placement of virtual network functions (VNFs) has been widely studied. We motivate our research by the following two observations. First, in many situations, such as enterprise and edge networks, the hosting servers may not have been deployed for VNF placement. Second, some VNFs may only process/cover a partial set of flows (such as elephant flows) but still improve network performance. To take the above observations into consideration, we propose to incrementally deploy servers and place VNFs with explicit performance guarantee. We formally define the Incremental Server and Virtual Network Function Co-Placement (InSCOPE) problem. We prove the NP-hardness of the InSCOPE problem. The results of extensive simulation show high performance of our proposed algorithm. For example, our proposed algorithm can reduce the number of servers by about 50%.
Virtual network
Network Functions Virtualization
Cite
Citations (0)
With the development of network function virtualization (NFV), the resource management of service function chains (SFC) in the virtualized environment has gradually become a research hotspot. Usually, users hope that they can get the network services they want anytime and anywhere. The network service requests are dynamic and real-time, which requires that the SFC in the NFV environment can also meet the dynamically changing network service requests. In this regard, this paper proposes an SFC deployment method based on traffic prediction and adaptive virtual network function (VNF) scaling. Firstly, an improved network traffic prediction method is proposed to improve its prediction accuracy for dynamically changing network traffic. Secondly, the predicted traffic data is processed for the subsequent scaling of the VNF. Finally, an adaptive VNF scaling method is designed for the purpose of dynamic management of network virtual resources. The experimental results show that the method proposed in this paper can manage the network resources in dynamic scenarios. It can effectively improve the availability of network services, reduce the operating overhead and achieve a good optimization effect.
Virtual network
Network service
Cite
Citations (2)
Network Function Virtualization (NFV) has revolutionized the way networking services are offered and deployed. Moving away from a rigid and hardware-centric approach, where expensive and dedicated network components are used, NFV is now leveraging standard x86 servers, where softwarized images of network functions (NFs) can be hosted as Virtual Machines (VNFs) or containers (CNFs). However, in terms of deploying, configuring, and interconnecting these softwarized images, a lot of manual intervention is required. To this end, the Intent-Based Networking (IBN) paradigm has emerged, which has as a goal to automate the network configuration by translating a high-level and abstract request of a network service into a detailed policy description. Usually, IBN and NFV are studied separately, even though in reality they are highly correlated and can benefit from each other. In particular, network services can be expressed as abstract service requirements from the users, where through an IBN System (IBNS) will be translated into specific network policies and a VNF/CNF deployment solution, called VNF Placement solution. Accordingly, in this paper, we aim to combine these two technologies together in order to automate the deployment of the VNFs in a Cloud-based infrastructure, while supporting multitenancy and intent refinement. Our results reveal that an IBN-based VNF placement solution can successfully offer network services, expressed as user intents, in such a way that the network services are automatically configured according to the quality of service and security requirements included in the intent.
Virtual network
Network service
Network Functions Virtualization
Cite
Citations (29)
Network function virtualization (NFV) provides flexible and scalable network services by leveraging software-based network appliances. In an NFV architecture, the virtual network function manager (VNFM) has a critical role in provisioning, configuring, and operating virtual network functions (VNFs) and service function chaining (SFC). In addition, the VNF manager should provide other functions to guarantee high availability of network services, such as fault management, monitoring and auto scaling. Currently, many open source projects aim to implement these functions, but there is no standard method of monitoring and detecting failure of SFC and VNFs. Therefore, we propose an alarm-based monitoring driver architecture to support high-availability VNF. This driver can activate an alarm upon VNF failure and will be used to support high-availability SFC. This monitoring driver is implemented in a cloud environment to show its feasibility and advantages.
Virtual network
Chaining
Provisioning
High availability
Network service
Network Functions Virtualization
Failover
Network monitoring
Fault management
Cite
Citations (6)
Network function virtualization (NFV) is a promising solution for cost reduction, service agility, and scalability to cope with the dynamic changes of traffic volume and business requirements. In the NFV framework, management and orchestration framework (MANO) plays an important roles in the automated provisioning, configuration, lifecycle management, and high availability of virtual network functions (VNF). In order to achieve high availability and auto-scaling for VNFs, the MANO needs to integrate with cluster management service for scaling VNF in and out on demand basis. In this paper, we introduce an architecture which integrates clustering service into the MANO framework and define policies for cluster management in TOSCA template. Our architecture exploits policies-based cluster management and TOSCA service description template to provide the high availability and scalability to the NFV framework. We also shows our implementation using related open source projects.
Orchestration
Virtual network
Provisioning
Network service
Cite
Citations (7)
Virtual Content Delivery Network (vCDN) orchestration is necessary to optimize the use of resources and improve the performance of the overall SDN/NFV-based CDN function in terms of network operator cost reduction and high streaming quality. It requires intelligent and enticed joint SDN/NFV orchestration algorithm due to the evident huge amount of traffic to be delivered to end customers of the network. In this paper, a global vCDN architecture and an exact approach for finding the optimal path orchestration(s) and vCDN component instantiation(s) (OCPA) are proposed. Moreover, several scenarios are considered to quantify the OCPA behavior and to compare its efficiency in terms of caching and streaming cost, orchestration time, vCDN replication number, and other cost factors. Then, it is implemented and evaluated under different deployment flavors. Several scenarios are considered to study the algorithm's behavior and to quantify the impact of both network and system parameters.
Orchestration
Virtual network
Cost reduction
Cite
Citations (12)
Network Function Virtualization (NFV) has been a prominent shift from dedicated network devices towards reusable Virtual Network Functions (VNF). Today's network services are defined as a graph of VNFs connected via Virtual Links (VL), which are realized by placing its elements on their corresponding substrate network nodes and links. Efficient placement of the network services is one of the most important challenges introduced by NFV. We propose a demonstration of the "Network Service Placer (NSPlacer)", a tool to enable the evaluation of the placement algorithms. By being able to adjust various parameters, related to the substrate network, the service and the placement algorithm, and thanks to the already implemented interesting placement algorithms, our online tool can provide near-optimal and optimal results to serve as a comparison with many placement solutions. Made available to the community, our tool aims to contribute to the reproducibility of the research results by offering a set of reference algorithms to which everyone can compare their solution and even integrate it into the set of available algorithms of the tool.
Cite
Citations (0)
Network Function Virtualization (NFV) is enabling the softwarization of traditional network services, commonly deployed in dedicated hardware, into generic hardware in form of Virtual Network Functions (VNFs), which can be located flexibly in the network. However, network load balancing can be critical for an ordered sequence of VNFs, also known as Service Function Chains (SFCs), a common cloud and network service approach today. The placement of these chained functions increases the ping-pong traffic between VNFs, directly affecting to the efficiency of bandwidth utilization. The optimization of the placement of these VNFs is a challenge as also other factors need to be considered, such as the resource utilization. To address this issue, we study the problem of VNF placement with replications, and especially the potential of VNFs replications to help load balance the network, while the server utilization is minimized. In this paper we present a Linear Programming (LP) model for the optimum placement of functions finding a trade-off between the minimization of two objectives: the link utilization and CPU resource usage. The results show how the model load balance the utilization of all links in the network using minimum resources.
Virtual network
Network service
Network Functions Virtualization
Replication
Cite
Citations (22)