By separating network functionalities from the dedicated hardware and implementing them in the form of Virtual Network Function (VNF), Network Function Virtualization (NFV) provides a new service paradigm called Service Function Chain (SFC). It is unrealistic to replace the running VNFs due to the stateful information accumulated, which makes the current VNF re-deployment or replacement approaches no longer effective enough. Such a situation is even worse when VNFs are shared. Instead of replacing VNFs, this paper proposes a VNF migration approach to maintain the network balance and performance by migrating VNFs from poor-state nodes to good-state nodes. In particular, two situations are considered, which are node-aware and link-aware VNF migrations respectively. Via migrating the VNFs that are related to the nodes or links causing Quality of Service (QoS) issues, the network and service performance can be optimized. In addition, the migration impact is also minimized by perceiving the global network states. The experimental results indicate that the service performance is improved and the load balance is achieved after the VNF migration.

Network Function Virtualization (NFV) provides a flexible way to provision new services by decoupling network functions from hardware and implementing them as Virtual Network Functions (VNFs). However, the rapid development of technologies greatly promotes the explosion of diverse services, which directly results in the exponential increase of heterogeneous traffic. In addition, such a tremendous amount of heterogeneous traffic will generate bursts in a more dynamic and unexpected manner, so it becomes extremely hard to satisfy the customer demands. Aiming at addressing these challenges, this work proposes a positive and elastic VNF deployment mechanism for service provisioning, which introduces three novelties: 1) a Gated Recurrent Unit (GRU) based traffic prediction model is established to predict the unexpected and dynamically changing traffic behaviors in advance with the accuracy over 98%; 2) a closed-loop system is formed, in which the prediction model can learn and evolve continuously to respond to more complex scenarios; 3) different states of VNF are introduced and dynamically switched to deal with the current demands with reduced cost by avoiding frequent VNF initialization and destroy. The experimental results indicate that the proposed mechanism outperforms the state-of-theart methods, which include achieving over 98% prediction accuracy, improving the service acceptance rate by more than 18%, and reducing the overall cost by more than 20%

Network service provisioning becomes flexible and programmable with the help of Network Function Virfitualization (NFV), since NFV abstracts various service functions into software components called Virtual Network Function (VNF) and VNFs can be flexibly and quickly composed to form new services. It is commonly known that sharing the same VNF among different services can improve the resource utilization. However, we should be aware that such sharing also leads to serious resource preemption. In addition, VNF sharing aggravates the generation of the performance bottleneck, which then causes the rate mismatch problem between the upstream and downstream VNFs belonging to the same service chain. In this article, we propose a dynamic and flexible algorithm to jointly address the VNF sharing resource allocation and the rate coordination between the upstream and downstream VNFs. Specifically, 1) the VNFs are shared among different service chains with a fairness factor considered for the purpose of reducing the resource preemption probability and improving the resource utilization; 2) the backpressure indicator of each VNF is defined to judge its pressure condition, based on which we can dynamically adjust the processing rates between it and its downstream or upstream VNFs by maximizing the idle resource utilization. The experimental results indicate that the proposed algorithm outperforms the other methods in terms of the average delay, the flow completion time, the throughput and the backlog, etc. Meanwhile, the proposed algorithm achieves more stable performance than the other methods.

Extensive redundant links and dedicated network components are deployed in existing networks to avoid network congestion caused by peak traffic. These network components are underutilized, leading to an extreme waste of electrical energy and negative environmental consequences. Most energy efficiency schemes have little consideration of the subsequent traffic load as well as the impact of energy saving on network performance, which brings many challenges, such as frequent flow rerouting, continual device state transition, and network load imbalance. In this article, we propose an energy efficiency optimization framework based on traffic prediction in software-defined networking, aiming at reducing network energy consumption while ensuring communication quality. First, we design a real-time traffic prediction mechanism based on gated recurrent unit neural network of deep learning to capture the temporal characteristics of network traffic and provide data basis for the deployment of energy-saving strategies. Second, we propose a heuristic algorithm for energy efficiency optimization to balance flow demand and energy consumption and to achieve dynamic load balancing and energy saving. Finally, simulation is carried out through Ryu controller, Mininet, and TensorFlow. Experimental results show that our algorithm achieves about 47.71% reduction in terms of the overall energy consumption with good network load balancing.

Opinion maximization in social networks is an optimization problem, which targets at determining some influential individuals (i.e., seed nodes), propagating the desired opinion to their neighbors, and eventually obtaining maximum opinion spread. Previous studies assume that influence power of one individual is mainly calculated by using some network structure properties and once the opinion of one individual is determined, its opinion usually keeps unchanged. However, in the real scenario, the influence power of one individual may be unknown and should be closely associated with the dynamic opinion formation process. In this paper, we propose a novel Influence Power-based Opinion Framework (IPOF) to solve the opinion maximization problem, which is composed of two phases: 1) influence power estimation, and 2) elimination of influence overlapping (EIO). Specifically, we design the exponential influence power and estimate the unknown parameter of influence power through maximum likelihood estimation due to its simplicity, practicability, and superior convergence in large samples. To generate the opinion series dynamically, the weighted voter model is proposed by leveraging influence power and intimate degree. Moreover, we also prove that the likelihood function is concave by using Hessian matrix. To determine the initial seed nodes and facilitate large opinion propagation, influence power-based EIO algorithm is proposed. Experimental results in six social networks demonstrate that the proposed approach outperforms the state-of-the-art benchmarks.