Network virtualization provides a promising solution for next-generation network management by allowing multiple isolated and heterogeneous virtual networks to coexist and run on a shared substrate network. A long-standing challenge in network virtualization is how to effectively and efficiently map these virtual nodes and links of heterogeneous virtual networks onto specific nodes and links of the shared substrate network, known as the Virtual Network Embedding (VNE) problem. Existing centralized VNE algorithms and distributed VNE algorithms both have advantages and disadvantages. In this paper, a novel cooperative VNE algorithm is proposed to coordinate centralized and distributed algorithms and unite their respective advantages and specialties. By leveraging the learning technology and topology decomposition, autonomous substrate nodes entrusted with detailed mapping solutions cooperate closely with the central controller with a global view and in charge of general management to achieve a successful embedding process. Besides a topology-aware resource evaluation mechanism and customized mapping management policies, Bloom filter is elaborately introduced to synchronize the mapping information within the substrate network, instead of flooding which generates massive communication overhead. Extensive simulations demonstrate that the proposed cooperative algorithm has acceptable and even better performance in terms of long-term average revenue and acceptance ratio than previous algorithms.

In the era of smart cities, all vehicle systems will be connected to enhance the comfort of driving, relieve traffic congestion, and enjoy in-vehicle multimedia entertainment. The vision of all vehicles connected poses a crucial challenge for an individual vehicle system to efficiently support these applications. Network virtualization is a very promising enabling solution, by allowing multiple isolated and heterogeneous virtual networks (VNs), to satisfy the different quality of service (QoS) requirements. Smart identifier network (SINET) may provide VNs through effective resources allocation and control based on its model of three layers and two domains. In this paper, we provide resource allocation and mapping of the vehicular networks through elastic network virtualization based on SINET. The appropriate vehicles are selected and generated as a function group for special service, by which the difference of heterogeneous vehicular resources is hided. Vehicular nodes are autonomously organized that each of them can evaluate others’ resource availability in a topology-aware way with information by leveraging the learning technology and make its own decision to realize the whole mapping process through a phasing virtual network embedding (PVNE) algorithm. The results demonstrate that our proposed mechanism has better performance in long-term acceptance ratio and average revenue than existing state-of-the-art solutions.

Network virtualization provides a promising tool for next-generation network management by allowing multiple heterogeneous virtual networks to run on a shared substrate network. A long-standing challenge in network virtualization is how to effectively map these virtual networks onto the shared substrate network, known as the virtual network embedding (VNE) problem. Most heuristic VNE algorithms find practical solutions by leveraging a greedy matching strategy in node mapping. However, greedy node mapping may lead to unnecessary bandwidth consumption and increased network fragmentation because it ignores the relationships between the mapped virtual network requests and the mapping ones. In this paper, we re-visit the VNE problem from a statistical perspective and explore the potential dependencies between every two substrate nodes. We define a well-designed dependency matrix that represents the importance of substrate nodes and the topological relationships between them, i.e., every substrate node’s degree of belief. Based on the dependency matrix generated from collecting and processing records of accepted virtual network requests, Bayesian inference is leveraged to iteratively select the most suitable substrate nodes and realize our novel statistical VNE algorithm consisting of a learning stage and an inference stage in node mapping. Due to the overall consideration of the relationships between the mapped nodes and the mapping ones, our statistical approach reduces unnecessary bandwidth consumption and achieves a better performance of embedding. Extensive simulations demonstrate that our algorithm significantly improves the long-term average revenue, acceptance ratio, and revenue/cost ratio compared to previous algorithms.

As a disruptive technology, Network Virtualization plays an increasingly important role in Cloud Computing by allowing multiple heterogeneous virtual networks to run on a shared infrastructure as independent slices. However, it is a big challenge to map multiple virtual networks onto specific nodes and links in the shared substrate network, known as the Virtual Network Embedding (VNE) problem. Most existing solutions are provided in a centralized way, which may have hot spot problem. In this paper, by leveraging the learning and inference technology, a novel VNE algorithm is proposed to achieve the embedding in a distributed way. Instead of flooding, Bloom Filter is introduced to synchronize the mapping information, without generating massive communication overhead. The results of comparison between existing algorithms and the proposed algorithm show that our distributed algorithm has acceptable, even better performance in terms of long-term average revenue and acceptance ratio.

Network Virtualization provides a promising tool to allow multiple heterogeneous virtual networks to run on a shared substrate network simultaneously. A long-standing challenge in Network Virtualization is the Virtual Network Embedding (VNE) problem: how to embed virtual networks onto specific physical nodes and links in the substrate network effectively and efficiently. Recent research presents several heuristic algorithms that only consider single network topological attribute, which may lead to decreased utilization of resources. In this paper, we introduce seven complementary characteristics that reflect different topological attributes, and propose three topology-aware VNE algorithms by leveraging their respective advantages. Due to overall considering topological attributes of substrate and virtual networks through multiple characteristics, our study better coordinates node and link embedding. Extensive simulations demonstrate that our algorithms improve the long-term average revenue, acceptance ratio, and revenue/cost ratio compared to previous algorithms.

网络虚拟化技术可以在共用的底层网络基础设施上同时构建多个彼此隔离的虚拟网络,为用户提供差异化服务,从而解决现有因特网的僵化问题.然而,一个重要的挑战是,如何在共用的基础设施中高效地映射多个具有不同拓扑的虚拟网络,即虚拟网络的嵌套映射问题.主要根据基础设施的构成方式对现有的虚拟网络映射算法进行了综述.首先,阐述了网络虚拟化的概念、特点以及相应的虚拟网络映射模型;其次,按照基础设施的构成方式、问题空间完整性、映射阶段数等方面梳理了嵌套映射算法的最新研究进展;最后,对虚拟网络映射算法在公平性、扩展性、高利用率、信任度等未来可能的发展方向进行了展望.

Network virtualization provides a powerful tool to allow multiple networks, each customized to a specific purpose, to run on a shared substrate. However, a big challenge is how to map multiple virtual networks onto specific nodes and links in the shared substrate network, known as virtual network embedding problem. Previous works in virtual network embedding can be decomposed to two classes: two-stage virtual network embedding and one-stage virtual network embedding. In this paper, by pruning the topology of virtual network using k-core decomposition, a hybrid virtual network embedding algorithm, with consideration of location constraints, is proposed to leverage the respective advantage of the two kinds of algorithm simultaneously in the mapping process. In addition, a time-oriented scheduling policy is introduced to improve the mapping performance. We conduct extensive simulations and the results show that the proposed algorithm has better performance in the long-term run.

Multiple heterogenous virtual networks are given the ability to run on a shared infrastructure simultaneously as independent slices in the network virtualization environment. However, a major challenge is how to map multiple virtual networks, with specific node and link constraints, onto the shared substrate network, known as virtual network embedding problem. By taking topology attribute into account, topology-aware virtual network embedding algorithms efficiently improve the performance by leveraging a node ranking method based on Markov chain. However, as the basis of node ranking, the resource evaluation of node which is calculated as the product of its CPU and bandwidth may be incorrect. Moreover, a greedy matching strategy is always applied in the node mapping stage, which may lead to unnecessary bandwidth consumption by ignoring the relationships between the mapped substrate nodes and the mapping one. In this paper, we re-think the topology-aware virtual network embedding from a statistical perspective by proposing a statistical method to generate a dependency matrix representing the importance of every node and the relationships between every two nodes in the substrate network. Based on this dependency matrix, bayesian network analysis is leveraged to iteratively select the substrate node, with the closest relationship to the selected ones, to achieve node mapping process. Extensive simulations were conducted and the results show that our proposed algorithm has better performance in the long-term run.

Network virtualization provides a powerful tool to allow multiple networks, each customized to a specific purpose, to run on a shared substrate. However, a big challenge is how to map multiple virtual networks onto specific nodes and links in the shared substrate network, known as virtual network embedding problem. Previous works in virtual network embedding can be decomposed to two classes: two-stage virtual network embedding and one-stage virtual network embedding. In this paper, by pruning the topology of virtual network using k-core decomposition, a hybrid virtual network embedding algorithm is proposed to leverage the respective advantage of the two kinds of algorithm simultaneously in the mapping process. In addition, a time-oriented scheduling policy is introduced to improve the mapping performance. We conduct extensive simulations and the results show that the proposed algorithm obtains more revenue in the long-term run.