Content uploaded by Bin Tang
Author content
All content in this area was uploaded by Bin Tang on Jul 24, 2014
Content may be subject to copyright.
A preview of the PDF is not available
Opportunistic Bandwidth Sharing for Virtual Network Mapping
Abstract and Figures
Network virtualization has emerged as a powerful way to fend off the current ossification of the Internet. A major challenge is virtual network mapping, which is to assign substrate resources to virtual networks (VNs) such that some predefined constraints are satisfied and substrate resources are utilized in an effective and efficient manner. Due to the NP-completeness of this problem, a variety of heuristic algorithms have been proposed. However, existing solutions rarely consider the inefficient utilization of bandwidth resources due to the network traffic fluctuation. In this paper, we study the opportunistic bandwidth sharing in a single physical link among multiple virtual links from different VNs. We formulate the problem of assigning time slots to dispensable sub-flows with constraints on the performance guarantee and the objective of minimizing the number of time slots used, as an optimization problem. Two heuristic algorithms HA-I and HA-II, which consider the problem from different perspectives, are presented. Extensive simulations are conducted to evaluate the effectiveness and efficiency of our algorithms.
Figures - uploaded by Bin Tang
Author content
All figure content in this area was uploaded by Bin Tang
Content may be subject to copyright.
... One of the significant problems for a substrate network provider is how to embed the requested virtual networks into the substrate network, and a variety of virtual network embedding (VNE) methods have been proposed [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] . However, these methods do not take into consideration substrate resource sharing among the multiple priority classes within each of the requested virtual networks. ...
... Most of these heuristic methods solve the virtual node and virtual link assignment separately. First, each virtual node is assigned to a substrate node using the heuristics related to the resource utilization and topology information around each substrate node [5][6][7][8][9][10] . Although each virtual link is generally assigned to the shortest substrate path between a pair of substrate nodes to which the virtual nodes are assigned, a virtual link may be assigned to multiple substrate paths by solving the multi-commodity flow problem [11,12] . ...
... Zhang et al. [7,8] have considered substrate resource sharing among multiple virtual networks. However, their method only considers substrate resource sharing within the same priority class and does not consider sharing among different priority classes while at the same time satisfying the different latency requirements. ...
... One of the significant problems for a substrate network provider is how to embed the requested virtual networks into the substrate network, and a variety of virtual network embedding (VNE) methods have been proposed [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]. However, these methods do not take into consideration substrate resource sharing among the multiple priority classes within each of the requested virtual networks. ...
... Most of these heuristic methods solve the virtual node and virtual link assignment separately. First, each virtual node is assigned to a substrate node using the heuristics related to the resource utilization and topology information around each substrate node [5][6][7][8][9][10]. Although each virtual link is generally assigned to the shortest substrate path between a pair of substrate nodes to which the virtual nodes are assigned, a virtual link may be assigned to multiple substrate paths by solving the multi-commodity flow problem [11,12]. ...
... Zhang et al. [7,8] have considered substrate resource sharing among multiple virtual networks. However, their method only considers substrate resource sharing within the same priority class and does not consider sharing among different priority classes while at the same time satisfying the different latency requirements. ...
Network virtualization is a promising approach for virtual network operators to configure their own inter-cloud networks flexibly on an inter-cloud substrate network. Virtual network operators specify multiple priority classes in order to satisfy different latency requirements for their inter-cloud network services with a variety of delay-sensitivities on their virtual networks. Furthermore, they may require substrate resource sharing among multiple priority classes in order to reduce the amount of substrate resources assigned to them. Meanwhile, it is desirable for a substrate network provider that multiple virtual networks can share substrate resources since the total amount of substrate resources required can be reduced due to the effect of statistical multiplexing. This paper formulates a novel virtual network embedding problem and proposes a heuristic virtual network embedding method to minimize the total substrate resources required when the virtual network operators request substrate resource sharing among multiple priority classes within their virtual networks. Based on the proposed method, multiple virtual networks can maximally share substrate resources with one another while sharing substrate resources on an equal basis. The effect of substrate resource sharing among multiple priority classes and multiple virtual networks is quantitatively assessed through extensive simulations, and advantageous topologies for requested virtual networks and substrate networks are also presented.
... Zhang et al. conducted a series of research that considers the bandwidth (BW) variant nature of VNs during the process of resources provisioning, [Zhang et al. 2011, Zhang et al. 2012, Zhang et al. 2014]. The authors modeled the time-variant nature of the VNs demands as the combination of a basic sub-requirement, which exists all through the VNs lifetime, and a variable sub-requirement, which exists with a probability. ...
... In [Zhang et al. 2011], Zhang et al. propose a bandwidth sharing technique that allocates bandwidth (BW) in accordance with VNs traffic fluctuation as detailed in the related works. The authors consider specific design of the SN as the following: the time is partitioned into frames of equal length, and each frame is further divided into slots of equal length. ...
... In our work, we inspire a suitable SN design from the work of Zhang et al. [Zhang et al. 2011] after adapting it in what matches our problem: i) in our work the virtual flows are of fixed BW demand during time (not opportunistic demands), thus, we disconsider collision probability; ii) the HSVNs demand synchrony once during T , see 3.1, so, we need only one time window during T that applies Zhang et al. technique. We name this time window synchronous frame. ...
Network virtualization has been proposed in the last years, and it received special attention from both networking and distributed systems communities. However, specific applications' requirements, such as topology, security, and resilience, pose different challenges to the network embedding problem. Among these requirements lays the one of synchrony, that some applications demand time bounds for processing and communication. In this sense, Hybrid Syn-chrony Virtual Networks (HSVN) has been proposed to fulfill specific synchrony requirements and better support a considerable class of distributed systems. In our previous works, we gave attention to " space " HSVNs that was addressed to support hybrid synchrony systems in space. In this work, analogously, we discuss hybrid synchrony virtual networks for the " time " dimension. We regard the time HSVNs abstractions and techniques as being a refinement of the space HSVNs, since it further defines repeating time windows of synchrony while allowing the resource to behave asynchronously for a period of time. The timely hybrid model leads to the possibility of further sparing synchronous resources, if compared to the space model, as will be presented in this paper.
... A mixed integer programming model has been built for it and deterministic and random algorithms are proposed to solve the problem in [19]. [20] realized the fluctuations of the network workload, therefore, they studied opportunistic bandwidth sharing on a single physical link among multiple virtual links in different VNs, assuming time division multiplexing. This sharing is achieved through assigning different time slots to each virtual links. ...
... The status of each physical node is randomly determined to be awake or asleep with equal probability 0.5. Each physical node is randomly assigned a value between [20,35] to indicate its maximum available resource following uniform distribution. To examine the impact of available resources on energy consumption, we randomly deduct a portion of available resources of each physical node to simulate the initial resource usage. ...
With the rapid proliferation of data centers, their energy consumption and green house gas emissions have significantly increased. Some efforts have been made to control and lower energy consumption of data centers such as proportional energy consuming hardware, dynamic provisioning and virtual-ization machine techniques. However, it is still common that many servers and network resources are often underutilized, and idle servers spend a large portion of their peak power consumption. We first built a novel model of a network virtualization in order to minimize energy usage in data centers for both computing and network resources by taking practical factors into consideration. Due to the NP-hardness of the proposed model, we have developed a heuristic algorithm for virtual network scheduling and mapping, considering expected energy consumption at different times, a data center architecture, and virtual network migration, as well as operation costs. Our evaluation results show that our algorithm could reduce energy consumption up to 40%, and take up to 57% higher number of virtual network requests over other existing virtual mapping schemes.
... There are some VNF-FGE methods mentioning that the demanded bandwidth of a mapped VL might change over time. Some works considered it as a stochastic variable [12] while others appropriately shared it among different flows [13,14]. Studying these issues requires separate analysis and is outside the scope of this work. ...
With the development of heterogeneous network structure, dynamic user requests as well as complex service types and applications scenarios, current networks may not accommodate the increasingly stringent requirements. As a result, the research of the beyond fifth generation (B5G) or the sixth generation (6G) networks has been put on the agenda. In B5G/6G networks, achieving the automatic, flexible, and cost-effective orchestration and management of network resources is a significant but challenging issue. Network function virtualization (NFV), as a promising paradigm to address this issue, has received considerable attention from both industry and academia. NFV leverages the virtualization technology to decouple network functions from dedicated hardware appliances to software middleboxes or called virtual network functions (VNFs) that run on the commodity servers. The demand for a network service becomes a request for running a set of VNFs deployed on the substrate network. The requested network service is orchestrated in the form of a VNF-forwarding graph (VNF-FG). The problem of embedding the VNF-FG into the substrate network is known as VNF-FG embedding (VNF-FGE). The efficiency and the management cost of a network are highly dependent on the optimization of VNF-FGE. This paper mainly presents a survey on solving the VNF-FGE problem. To this end, we present a general formulation and several objectives of the VNF-FGE problem. In the meanwhile, we summarize its different application scenarios from four perspectives and divide the approaches into four main categories based on the optimization methods. The main challenges and potential future directions due to the appearance of B5G/6G are also discussed.
... A VN embedding algorithm while addressing the delay, routing and location constraints has been worked upon using a centralized approach in [24]. Another centralized heuristic is proposed in [25] based on resource sharing that deals with load fluctuation of the multiple virtual network requests. Fajjari et al. [26] proposed a max-min ant colony meta-heuristic based centralized virtual network embedding (VNE) approach. ...
The multiple infrastructure provider network virtualization system is a self-governing system that independently figures out the appropriate number of infrastructure providers and selects the most suitable infrastructure providers in the group to map the virtual network request. This work applies a fault tolerant market-based strategy for efficient working of a network virtualization environment by cooperation of multiple infrastructure providers while dealing with restricted or absolute failures of infrastructure providers during embedding in an unknown, random and dynamic virtual network system. The approach uses auctioning mechanism to decide on the infrastructure provider to serve the virtual network requests, sent by the service providers. A request can be mapped either by a single infrastructure provider or cooperatively by multiple infrastructure providers, depending on the requirements of the service provider’s request and the network resources available with the infrastructure provider. The infrastructure provider that matches best the demands of the service provider is chosen by the system. The feasibility of the proposed methodology is tested by implementing the approach on a group of multiple infrastructure providers that participate in the auctioning mechanism to serve multiple service providers’ virtual network requests.
In a datacenter, complex and time-varying interactions between various tiers and services of web applications, and the contention of shared resources among co-located virtual machines have significant impact on the user perceived performance and power consumption of the underlying system. We propose and develop APPLEware, an autonomic middleware for joint performance and power control of co-located web applications in virtualized datacenters. It features a distributed control structure that provides predictable performance and energy efficiency for large complex systems. It applies machine learning based self-adaptive modeling to capture the complex and time-varying relationship between the application performance and allocation of resources to various application components, in the face of highly dynamic and bursty workloads. The distributed controllers coordinate with each other and allocate resources to meet the service level agreements of applications in an agile and energy-efficient manner. Experimental results based on a testbed implementation with benchmark applications and large scale simulations demonstrate APPLEware's effectiveness, energy efficiency and scalability.
Nowadays, cloud data centers play an important role in modern Information Technology (IT) infrastructures, being progressively adopted in different scenarios. The proliferation of cloud has led companies and resource providers to build large warehouse-sized data centers, in an effort to respond to costumers demand for computing resources. Operating with powerful data centers requires a significant amount of electrical power, which translates into more heat to dissipate, possible thermal imbalances, and increased electricity bills. On the other hand, as data centers grow in size and in complexity, failure events become norms instead of exceptions. However, failures contribute to the energy waste as well, since preceding work of terminated tasks is lost. Therefore, today's cloud data centers are faced with the challenge of reducing operational costs through improved energy utilization while provisioning dependable service to customers. This chapter discusses the causes of power and energy consumption in data centers. The advantages brought by cloud computing on the management of data center resources are discussed, and the state of the art on schemes and strategies to improve power and energy efficiency of computing resources is reviewed. A practical case of energy-efficient and service-level agreement (SLA)-based management of resources, which analyzes and discusses the performance of three state-of-the-art scheduling algorithms to improve energy efficiency, is also included. This chapter concludes with a review of open challenges on strategies to improve power and energy efficiency in data centers.
MapReduce is widely used in cloud applications for large-scale data processing. The increasing number of interactive cloud applications has led to an increasing need for MapReduce real-time scheduling. Most MapReduce applications are data-oriented and nonpreemptively executed. Therefore, the problem of MapReduce real-time scheduling is complicated because of the trade-off between run-time blocking for nonpreemptive execution and data-locality. This paper proposes a data-locality-aware MapReduce real-time scheduling framework for guaranteeing quality of service for interactive MapReduce applications. A scheduler and dispatcher that can be used for scheduling two-phase MapReduce jobs and for assigning jobs to computing resources are presented, and the dispatcher enable the consideration of blocking and data-locality. Furthermore, dynamic power management for run-time energy saving is discussed. Finally, the proposed methodology is evaluated by considering synthetic workloads, and a comparative study of different scheduling algorithms is conducted.
In order to improve the host energy efficiency in IaaS, we proposed an adaptive host resource provisioning method, CoST, which is based on QoS differentiation and VM resizing.
The control model can adaptively adjust control parameters according to real time application performance, in order to cope with changes in load.
CoST takes advantage of the fact that different types of applications have different sensitivity degrees to performance and cost. It places two different types of VMs on the same host and dynamically adjusts their sizes based on the load forecasting and QoS feedback. It not only guarantees the performance defined in SLA, but also keeps the host running in energy-efficient state. Real Google cluster trace and host power data are used to evaluate the proposed method. Experimental results show that CoST can provide performance-sensitive application with a steady QoS and simultaneously speed up the overall processing of performance-tolerant application by 20~66%. The host energy efficiency is significantly improved by 7~23%.
The cutting-stock problem is the problem of filling an order at minimum cost for specified numbers of lengths of material to be cut from given stock lengths of given cost. When expressed as an integer programming problem the large number of variables involved generally makes computation infeasible. This same difficulty persists when only an approximate solution is being sought by linear programming. In this paper, a technique is described for overcoming the difficulty in the linear programming formulation of the problem. The technique enables one to compute always with a matrix which has no more columns than it has rows.
In this paper, the methods for stock cutting outlined in an earlier paper in this Journal [Opns Res 9, 849--859 1961] are extended and adapted to the specific full-scale paper trim problem. The paper describes a new and faster knapsack method, experiments, and formulation changes. The experiments include ones used to evaluate speed-up devices and to explore a connection with integer programming. Other experiments give waste as a function of stock length, examine the effect of multiple stock lengths on waste, and the effect of a cutting knife limitation. The formulation changes discussed are i limitation on the number of cutting knives available, n balancing of multiple machine usage when orders are being filled from more than one machine, and m introduction of a rational objective function when customers' orders are not for fixed amounts, but rather for a range of amounts. The methods developed are also applicable to a variety of cutting problems outside of the paper industry.
Virtualization has been proposed as a vehicle for overcoming the growing problem of internet ossification (1). This paper studies the problem of mapping diverse virtual networks onto a common physical substrate. In particular, we develop a method for mapping a virtual network onto a substrate network in a cost-efficient way, while allocating sufficient capacity to virtual network links to ensure that the virtual network can handle any traffic pattern allowed by a general set of traffic constraints. Our approach attempts to find the best topology in a family of backbone-star topologies, in which a subset of nodes constitute the backbone, and the remaining nodes each connect to the nearest backbone node. We investigate the relative cost-effectiveness of different backbone topologies on different substrate networks, under a wide range of traffic conditions. Specifically, we study how the most cost-effective topology changes as the tightness of pairwise traffic constraints and the constraints on traffic locality are varied. In general, we find that as pairwise traffic constraints are relaxed, the least-cost backbone topology becomes increasingly "tree-like". We also find that the cost of the constructed virtual networks is usually no more than 1.5 times a computed lower bound on the network cost and that the quality of solutions improves as the traffic locality gets weaker.
A new approach to the one-dimensional cutting stock problem is described and compared to the classical model for which Gilmore and Gomory have developed a special column-generation technique. The new model is characterized by a dynamic use of simply structured cutting patterns. Nevertheless, it enables the representation of complex combinations of cuts. It can be advantageous in practical applications where many different stock lengths or a relatively large number of order lengths have to be dealt with. The new approach is applied to a real problem where the ″trim loss″ is not valueless, since it can be used for further demands arising in later planning periods.
Recently network virtualization has been proposed as a promising way to overcome the current ossification of the Internet by allowing multiple heterogeneous virtual networks (VNs) to coexist on a shared infrastructure. A major challenge in this respect is the VN embedding problem that deals with efficient mapping of virtual nodes and virtual links onto the substrate network resources. Since this problem is known to be NP-hard, previous research focused on designing heuristic-based algorithms which had clear separation between the node mapping and the link mapping phases. This paper proposes VN embedding algorithms with better coordination between the two phases. We formulate the VN embedding problem as a mixed integer program through substrate network augmentation. We then relax the integer constraints to obtain a linear program, and devise two VN embedding algorithms D-ViNE and R-ViNE using deterministic and randomized rounding techniques, respectively. Simulation experiments show that the proposed algorithms increase the acceptance ratio and the revenue while decreasing the cost incurred by the substrate network in the long run.
This book covers the dominant theoretical approaches to the approximate solution of hard combinatorial optimization and enumeration problems. It contains elegant combinatorial theory, useful and interesting algorithms, and deep results about the intrinsic complexity of combinatorial problems. Its clarity of exposition and excellent selection of exercises will make it accessible and appealing to all those with a taste for mathematics and algorithms.
Richard Karp,University Professor, University of California at Berkeley
Following the development of basic combinatorial optimization techniques in the 1960s and 1970s, a main open question was to develop a theory of approximation algorithms. In the 1990s, parallel developments in techniques for designing approximation algorithms as well as methods for proving hardness of approximation results have led to a beautiful theory. The need to solve truly large instances of computationally hard problems, such as those arising from the Internet or the human genome project, has also increased interest in this theory. The field is currently very active, with the toolbox of approximation algorithm design techniques getting always richer.
It is a pleasure to recommend Vijay Vazirani's well-written and comprehensive book on this important and timely topic. I am sure the reader will find it most useful both as an introduction to approximability as well as a reference to the many aspects of approximation algorithms.
László Lovász, Senior Researcher, Microsoft Research
Inhaltsverzeichnis Introduction.- I. Combinatorial Algorithms: Set cover. Steiner tree and TSP. Multiway cuts and k-cuts. k-center. Feedback vertex set. Shortest superstring. Knapsack. Bin packing. Minimum makespan scheduling. Euclidean TSP.- II. LP-Based Algorithms: Introduction to LP-duality. Rounding applied to set cover. LP-duality based analysis for set cover. The primal-dual schema. Maximum satisfiability. Scheduling on unrelated parallel machines. Multicut and integer multicommodity flow in trees. Multiway cut. Multicut in general graphs. Sparsest cut. Steiner forest. Steiner network. Facility location. k-median. Semidefinite programming.- III. Other Topics: Counting problems. Shortest vector. Hardness of approximation. Open problems.
Due to the existence of multiple stakeholders with conflicting goals and policies, alterations to the existing Internet architecture are now limited to simple incremental updates; deployment of any new, radically different technology is next to impossible. To fend off this ossification, network virtualization has been propounded as a diversifying attribute of the future inter-networking paradigm. By introducing a plurality of heterogeneous network architectures cohabiting on a shared physical substrate, network virtualization promotes innovations and diversified applications. In this paper, we survey the existing technologies and a wide array of past and state-of-the-art projects on network virtualization followed by a discussion of major challenges in this area.