The objectives of this survey are to provide an in-depth coverage of a few selected research papers that have made significant contributions to the development of Network Function Virtualization (NFV), and to provide readers insights into the key advantages and disadvantages of NFV and Software Defined Networks (SDN) when compared to traditional networks. The research papers covered are classified into four categories: NFV Infrastructure (NFVI), Network Functions (NFs), Management And Network Orchestration (MANO), and service chaining. The NFVI papers describe “framework” software that implement common functions, such as dynamic scaling and load balancing, required by NF developers. Papers on NFs are classified as offering solutions for software switches or middleboxes. MANO papers covered in this survey are primarily on resource allocation (virtual network embedding), which is an orchestrator function. Finally, service chaining papers that offer examples and extensions are reviewed. Our conclusions are that with the current level of investment in NFV from cloud and Internet service providers, the promised cost savings are likely to be realized, though many challenges remain.

Predictive spatial-monitoring, which predicts spatial information such as road traffic, has attracted much attention in the context of smart cities. Machine learning enables predictive spatial-monitoring by using a large amount of aggregated sensor data. Since the capacity of mobile networks is strictly limited, serious transmission delays occur when loads of communication traffic are heavy. If some of the data used for predictive spatial-monitoring do not arrive on time, prediction accuracy degrades because the prediction has to be done using only the received data, which implies that data for prediction are ‘delay-sensitive’. A utility-based allocation technique has suggested modeling of temporal characteristics of such delay-sensitive data for prioritized transmission. However, no study has addressed temporal model for prioritized transmission in predictive spatial-monitoring. Therefore, this paper proposes a scheme that enables the creation of a temporal model for predictive spatial-monitoring. The scheme is roughly composed of two steps: the first involves creating training data from original time-series data and a machine learning model that can use the data, while the second step involves modeling a temporal model using feature selection in the learning model. Feature selection enables the estimation of the importance of data in terms of how much the data contribute to prediction accuracy from the machine learning model. This paper considers road-traffic prediction as a scenario and shows that the temporal models created with the proposed scheme can handle real spatial datasets. A numerical study demonstrated how our temporal model works effectively in prioritized transmission for predictive spatial-monitoring in terms of prediction accuracy.

The machine-to-machine (M2M) service network platform that accommodates and controls various types of Internet of Things devices has been presented. This paper investigates program file placement strategies for the M2M service network platform that achieve low blocking ratios of new task requests and accommodate as many tasks as possible in the dynamic scenario. We present four strategies for determining program file placement, which differ in the computation method and whether the relocation of program files being used by existing tasks is allowed or not. Simulation results show that a strategy based on solving a mixed-integer linear programming model achieves the lowest blocking ratio, but a heuristic algorithm-based strategy can be an attractive option by allowing recomputation of the placement when the placement cannot be obtained at the timing of new task request arrival.

The Machine-to-Machine (M2M) service network platform accommodates M2M communications traffic efficiently by using tree-structured networks and the computation resources deployed on network nodes. In the M2M service network platform, program files required for controlling devices are placed on network nodes, which have different amounts of computation resources according to their position in the hierarchy. The program files must be dynamically repositioned in response to service quality requests from each device, such as computation power, link bandwidth, and latency. This paper proposes a Program File Placement (PFP) method for the M2M service network platform. First, the PFP problem is formulated in the Mixed-Integer Linear Programming (MILP) approach. We prove that the decision version of the PFP problem is NP-complete. Next, we present heuristic algorithms that attain sub-optimal but attractive solutions. Evaluations show that the heuristic algorithm based on the number of devices that share a program file reduces the total number of placed program files compared to the algorithm that moves program files based on their position.

The Software-defined network (SDN) uses a centralized SDN controller to store flow entries in the flow table of each SDN switch; the entries in the switch control packet flows. When a multicast service is provided in an SDN, the SDN controller stores a multicast entry dedicated for a multicast group in each SDN switch. Due to the limited capacity of each flow table, the number of flow entries required to set up a multicast tree must be suppressed. A conventional multicast routing scheme suppresses the number of multicast entries in one multicast tree by replacing some of them with unicast entries. However, since the conventional scheme individually determines a multicast tree for each request, unicast entries dedicated to the same receiver are distributed to various SDN switches if there are multiple multicast service requests. Therefore, further reduction in the number of flow entries is still possible. In this paper, we propose a multicast routing model for multiple multicast requests that minimizes the number of flow entries. This model determines multiple multicast trees simultaneously so that a unicast entry dedicated to the same receiver and stored in the same SDN switch is shared by multicast trees. We formulate the proposed model as an integer linear programming (ILP) problem. In addition, we develop a heuristic algorithm which can be used when the ILP problem cannot be solved in practical time. Numerical results show that the proposed model reduces the required number of flow entries compared to two benchmark models; the maximum reduction ratio is 49.3% when the number of multicast requests is 40.

This paper proposes a defragmentation scheme using reroutable backup paths in toggled-based quasi 1+1 path protected elastic optical networks (EONs) to improve the efficiency of defragmentation and suppress the fragmentation effect. The proposed scheme can reallocate spectrum slots of backup paths and reroute of backup paths. The path exchange function of the proposed scheme makes the primary paths become the backup state while the backup paths become the primary. This allows utilization of the advantages of defragmentation in both primary and backup paths. We formulate a static spectrum reallocation problem with rerouting (SSRR) in the toggled-based quasi 1+1 path protected EON as an integer linear programming (ILP) problem. The decision version of SSRR is proven to be an NP-complete problem. A heuristic algorithm is introduced to solve the problem for large networks networks where the ILP problem is not tractable. For a dynamic traffic scenario, an approach that suppresses the fragmentation considering rerouting and path exchanging operations is presented. We evaluate the performances of the proposed scheme by comparing it to the conventional scheme in terms of dependencies on node degree, processing time of network operations and interval time between scheduled defragmentations. The numerical results obtained from the performance evaluation indicate that the proposed scheme increases the traffic admissibility compared to the conventional scheme.

This paper proposes a scalable active optical access network using high-speed Plumbum Lanthanum Zirconate Titanate (PLZT) optical switch/splitter. The Active Optical Network, called ActiON, using PLZT switching technology has been presented to increase the number of subscribers and the maximum transmission distance, compared to the Passive Optical Network (PON). ActiON supports the multicast slot allocation realized by running the PLZT switch elements in the splitter mode, which forces the switch to behave as an optical splitter. However, the previous ActiON creates a tradeoff between the network scalability and the power loss experienced by the optical signal to each user. It does not use the optical power efficiently because the optical power is simply divided into 0.5 to 0.5 without considering transmission distance from OLT to each ONU. The proposed network adopts PLZT switch elements in the variable splitter mode, which controls the split ratio of the optical power considering the transmission distance from OLT to each ONU, in addition to PLZT switch elements in existing two modes, the switching mode and the splitter mode. The proposed network introduces the flexible multicast slot allocation according to the transmission distance from OLT to each user and the number of required users using three modes, while keeping the advantages of ActiON, which are to support scalable and secure access services. Numerical results show that the proposed network dramatically reduces the required number of slots and supports high bandwidth efficiency services and extends the coverage of access network, compared to the previous ActiON, and the required computation time for selecting multicast users is less than 30 msec, which is acceptable for on-demand broadcast services.

The energy consumption of network virtualization environments (NVEs) has become a critical issue. In this paper, we focus on reducing the data switching energy consumption of NVE. We first analyze the data switching energy of NVE. Then, we propose a dynamic energy efficient virtual link resource reallocation (eEVLRR) approach for NVE. eEVLRR dynamically reallocates the energy efficient substrate resources (s-resources) for virtual links with dynamic changes of embeddable s-resources to save the data switching energy. In order to avoid traffic interruptions while reallocating, we design a cross layer application-session-based forwarding model for eEVLRR that can identify and forward each data transmission flow along the initial specified substrate data transport path until end without traffic interruptions. The results of performance evaluations show that eEVLRR not only guarantees the allocated s-resources of virtual links are continuously energy efficient to save data switching energy but also has positive impacts on virtual network acceptance rate, revenues and s-resources utilization.

Software defined networking (SDN) and OpenFlow, which enables the abstraction of vendor/technology-specific attributes, improve the control and management flexibility of optical transport networks. In this paper, we present an interoperability demonstration of SDN/OpenFlow-based optical path control for multi-domain/multi-technology optical transport networks. We also summarize the abstraction approaches proposed for multi-technology network integration at SDN controllers.

One of the features of a software-defined network (SDN) is a logically centralized control plane hosting one or more SDN controllers. As SDN controller placement can impact network performance, it is widely studied as the controller placement problem (CPP). For a cost-effective network design, network providers need to minimize the number of SDN controllers used in the network since each SDN controller incurs installation and maintenance costs. Moreover, the network providers need to deal with the failure of SDN controllers. Existing studies that consider SDN controller failures use the scheme of connecting each SDN switch to one Master controller and one or more Slave controllers. The problem with this scheme is that the computing capacity of each SDN controller cannot be used efficiently since one SDN controller handles the load of all SDN switches connected to it. The number of SDN controllers required can be reduced by distributing the load of each SDN switch among multiple SDN controllers. This paper proposes a controller placement model that allows the distribution against SDN controller failures. The proposed model determines the ratios of computing capacity demanded by each SDN switch on the SDN controllers connected to it. The proposed model also determines the number and placement of SDN controllers and the assignment of each SDN switch to SDN controllers. Controller placement is determined so that a network provider can continue to manage all SDN switches if no more than a certain number of SDN controller failures occur. We develop two load distribution methods: split and even-split. We formulate the proposed model with each method as integer linear programming problems. Numerical results show that the proposed model reduces the number of SDN controllers compared to a benchmark model; the maximum reduction ratio is 38.8% when the system latency requirement between an SDN switch and an SDN controller is 100[ms], the computing capacity of each SDN controller is 6 × 106[packets/s], and the maximum number of SDN controllers that can fail at the same time is one.

Multi-core fiber (MCF) is one of the promising space-division multiplexing technologies to increase the capacity of optical networks. MCF-based networks have two challenges. One is the inter-core crosstalk (XT) that degrades the quality of optical signals in two neighboring fiber cores. The other is network protection against link failures that cause massive data loss. One way to protect against multiple link failures is to prepare physically separated links as a backup network. Probabilistic protection improves the efficiency of protection by allowing a certain probability of protection failure. Existing studies on backup network design with probabilistic protection do not target MCF-based networks, which raises problems such as protection failure due to the inter-core XT and excessive consumption of optical resources. To address these problems, this paper proposes a XT-aware backup network design model for the MCF optical path networks. The proposed model protects the network against probabilistic multiple link failures. We adopt probabilistic protection that allows a certain probability of protection failure due to the inter-core XT and minimizes the required number of links in the backup network. We present an algorithm to satisfy the probabilistic protection requirement and formulate the model as an integer linear programming problem. We develop a heuristic approach to apply the proposed model to larger networks. Numerical results observe that the proposed model requires fewer links than the dedicated allocation model, which provisions the backup paths in the same manner as the primary paths.

Mobile edge computing (MEC) is a key technology for providing services that require low latency by migrating cloud functions to the network edge. The potential low quality of the wireless channel should be noted when mobile users with limited computing resources offload tasks to an MEC server. To improve the transmission reliability, it is necessary to perform resource allocation in an MEC server, taking into account the current channel quality and the resource contention. There are several works that take a deep reinforcement learning (DRL) approach to address such resource allocation. However, these approaches consider a fixed number of users offloading their tasks, and do not assume a situation where the number of users varies due to user mobility. This paper proposes Deep reinforcement learning model for MEC Resource Allocation with Dummy (DMRA-D), an online learning model that addresses the resource allocation in an MEC server under the situation where the number of users varies. By adopting dummy state/action, DMRA-D keeps the state/action representation. Therefore, DMRA-D can continue to learn one model regardless of variation in the number of users during the operation. Numerical results show that DMRA-D improves the success rate of task submission while continuing learning under the situation where the number of users varies.