Empowered by the optical orthogonal frequency-division multiplexing (O-OFDM) technology, flexible online service provisioning can be realized with dynamic routing, modulation, and spectrum assignment (RMSA). In this paper, we propose several algorithms that incorporate RMSA a hybrid single-/multi-path routing (HSMR) scheme. We investigate two types of HSMR schemes, namely using path computation (HSMR-OPC) fixed sets (HSMR-FPS). Moreover, for HSMR-FPS, analyze selection policies to optimize...

This article discusses the technologies for realizing highly efficient data migration and backup big applications in elastic optical inter-data-center (inter-DC) networks. We first describe impacts of on underlying network infrastructure introduce concept flexible-grid inter-DC Then we model such networks as dynamic anycast propose several algorithms. Joint resource defragmentation is also discussed to further improve performance. For backup, leverage a mutual investigate how avoid prolonged...

Based on the concept of infrastructure as a service, optical network virtualization can facilitate sharing physical among different users and applications. In this paper, we design algorithms for both transparent opaque virtual embedding (VONE) over flexible-grid elastic networks. For VONE, first formulate an integer linear programming (ILP) model that leverages all-or-nothing multi-commodity flow in graphs. Then, to consider continuity consecutiveness substrate fiber links' (SFLs') spectra,...

Recently, optical orthogonal frequency-division multiplexing technology has attracted intensive research interest because spectrum-sliced elastic networks (EONs) can be constructed based on it. In this paper, we investigate how to serve multicast requests over EONs with multicast-capable routing, modulation level, and spectrum assignment (RMSA). Both EON planning static traffic provisioning dynamic are studied. For planning, formulate two integer linear programming (ILP) models, i.e., the...

This paper investigates the spectrum fragmentation issue, which undermines bandwidth efficiency in elastic optical networks. After categorizing two-dimensional problem as and misalignment subproblems, this proposes joint routing assignment (RSA) algorithms to alleviate spectral lightpath provisioning process. The time complexity of two proposed are analyzed detail, both can run O(kdnC log C) time, where k is number shortest path algorithm, d maximum node degree network, n nodes C link...

Network function virtualization (NFV) is a promising technology to decouple the network functions from dedicated hardware elements, leading significant cost reduction in service provisioning. As more and users are trying access their services wherever whenever, we expect NFV-related chains (SFCs) be dynamic adaptive, i.e., they can readjusted adapt requests' dynamics for better user experience. In this paper, study how optimize SFC deployment readjustment situation. Specifically, try jointly...

This paper proposes DeepRMSA, a deep reinforcement learning framework for routing, modulation and spectrum assignment (RMSA) in elastic optical networks (EONs). DeepRMSA learns the correct online RMSA policies by parameterizing with neural (DNNs) that can sense complex EON states. The DNNs are trained experiences of dynamic lightpath provisioning. We first modify asynchronous advantage actor-critic algorithm present an episode-based training mechanism namely, DeepRMSA-EP. DeepRMSA-EP divides...

In this paper, after proposing a novel metric, i.e., global resource capacity (GRC), to quantify the embedding potential of each substrate node, we propose an efficient heuristic virtual network (VNE) algorithm, called as GRC-VNE. The proposed algorithm aims maximize revenue and minimize cost infrastructure provider (InP). Based on GRC, applies greedy load-balance manner embed node sequentially, then adopts shortest path routing link. Simulation results demonstrate that our GRC-VNE achieves...

This paper tries to solve the location-constrained virtual network embedding (LC-VNE) problem efficiently. We first investigate complexity of LC-VNE, and by leveraging graph bisection problem, we provide formal proof NP-completeness inapproximability result LC-VNE. Then, propose two novel LC-VNE algorithms based on a compatibility (CG) achieve integrated node link mapping. In particular, in CG, each represents candidate substrate path for link, indicates compatible relation between its...

Elastic optical networks (EONs) enable network operators to have agile bandwidth management in the layer. In this paper, we investigate dynamic and adaptive defragmentation (DF) EONs with time-varying traffic using connection reconfigurations. We consider how design DF procedure a systematic way, study problems that not been fully explored so far. Basically, divide into four subproblems: “How reconfigure?,” “ How migrate traffic?,” “When “What solve them sequentially. For combination of...

The elastic optical network (EON) based on the orthogonal frequency-division multiplexing (O-OFDM) technology is a promising candidate for accommodating uncertainty and heterogeneity of traffic across future Internet. Advance reservation (AR) essential to support initial-delay-tolerant services, such as e-science grid computing, which are making significant contributions Internet traffic. Therefore, we expect that it necessary EONs AR requests. In this paper, study dynamic service...

We propose a novel two-population genetic algorithm (MPGA) to optimize the routing, modulation and spectrum assignments (RMSA) in optical orthogonal frequency-division multiplexing (O-OFDM) networks. The proposed MPGA makes two populations evolve parallel with different selection mutation strategies, incorporates migration operation exchange individuals between them. Performance evaluations show that outperforms several existing algorithms.

It is known that by incorporating network function virtualization (NFV) in inter-datacenter (inter-DC) networks, service providers can use their resources more efficiently and adaptively expedite the deployment of new services. This paper studies provisioning algorithms to realize tree-type virtual forwarding graphs (VNF-FGs), i.e., multicast NFV trees (M-NFV-Ts), inter-DC elastic optical networks (IDC-EONs) cost-effectively. Specifically, we try optimize VNF placement routing spectrum...

Software-defined networking separates the control and forwarding planes of a network to make it more programmable application- aware. As one initial implementations SDN, OpenFlow abstracts device as flow table realizes processing by applying "match-and-act" principle. However, protocol-dependent nature still limits programmability plane. Hence, in this article, we discuss how leverage protocol-oblivious (POF) further enhance such that plane becomes protocol-independent can be dynamically...

The combination of elastic optical networks (EONs) and software-defined networking (SDN) leads to SD-EONs, which bring a new opportunity for enhancing programmability flexibility with more freedom network operators customize their infrastructure dynamically. In this paper, we investigate how apply multidomain scenarios SD-EONs. We design the functionalities in control plane facilitate tasks, propose an interdomain protocol enable OpenFlow controllers different SD-EON domains operate...

In elastic optical networks (EONs), bandwidth fragmentation refers to the existence of non-aligned, isolated and small-sized blocks contiguous subcarrier slots in spectrum. As they are neither spectrum domain nor aligned along routing paths, network operator will have difficulty use these for future connections. this work, we analyze effect on blocking probability EONs. Our theoretical analysis indicates that two factors related effects probability: 1) extent available slot-blocks (i.e.,...

We study how to allocate spectrum and IT resources jointly for realizing efficient virtual network function (VNF) service chaining in inter-datacenter elastic optical networks. first formulate an integer linear programming model solve the problem exactly, then a longest common subsequence-based heuristic is proposed. The simulation results indicate that proposed algorithms can reuse deployed VNFs efficiently arrange utilization much more load-balanced manner.

This paper studies the problem of service chain (SC) deployment. Specifically, we try to place virtual network functions (vNFs) on nodes and connect vNFs in sequence through link mapping. We start with offline problem. An integer linear programming (ILP) model is formulated minimize total SC deployment cost. With ILP, prove that NP-hard propose a time-efficient heuristic based affiliation-aware vNF placement. Then, move online problem, design forecast-assisted algorithm includes prediction...

This paper proposes a self-taught anomaly detection framework for optical networks. The proposed makes use of hybrid unsupervised and supervised machine learning scheme. First, it employs an data clustering module (DCM) to analyze the patterns monitoring data. DCM enables self-learning capability that eliminates requirement prior knowledge abnormal network behaviors therefore can potentially detect unforeseen anomalies. Second, we introduce mechanism transfers learned by regression...

This work addresses the relatively long setup latency and complicated network control management caused by on-demand virtual function service chain (vNF-SC) provisioning in inter-datacenter elastic optical networks. We first design a framework with resource pre-deployment to resolve aforementioned challenge. Specifically, is designed as discrete-time system, which operations are performed periodically fixed time slots (TS). Each TS includes phase followed phase. In phase, deep-learning (DL)...

This paper demonstrates Deep-RMSA, a deep reinforcement learning based self-learning RMSA agent that can learn successful policies from dynamic network operations while realizing cognitive and autonomous in EONs.

We investigate the principle of how dynamic service provisioning fragments spectral resources on links along a path, and propose corresponding RMSA algorithms to alleviate spectrum fragmentation in network environments.

In an optical inter-datacenter (inter-DC) network, for preventing data loss, a cloud system usually leverages multiple DCs obtaining sufficient redundancy. order to improve the data-transfer efficiency of regular DC backup, this paper investigates fast and coordinated backup in geographically distributed (geo-distributed) inter-DC networks. By considering mutual model, which can serve as sites each other, we study how finish within shortest time duration (i.e., window (DC-B-Wnd))....

Bandwidth defragmentation, i.e., the operation to reconfigure existing connections for making spectrum usage less fragmented and misaligned, has recently been recognized as one of most important features elastic optical networks (EONs). In this paper, we propose a novel comprehensive bandwidth defragmentation algorithm that considers problems 1) When defragment? 2) What 3) How jointly. The proposed accomplishes through proactive network reconfiguration only reroutes portion connections. each...

Optical orthogonal frequency-division multiplexing (O-OFDM) technology has the elastic feature of allocating spectrum resources based on subcarrier slots with bandwidths at a few GHz or even narrower. This enables us to utilize link capacity more efficiently by splitting connection's traffic over multiple routing paths. In this paper, we propose novel dynamic multi-path provisioning algorithm for O-OFDM optical networks. The tries set up connections single-path in best-effort manner. When...