Survivable Control Plane Establishment With Live Control Service Backup and Migration in SD-EONs

Abstract

It is known that software-defined elastic optical networks (SD-EONs) are programmable and application-aware. However, due to the centralized network control and management, SD-EONs are vulnerable to the network failures that can affect control plane operations. In this paper, we study the problem of survivable control plane establishment (SCPE), i.e., the controller placement and related communication planning for control service backup and migration to protect the control plane of an SD-EON against single node failures. We first propose a novel mutual backup model to improve the survivability of the control plane with high protection efficiency. An integer linear programming (ILP) model is then formulated to solve the SCPE problem with the objective to minimize both the number of deployed OpenFlow controllers (OF-Cs) and the average control channel latency. We also propose a time-efficient heuristic and use simulations to verify that it can obtain similar solutions to those of the ILP. On top of the theoretical investigation, we design and implement the system to facilitate live control service backup and migration with SCPE in an SD-EON control plane testbed. Experimental results demonstrate that the proposed scheme works efficiently, and compared with that using a single OF-C, our scheme achieves much shorter average provisioning latency in dynamic provisioning.Index Terms-Controller placement; Control plane resiliency; Mutual backup; OpenFlow; Software-defined elastic optical networks (SD-EONs).It is known that by decoupling the control and data planes, software-defined networking (SDN) can make the NC&M in optical networks programmable, adaptive, and application-aware [5]. Specifically, software-defined optical networks (SDONs) can typically be realized by leveraging protocols such as the OpenFlow (OF) [6], the network configuration protocol (NETCONF) with YANG as a data modeling language [7], and the path computation element communication protocol (PCEP) [8]. Here, OF is an open standard protocol that incorporates flow-based switching and centralized controller(s) to facilitate software-defined routing, switching, and network management. Moreover, to support software-defined EONs (SD-EONs), the latest OF specification (version 1.5) [6] has included the extensions for identifying flexible-grid optical flows. Hence, by implementing the OF protocol in the control plane, the network operator can manage the data plane elements (i.e., BV-Ts and BV-WSSs) in an EON intelligently with one or more controllers and realize a SD- EON [5,[9][10][11][12].In the meantime, as an optical fiber can carry over Tb/s traffic, ensuring network survivability is vital in optical networks. Hence, researchers have considered the data plane failures in SD-EONs and proposed a few schemes to improve the data plane resiliency [13][14][15]. Note that in a practical SD-EON, the control plane is definitely not failure-proof without specific considerations on the control plane resiliency. For instance, as it counts on the centralized control...

Cuckoo-PC: An Evolutionary Synchronization-Aware Placement of SDN Controllers for Optimizing the Network Performance in WSNs

“…In [28], the authors focused on establishing a survivable control plane for SDN networks by using a novel mutual backup model. Moreover, they formulated the problem as an ILP model and designed a heuristic approach.…”

A Novel Resource Allocation and Spectrum Defragmentation Mechanism for IoT-Based Big Data in Smart Cities

“…SDN support centralized the control of network control and management, giving service providers greater freedom to dynamically define network devices [35]. The authors in [36] proposed a new interoperability backup model to improve the survivability of the control plane in SDN. However, in order to consider survivability, the allocation of spectrum resources needs further research.…”

Multi-Gaussian Schell-model source with a complex coherence state

“…The flat-top MGSM beams have been very well studied: their non-paraxial propagation was discussed in [25]; their electromagnetic extension was introduced in [26]; their versions with a twist phase and a vortex were suggested in [27] and [28], respectively; the arrays with the flat-top MGSM elements were examined in [29]. The interaction of the MGSM beams with crystals [30], deterministic scatterers [31,32], turbulent atmosphere [33] and turbulent ocean [34] is also known. However, all of these studies were concerned with the MGSM beams with real-valued coherence states.…”