Software-Defined Cross-Layer Optical Networks Control Technology Research

With the advent of big data era, data centers are widely deployed. Large data center resource flood in the network. Current network is facing problems of complex control and inefficient use of resources. However, inter network is urgent need in order to make full use of the data center resource. Cross-layer control technology which support coordinated scheduling between data center and network resources can greatly improve the overall network performance. The software-defined network can establish intelligent optical network control functions, provide dynamic flexible intelligence services and more accurate way to utilize and allocate resources convenient.Based on the above situation, the network control architecture and resource scheduling strategy has carried out in-depth research on software-defined optical cross-layer network control technology in this paper. This paper has designed and implemented the new software-defined optical cross-layer network control architecture and introduce the time factor into cross-layer control techniques so that the resource granularity is smaller and control architecture is more intelligent to enhance the network performance. The main work and achievement are described as below.This paper designs a new module of cross-layer control architecture based on the optimization of both application resources and network resources. Experiment results show that application plane, control plane and transport plane can delivery and process service intelligent by dynamically managing data center and network virtual resources. The core controller of software-defined optical cross-layer network control architecture is designed and implemented in this paper, including the application controller of time-resource double elastic resource scheduling and programmed enhance NOX Controller. Experiments show that this architecture can greatly reduce the complexity of network control, improve the resource scheduling intelligence and service flexibility.This paper proposes a flexible and elastic resource scheduling strategy which focuses on the application and network resources in flexi-grid optical network. A time factor is introduced into the scheduling of whole network resources so that resources allocation and management of the whole network are more accurate and efficient. Simulation results show that the strategy improves the network data processing capability and the balance of whole network resources. It also can reduce the network congestion rate and resource utilization.