Research On The Network Node Energy Saving Problem Based On Reconfigure Technology

The internet has been one of important infrastructure to support Modern social and economic development, social progress and scientific and technological innovation. The Internet to meet people on the network size, functionality and performance needs, while also gradually revealed some problems. In energy consumption, information and communication technology industry as one of the fastest growing industries worldwide, Carbon emissions are also growing with the growth of the industry. The carbon emissions of information and communication technology accounted for2%of the world, which will be double in2020. Network infrastructure, including routers, servers, switches, cooling facilities and other equipment consume868billion kWh of electricity in2008, Accounting for5.3%of the world’s total electricity consumption. With the current growth trends, in2025, The average energy consumption of the IT industry will reach five times that of2006, network areas will reach13times. Energy consumption has become a major obstacle to sustainable development in information and communication technology. National863Program IT major projects in the field of "A new generation of high-trusted network" proposed the reconfigurable flexible network idea. Committed to building capable of carrying a new business, provide a reliable service support to QoS, user scale planning, network resources can be managed, node service capabilities can reconfigure, building a green next-generation networks and services national test bed. On the demand of Green Network, this article relies on research of reconfigurable flexible network, studying the energy-saving technologies of the network nodes.In this paper, we proposed an idea of node device energy consumption following the traffic load changes. Adjusting node energy consumption can be abstract to the physical nature of the resource allocation and resource adjustment. First, study reconfigurable theory based component computing system, second, study the energy consumption of the component perceptual model. Third, research a fine-grained adjustment method component-based reconstruction of the energy consumption. Last, design a low power scheme for the network nodes forwarding. Specifically, the main findings of this article are as follows:This paper focuses on the lacks at methodology of describing formal model and reconfigurable attributes of the reconfigurable system, abstracts and describes the attributes and the behaviors of reconfigurable system, component combination and reconfigurable method by algebraic method. By understanding the component combination an operation, that is a new idea, and extending the calculus in process algebraic, some component combination operators are defined and then a formal algebraic model of reconfigurable system is proposed. Based this model some reconfigurable attributes are analyzed and a few reconfigurable nomal formats is proposed. All above viewpoints construct theoretical footstone for designing reconfigurable system.This paper focuses on the energy-aware problem of components, starts at a software component architecture level, considers the functional relation between software component characteristic quantities and software energy as nonlinear (linear functional relation can be considered as a special nonlinear functional relation). Next, the paper presents an energy model at architecture level by using BP neural network. The energy model measures three software characteristic quantities at architecture level and uses BP neural network to fit the functional relation between software characteristic quantities and software energy. Experimental results show that this model is effective. Based on the energy aware model, energy aware component model is proposed. In this model, such as energy aware, decision, and execution are into components, and then supports their dynamic configuration with the reconfigure technology. Based on principles of the Separation of Concerns and the Dynamic Software Architecture (DSA) technology, the component model can help components running and reconfiguring.For the delay-sensitive traffic, paper introduce the frequency scaling consistency and the algorithm assisted with could make dynamic frequency scaling by automatically inserting dynamic frequency scaling source code in presented. The algorithm has been realized in Zebra container prototype and Wattch emulator software. The results show that we can save20%-40%energy consumption of components. The meticulous-grained device energy scaling is implemented to validate that the components energy can changed with application load variety.In order to solve the energy saving problem of no-delay-sensitive traffic, a multiscale reconfigure decision algorithm is proposed. The effective model of self-similarity or multiracial traffic was obtained using the Hurst index; the effective envelope is obtained using the effective bandwidth and effective envelope theorem. Then, traffic parameters are obtained using network calculus to perform reconfigures decision. The simulations prove that the traffic model can be selected accurately; the rate of success and accurate of reconfigure are improved; the effect to traffic delay is reduced. This paper proposes an energy meticulous-grained scaling algorithm to deal with the routers’energy problem, which is based on the idea that components energy can changed with application load variety. It first predicts the future workloads of the applications with Brown’s quadratic exponential smoothing method to make reconfiguration catch up with loads. Next, it adopts a genetic algorithm to parallel find the optimal reconfiguration policy. The real network traffic is used to check the algorithm. Experimental results demonstrate the approach can adapt the router’s energy according to the change of network traffic, save the algorithm computing time60%and greatly reduce the energy consumption25%.We focus on the low-energy architecture designing problem of reconfigurable flexible network forwarding plane. In this paper, we consider building a FIS mechanism based forwarding in switching idea. The memory requiring and accessing reduced by blurredly forwarding and piping switching. The multilevel line framework in FIS is made by many slow processing units. It can obtain higher performance and lookup speed through cosmically running. Simultaneously, it can reduce the complexity of hardware implementation. Energy model of FIS and FBS is established. Experiments using real-life routing tables demonstrate that our FIS solution can reduce the power of routing lookup12.5%than FBS clearly.The conclusions of this thesis offer a fine-grained adjustment method component-based reconstruction of the energy consumption which contains the energy resource allocation and resource adjustment, which is great valuable to accurately grasp the macro features of component reconfigurable system. And it can provide theoretical foundation and research model directly for the research on the green network based on the idea of node device energy consumption following the traffic load changes.