Energy-Efficient Virtual Network Function Reconfiguration Strategy Based on Short-Term Resources Requirement Prediction

In Network Function Virtualization, the resource demand of the network service evolves with the change of network traffic. VNF dynamic migration has become an effective method to improve network performance. However, for the time-varying resource demand, how to minimize the long-term energy consumption of the network while guaranteeing the Service Level Agreement (SLA) is the key issue that lacks previous research. To tackle this dilemma, this paper proposes an energy-efficient reconfiguration algorithm for VNF based on short-term resource requirement prediction (RP-EDM). Our algorithm uses LSTM to predict VNF resource requirements in advance to eliminate the lag of dynamic migration and determines the timing of migration. RP-EDM eliminates SLA violations by performing VNF separation on potentially overloaded servers and consolidates low-load servers timely to save energy. Meanwhile, we consider the power consumption of servers when booting up, which is existing objectively, to avoid switching on/off the server frequently. The simulation results suggest that RP-EDM has a good performance and stability under machine learning models with different accuracy. Moreover, our algorithm increases the total service traffic by about 15% while ensuring a low SLA interruption rate. The total energy cost is reduced by more than 20% compared with the existing algorithms.

Resource Management for Minimizing Energy and Cost of Geo-Distributed Data Centers

Geo-distributed data centers (GDCs) house computing resources and provide cloud services across the world. As cloud computing flourishes, energy consumption and electricity cost for powering servers of GDCs also soar high. Energy consumption and cost minimization for GDCs has become the main challenge for the cloud service providers. This paper proposes a resource management framework that accomplishes resource demand prediction, ensuring service level objective (SLO), electricity price prediction, and energy-efficient and cost-effective resource allocation through GDCs. This paper also proposes an energy-efficient and cost-effective resource allocation (EECERA) algorithm which deploys energy efficiency factors and incorporates the electricity price diversity of GDCs. Extensive evaluations were performed based on real-world workload traces and real-life electricity price data of GDC locations. The evaluation results showed that the resource demand prediction model could predict the right amount of dynamic resource demand while achieving SLO, and also, the electricity price prediction model could provide promising accuracy. The performances of resource allocation algorithms were evaluated on CloudSim. This work contributes to minimizing the energy consumption and the average turnaround time taken to complete the task and offers cost-saving. HIGHLIGHTS SLO guaranteed, energy-efficient and cost-effective resource management framework Energy-efficient and cost-effective resource allocation (EECERA) algorithm Extensive evaluations based on real-world workload traces and real-life electricity price data of GDC locations Performances of resource allocation algorithms evaluated on CloudSim Minimizing the energy consumption and the average turnaround time taken to complete the task and also cost-saving GRAPHICAL ABSTRACT

Analysis of Enterprise Human Resources Demand Forecast Model Based on SOM Neural Network

Human resource planning is the prerequisite of human resource management, and the basic work of human resource planning is to predict human resource demand. Scientific and reasonable human resource demand forecasting results can provide important data support for enterprise human resource planning and strategic decision-making so that human resources management can play a better role in the realization of corporate goals. Because human resource demand is affected by many factors, there is a high degree of nonlinearity and uncertainty between each factor and personnel demand, as well as the incompleteness and inaccuracy of corporate human resource data. In this paper, the self-organizing feature mapping (SOM) artificial neural network prediction model is selected as the prediction model, and the input and output process of sample data is converted into the optimal solution process of the nonlinear function. In the application of the model, the human resource demand prediction index system is used as the input of the SOM neural network and the total number of employees in the enterprise is used as the output so that the problem of nonlinear fitting between human resource demand-influencing factors and human resource demand can be solved. Finally, through the empirical analysis of the enterprise, the model forecasting process is explained and the human resource demand forecast is realized.

SARDE

Resource demands are crucial parameters for modeling and predicting the performance of software systems. Currently, resource demand estimators are usually executed once for system analysis. However, the monitored system, as well as the resource demand itself, are subject to constant change in runtime environments. These changes additionally impact the applicability, the required parametrization as well as the resulting accuracy of individual estimation approaches. Over time, this leads to invalid or outdated estimates, which in turn negatively influence the decision-making of adaptive systems. In this article, we present SARDE , a framework for self-adaptive resource demand estimation in continuous environments. SARDE dynamically and continuously tunes, selects, and executes an ensemble of resource demand estimation approaches to adapt to changes in the environment. This creates an autonomous and unsupervised ensemble estimation technique, providing reliable resource demand estimations in dynamic environments. We evaluate SARDE using two realistic datasets. One set of different micro-benchmarks reflecting different possible system states and one dataset consisting of a continuously running application in a changing environment. Our results show that by continuously applying online optimization, selection and estimation, SARDE is able to efficiently adapt to the online trace and reduce the model error using the resulting ensemble technique.