基于大数据切片流概率稳定调控机制的云网络数据评测算法

A Cloud Network Data Evaluation Algorithm Based on Large Data Slice Flow Probability Stability Control Mechanism

针对当前云网络数据评测算法普遍存在的评测维度高,切片评测困难,以及在评测过程中存在显著的网络拥塞难题等不足,论文提出了一种基于大数据切片流概率稳定调控机制的云网络数据评测算法。首先,使用切片流随机分发机制,结合云网络中大数据具有的业务带宽切片等特性,实现云网络中数据评测的控制与测量,有效降低了传统方案存在的评测冲突问题;随后,根据梯度覆盖思想,对业务流传输带宽进行二次评测,进一步提高了云网络节点的数据传输能力,降低了数据评测的错误率。仿真数据表明,与当前云网络数据评测中常用的超带宽评估传输算法(Ultra Bandwidth EvaluationTransmission Algorithm,UBET算法)、线性粒度评测传输算法(Linear Granularity Evaluation Transmission Algorithm,LGET算法)相比,论文算法具有更低的评测错误率与更高的切片传输带宽。

In order to solve the evaluation dimension of the ubiquitous cloud network data evaluation algorithm,section evaluation difficulties,lack of significant network congestion problems exist and the evaluation process,this paper proposes a cloud network data flow evaluation of probabilistic stability regulation mechanism based on large data slicing algorithm. The first use of sections flow random distribution method,combined with the big data cloud network with bandwidth slice characteristics,realize data cloud in the network and can be measured,greatly reduces the problem of evaluation conflicts between traditional schemes exist;then according to the gradient overlay thought on business flow transmission bandwidth for the two time evaluation,improve the data the transmission ability of cloud network nodes to further reduce the data error rate evaluation. The simulation time and the surface,the current cloud network evaluation data used in ultra bandwidth evaluation（Ultra Bandwidth Evaluation Transmission algorithm,UBET algorithm）,linear size evaluation（Linear Granularity Evaluation Transmission algorithm,LGET algorithm）,this algorithm has the advantage of evaluating performance significantly,can significantly improve the network bandwidth and improve transmission a node running stability,it is of significant practical value.