基于粗糙K均值的服务质量相关弹性流聚集

面对多变的网络环境，现有的网络服务质量(QoS)映射中流聚集方法缺乏灵活性。针对现有聚集方法的缺陷，该文提出一种动态聚集方法。使用增强粗糙K均值算法(ERKM)，按照网络流的QoS属性将网络流进行合理聚集，并且在网络处于高负载状况时，通过隶属度弹性聚集网络流，从而适应网络的变化，使得网络流聚集具有灵活性。最后进行了网络流聚集实验和调度实验。实验表明，相比于现有的方法，该方法能够更加弹性地应对不同网络状态，并且更好地保障网络流的QoS指标。此外，还进一步验证了该文方法在不同网络环境下的QoS类聚集的一致性。

Quality of Service-aware Elastic Flow Aggregation Based on Enhanced Rough K-Means

Facing changeable network environment, current Quality of Service (QoS)-aware flow aggregation scheme is lack of flexibility. A dynamic flow aggregation method to overcome present problems is proposed. An Enhanced Rough K-Means (ERKM) algorithm is used to aggregate network flows properly. Importantly, it is able to adjust degree of membership to face ever-changing internet environment to make algorithm more flexible. Internet scheduler experiment is carried out and a comparison is made with existing methods. Experimental results suggest that proposed method has advantages not only on flexibility of aggregation, but also on assurance of QoS of Internet flows. In addition, the consistency of QoS allocation under different network environment is investigated.