Accurate and efficient measurements of IP level performance to drive interface selection in heterogeneous wireless networks

摘要

Optimal interface selection is a key mobility management issue inheterogeneous wireless networks. Measuring the physical or link levelperformance on a given wireless access networks does not provide a reliableindication of the actually perceived level of service. It is therefore neededto take measurements at IP level, on the (bidirectional) paths from the MobileHost to the node that is handling the mobility, over different heterogeneousnetworks. In this paper, we propose and analyze mechanisms for connectivitycheck and performance (network delay and packet loss) monitoring over IP accessnetworks, combining active and passive monitoring techniques. We evaluate theaccuracy and timeliness of the performance estimates and provide guidelines fortuning up the parameters. From the implementation perspective, we show thatusing application level measurements is highly CPU intensive, while a kernelbased implementation has comparably a very low CPU usage. The Linux kernelimplementation results in an efficient use of batteries in Mobile Hosts andintermediate Mobility Management Nodes can scale up to monitoring thousands offlows. The proposed solutions have been implemented in the context of aspecific mobility management solution, but the results are of generalapplicability. The Linux implementation is available as Open Source.