面向低轨卫星网络的星上分布式Sketch测量框架OA

Frequent handovers in low earth orbit(LEO)satellite networks fragment traffic and complicate the tracking and aggregation of long flows.To address this issue,a distributed Sketch-based measurement framework was proposed.The framework employed a distributed hash table to record the satellite trajectory of each long flow.During a query,it first located the trajectory sequence and then aggregated the traffic fragments collected from the access satellites.A coop-erative architecture consisting of a satellite data plane and a ground control plane was designed.The satellite side de-ployed Sketch for local traffic counting and used a Bloom filter together with a distributed hash table to detect the first packet and record trajectory information.The ground side supported on-demand queries,result aggregation,and param-eter configuration,and it exploited the linearity of Sketch to efficiently merge distributed traffic fragments.Experimental results demonstrate that,in a typical scenario of a 792-satellite Walker constellation with 3 measurement satellites tra-versed and 1 200 cross-satellite long flows,the proposed scheme achieves an average end-to-end query latency of 128 ms and a per-satellite storage overhead of less than 100 KB.