Data streaming algorithms for efficient and accurate estimation of flow size distribution

By Abhishek Kumar, Minho Sung, Jun Xu, Jia Wang. SIGMETRICS '04/Performance '04, 2004

315 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications

View Paper: https://www.semanticscholar.org/paper/3d48d4a133f6268a013d34ff9ec333950af7200c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Qun Huang
2. Ian F. Akyildiz
3. Vyas Sekar
4. Peng Liu
5. Vladimir Braverman
6. Rui Miao
7. D. Andersen
8. Xiaoming Li

In Computer Networks and Communications

1. Vyas Sekar
2. Qun Huang
3. Ian F. Akyildiz
4. Min Luo
5. Peng Liu
6. Ahyoung Lee
7. Vladimir Braverman
8. Rui Miao

Citations in the corpus

• Elastic sketch: adaptive and fast network-wide measurements (2018, methodology): “• WMRE (Weighted Mean Relative Error) [12, 17]: ∑z i=1 |ni−n̂i |∑z i=1( ni +n̂i 2 ) , where z is the maximum flow size, and ni and n̂i are the true and estimated numbers of flows of size i respectively.”
• SketchVisor: Robust Network Measurement for Software Packet Processing (2017, unknown): “In contrast, SketchVisor achieves over 17Gbps for all sketches, and almost 40Gbps for MRAC in particular.”
• Software Defined Traffic Measurement with OpenSketch (2013, unknown): “Picking packets with different granularity: Many streaming algorithms (such as flow size distribution [27], counting distinct destinations with the multiresolution bitmap [21], and latency tracking with arbitrary loss [26]) often make tradeoffs between flow space coverage and accuracy.”
• One Sketch to Rule Them All: Rethinking Network Flow Monitoring with UnivMon (2016, background): “Each such management task requires accurate and timely statistics on different application-level metrics of interest; e.g., the flow size distribution [37], heavy hitters [10], entropy measures [38,50], or detecting changes in traffic patterns [44].”
• An empirical evaluation of entropy-based traffic anomaly detection (2008, methodology): “The flow-header features are addresses (source and destination), ports (source and destination), and the flow size distribution (FSD) [10, 5, 9 ]. The behavioral features are the in and out-degree distributions (degree of an end-host X is the number of distinct IP addresses that X communicates with”
• A roadmap for traffic engineering in SDN-OpenFlow networks (2014, methodology): “Sketches can be used for many measurement programs such as heavy hitters [99,100], the traffic change detection [101], the flow size distribution estimation [102], the global iceberg detection [103], and the fine-grained delay measurement [104].”