Imperial College London
Portrait Picture

Dr Ed Cohen

Faculty of Natural Sciences, Department of Mathematics

Reader in Statistics
 
 
 
//

Contact

 

+44 (0)20 7594 3986e.cohen

 
 
//

Location

 

536Huxley BuildingSouth Kensington Campus

//

Summary

 

Publications

Citation

BibTex format

@inproceedings{Gibberd:2018,
author = {Gibberd, A and Nobel, J and Cohen, E},
publisher = {ITISE},
title = {Characterising dependency in computer networks using spectral coherence},
url = {http://hdl.handle.net/10044/1/62632},
year = {2018}
}
Download

RIS format (EndNote, RefMan)

TY  - CPAPER
AB  - The  quantification  of  normal  and  anomalous  traffic  flowsacross computer networks is a topic of pervasive interest in network se-curity, and requires the timely application of time-series methods. Thetransmission or reception of packets passing between computers can berepresented in terms of time-stamped events and the resulting activityunderstood in terms of point-processes. Interestingly, in the disparate do-main  of  neuroscience,  models  for  describing  dependent  point-processesare  well  developed.  In  particular,  spectral  methods  which  decomposesecond-order  dependency  across  different  frequencies  allow  for  a  richcharacterisation  of  point-processes.  In  this  paper,  we  investigate  usingthe spectral coherence statistic to characterise computer network activ-ity, and determine if, and how, device messaging may be dependent. Wedemonstrate on real data, that for many devices there appears to be verylittle dependency between device messaging channels. However, when sig-nificant coherence is detected it appears highly structured, a result whichsuggests coherence may prove useful for discriminating between types ofactivity at the network level.
AU  - Gibberd,A
AU  - Nobel,J
AU  - Cohen,E
PB  - ITISE
PY  - 2018///
TI  - Characterising dependency in computer networks using spectral coherence
UR  - http://hdl.handle.net/10044/1/62632
ER  -
Download