Multiwavelet based-approach to detect shared congestion in computer networks | ||
Journal of Engineering | ||
Article 1, Volume 18, Issue 11, November 2012, Pages 1219-1228 | ||
Authors | ||
Tarik Zeyad Ismaeel; Ahmed A. Mahdi A. Kareem | ||
Abstract | ||
Internet paths sharing the same congested link can be identified using several shared congestion detection techniques. The new detection technique which is proposed in this paper depends on the previous novel technique (delay correlation with wavelet denoising (DCW) with new denoising method called Discrete Multiwavelet Transform (DMWT) as signal denoising to separate between queuing delay caused by network congestion and delay caused by various other delay variations. The new detection technique provides faster convergence (3 to 5 seconds less than previous novel technique) while using fewer probe packets approximately half numbers than the previous novel technique, so it will reduce the overload on the network caused by probe packets. Thus, new detection technique will improve the overall performance of computer network. | ||
Keywords | ||
Congestion; computer network; shared congestion; DMWT; DWT | ||
Statistics Article View: 179 PDF Download: 98 |