Denoising network tomography estimations

In this paper, we apply the technique of sparse shrinkage coding (SCS) to denoise the network tomography model with errors. SCS is used in the field of image recognition for denoising of the image data and we are the first one to apply this technique for estimating error free link delays from erroneous link delay data. To make SCS properly adoptable in network tomography, we have made some changes in the SCS technique such as the use of Non Negative Matrix Factorization (NNMF) instead of ICA for the purpose of estimating sparsifying transformation. Our technique does not need the knowledge of the routing matrix which is assumed known in conventional tomography. The estimated error free link delays are compared with the original error free link delays based on the data obtained from a laboratory test bed. The simulation results reveal that denoising of the tomography data has been carried out successfully by applying SCS.