A Top Down Approach to Estimate Network Loss Rate

Loss tomography has received considerable attention in recent years. A number of methods, either based on maximum likelihood (ML) or Bayesian reasoning, have been proposed to estimate the loss rates of a network. Almost all implementations use an iterative approximating method, (e.g. EM algorithm) to search for the maximum in a multi-dimensional space, which crates two concerns: scalability and accuracy since the time used in searching increases exponentially as the number of links and the search may stop at a local maximum. To overcome the problems, we propose a top down approach to replace the iterative approach, which significantly reduces the time spent on estimation, and ensures the solution identified is the maximum likelihood estimate (MLE). We present simulation results to show the efficiency and accuracy of the method.