Statistical Analysis of Path Length in Optical Networks

Abstract

Optical networks are the main traffic carriers of all the modern communications. All the core and regional networks are very much optical these days. Even in the metro and access area networks, we now have a lot of presence of optical fibers. With the growth of traffic, the complexities of management and control aspects of the optical networks increase. In optical networks, shortest path and the minimum hop paths are not the only choices when the network operates at its optimum capacity level. Rather, several other paths are chosen for routing and traffic management, which are neither the shortest nor minimum hop path lengths. So, there is a requirement to understand the statistics of these paths. In this paper, we analyze the path lengths of 35 real optical transport networks (OTNs). For this study, we used 65 different statistical distributions. We found that both Wakeby and Johnson SB distributions are very much suitable for the modeling of path lengths in OTNs. The validity of our statistical measurements was checked using Kolmogorov–Smirnov statistic (KSS). For both Wakeby and Johnson SB distributions, all the KSS values obtained are valid at 95% confidence interval for all 35 networks.