There are many ways to model advanced broadband networks, and a growing number of ways to simulate their behavior, based on available information about performance characteristics. This paper, co-authored by Comcast’s Ramya Narayanaswamy, Karthik Subramanya, Richard Prodan and Larry Wolcott, will explore the intersection of theoretical modeling, practical proactive network maintenance (PNM) and modern data science – a potent combination that is well suited for the sophisticated modeling and simulation needs of cable 10G networks.
Using traditional RF reflection measurements like scattering parameters / S-parameter matrices, fed by real-world field PNM (Proactive Network Management) data into a graph topology, the authors will show how advanced 10G networks, and particular bidirectional signal flows, can be simulated for both existing and proposed networks.
Having a graph topology of the network, along with cable and component specifications, provides a means to apply transmission line theory as never before. Most cable operators have this RF and reflection data readily available for their systems. This simulation of cable systems, as-built, can predict end-to-end performance which can be compared against actual measured performance using PNM tools. This technique can provide a method for the evaluation of full-duplex (FDX), extended spectrum (ES) and traditional RF transmission performance of existing plant, as well as experimental designs. Beyond that, it can solve for fault detection, and many other previously unsolvable RF mysteries within our cable universe.