Kickstarting Proactive Network Maintenance with the Proactive Operations Platform and Example Application (2019)

By Jason Rupe, Ph.D. & Jingjie Zhu, CableLabs

As part of its long standing proactive network maintenance (PNM) project, CableLabs® has been assessing the needs of operators and vendors in the area of PNM with the goal of reducing adoption friction for members and vendors. We identified a few main issues, an important one of which is addressed by the work presented in this paper: the Proactive Operations (ProOps) platform.

ProOps is a platform (environment, framework) for turning data into operations action. That includes proaction, when the data allow it. PNM data enables proaction, so we built an example application that comes with ProOps, which serves multiple purposes: as an example to show how to use ProOps, as a starting point for trying basic network data-driven PNM and reactive operations, and as a launch point for implementing and sharing PNM best practices.

To turn data into action, most operators rely on engineering and technician expertise. It is common to simply gather and plot the data, then look at the output. CableLabs built the cable modem validation application (CMVA) for that latter purpose (as well as for cable modem (CM) certification test automation and sharing). But without a human expert sifting through the data, not much can be done with it. CMVA is great for developing PNM ideas, but it still requires experts to do the next step, and developers to build solutions to try. That requires investment risk that we surmise is a roadblock to implementation of PNM. But for many operators, and some vendors, there just aren’t enough available experts to do the work manually. The industry needs help getting over the hurdle of turning the data we exposed into action we can take with confidence.

ProOps was built to facilitate the automation of turning data into operations action. Generally, we identified the steps to accomplish that task as 1) data extraction (observation), 2) analysis across time and network elements (orient), 3) correlating problems and measuring severity (decide), and 4) defining work items that are worthy of attention (act). The steps can be labled as observe, orient, decide, and act (OODA) to roughly follow the OODA process or OODA loop, which is a cyclic process developed by US Air Force Colonel John Boyd [1,2]. Combat operations resembles network operations more than we care to admit perhaps, so the labels fit. Boyd systematized the combat operations process as a rapid cycle of the OODA loop. Likewise, network operations follows a similar process, and the concept helps explain how ProOps works.

In the future, we expect to release applications and modules to enhance existing applications, in the ProOps environment. Once ProOps is installed, new applications will work like updates to ProOps, making the operations impact even less. New applications and modules will interwork with existing applications in the same deployment of ProOps, or parallel deployments utilizing the same common collection framework (CCF) instance are possible too. Multiple deployment models are available today, and more to come as vendor and operator members request.

ProOps is currently available for use by CableLabs members and vendors under nondisclosure agreement (NDA) and intellectual property rights (IPR), with the additional common code collection (C3) community agreement.

In the rest of this paper, we explain in-depth the structure and function of ProOps, and the example application that comes with it today. This will lead us to explain some of the basic ways you can configure and build your own solutions in ProOps to support your operations improvement experiments and inventions. We also cover use cases for ProOps that we envision, which hopefully will interest you or spur your own imagination to invent use cases we haven’t thought of yet.

By clicking the "Download Paper" button, you are agreeing to our terms and conditions.

Similar Papers

Profile Management Informed Proactive Network Maintenance
By Jason Rupe & Jingjie Zhu, CableLabs
A General-Purpose Operations Cost Model to Support Proactive Network Maintenance and More
By Jason Rupe, Ph.D., CableLabs
Proactive Network Maintenance Evolution to the Optical Domain in Coherent Optics
By L. Alberto Campos, Ph.D. & Zhensheng (Steve) Jia, Ph.D., CableLabs & Larry Wolcott, Comcast
Proactive Customer Maintenance
By Andrew Joseph Milley, Cox Communications Inc
Convolutional Neural Networks for Proactive Network Management
By Jude Ferreira, Maher Harb, Karthik Subramanya, Bryan Santangelo & Dan Rice, Comcast
Using Historical Traffic Data to Schedule Service Interruptions for Minimum Customer Impact
By Jason Rupe & Colin Justis, CableLabs
Proactive Network and Technical Facilities Monitoring Using Standardized Scorecards
By Dr. Franklin Lartey, Cox Communications
A Proactive Network Management Scheme for Mid-Split Deployment
By Lei Zhou, Robert Thompson, Robert Howald, John Chrostowski & Daniel Rice, Comcast Cable
A PNM System Using Artificial Intelligence, HFC Network Impairment, Atmospheric and Weather Data to Predict HFC Network Degradation and Avert Customer Impact
By Larry Wolcott, Michael O'Dell, Peter Kuykendall, Vishnu Gopal, Jason Woodrich & Nick Pinckernell, Comcast
OpenCable Application Platform Architecture
By Allen R. Schmitt-Gordon, Ph.D., Cable Television Laboratories, Inc.
More Results >>