Keeping Crabgrass Out of Your Network

It’s hard enough to keep problems in your own network from spreading like crabgrass, but operators also need to keep the neighbors’ crabgrass from creeping in as well.

Time Warner Cable‘s Thomas Staniec, vice president of transport network engineering at TWC’s advanced technology group, used the above metaphor during a Webcast sponsored by JDSU and hosted by Communications Technology.

To back up a bit, Staniec explained the problem with crabgrass by using another metaphor. In the past, the "barn" was the main headend in a system or division while the "corral" was the franchise boundary. With the advent of VoIP, and to a lesser extent high-speed data, the proverbial horse has left both the barn and the corral. VoIP crosses more boundaries, including more with the PSTN, and it requires more than best-effort practices.

"If we look at a simple VoIP picture, there are six boundaries to cross, and of those six, only two of them may be owned by you (a cable operator)," Staniec said. "If a cable operator needs to go through a SIP cloud to another cable operator, and there’s a customer issue, where is the problem, and how do you fix it?"

Part of the solution is checking your own network for problem areas or "network crabgrass," Staniec said. Not only do you have to check your network for crabgrass; you also need to check your boundaries to make sure other networks’ crabgrass isn’t creeping in.

Metaphors aside, Staniec said that while key performance indicators (KPIs) are changing the face of monitoring, the industry needs standard measurements that are presented in open OSSs so that the KPIs mean the same thing to everyone who sees them.

"If the issue can be pinned down quickly, then we get the right engineer there at the right time," Staniec said. "The other side of a network also affects how our customers hear their service. We need to be able to tell our customers when there’s a problem on the network boundary and to get that information to whoever is outside of our network."

Staniec said Time Warner Cable has spent considerable time and energy on creating a monitoring umbrella across its systems and divisions. To build a network shield, TWC has standardized on an open OSS and made sure that everyone can write APIs, as well as standardize critical metrics.

"You take that information and data and correlate it to make it consistent to every location," he said. "Then you make sure that everyone interprets those metrics the same way. When all of the information and data is collected the same way and put on an open network, then when you see a difference you can find out right away why it’s different there." Finding the right tools Kevin Oliver, JDSU’s vice president of marketing, cable networks, said there are several best practice techniques that operators should have in place, including:

• Use cumulative data sources to view plant heath from end-to-end

• Summarize customer VoIP call quality analysis to rapidly sectionalize and proactively manage the calls

• Use VoIP and RF reports to "get proactive with HFC node maintenance"

Oliver said one way of doing the last item is by ranking top offending nodes by VoIP and by daily ranking of RF performances.

To see the free on-demand version of the Webcast, go to http://www.iian.ibeam.com/events/acce004/21173/. Note that registration is required. – Mike Robuck