Best Practices General

Fair Share for Bandwidth Management

By Incognito on March, 17 2015

Stay up to date

By now, we know that Internet Protocol Detail Record (IPDR) is a highly efficient protocol that delivers per-subscriber service usage information directly from the CMTS. This protocol is integrated with DOCSIS, so statistics are periodically collected and reported to an external application without overstraining the network. Unlike more invasive data collection methods — such as deep packet inspection (DPI) — IPDR protects user privacy while still providing a wealth of information that aids fair share management and network investment. IPDR data contains information about every flow inside a CMTS and lets you see where and when bandwidth is being used.

But even with IPDR statistics in hand, analyzing Big Data can be a challenge because you need to review many data sets to gain meaningful insight. Some providers also encounter inaccuracies with the data — for example, unreasonably high values or CMTS errors.

You can avoid these problems and ensure accuracy by finding an IPDR collector and normalizer that filters IPDR results to collect key metrics. Go one step further with a bandwidth management solution that not only lets you see at a glance where congestion occurs but also automates business policies of your choice to deal with congestion.

For example, enforce a temporary slowdown for the handful of abusive users who are hogging bandwidth in a region, or create node splits to alleviate congestion in real time. Alternatively, you might choose to build subscriber profiles to create better service offerings or offer heavy users the chance to upgrade.

The opportunities for fair share policies will depend on the region you operate in, but when it comes to reducing congestion, there are plenty of options. The first step is to figure out where and when bandwidth is being consumed. So get creative, look at fair access policies or simply start thinking about how and where you should invest in network expansions and node splits — your subscribers will thank you.

Submit a Comment

Get latest articles directly in your inbox, stay up to date