When Life Hands You Bananas, Make a Banana Split

By Jeff Finkelstein

2020 has been challenging us all, both personally and professionally. I hope everyone, their families, co-workers, and friends are safe. While we are often brought down by the times we live in, we can be equally uplifted by the kindness of others. Please keep yourself and loved ones safe by practicing social distancing. Keep your social distance from your refrigerator as well.

Memories of a childhood well spent

Those of us who have been in cable a long time remember all too well the “Radio Shack” effect, when consumers became their own tech support purchasing splitters, cables, and connectors from the local bastion of nerds known formally as RadioShack. In fact, many of us got our start as engineers with RadioShack kits. We saved our money to buy them and spent many happy hours assembling them to build working electronics. In a real-life “okay boomer” moment, how many of us remember those times fondly? Today’s generation of engineers will never know the joy of powering up our AM or FM radio for the first time or the ignominy of defeat when they shorted out forcing us to go back and buy whatever components we fried. To paraphrase Lieutenant Colonel Kilgore from “Apocalypse Now,” “I love the smell of burning capacitors in the morning. It smells like victory.”

For anyone who has worked as an installer, plant, or headend tech, we know all-too-well the impact of these components after a self-install by consumers. Many hours and days were (and are) spent troubleshooting ingress and egress, figuring out where the problems were located only to find out customers did not make optimal choices when modifying their home cabling.

There are so many components that are needed to keep the outside cable plant going it can appear at times daunting when we talk about plant upgrades. It is not just the actives and passives we consider when looking at future plant changes, it is the hardline and drop cables plus the inside the home cabling, house amplifiers, with a good sprinkling of the unknown as we do not know where customers at times decided to install or move equipment.

May you live in interesting times

While our customers, companies, employees, us, and our networks are stressed from the effect of isolation, there are learnings we can take away even in these times. I’ll give you my list of top three learnings to date and what it may mean for future planning. As we slowly dig our way out of the current conundrum, I’ll add to the list in future columns.

Learning 1:

You can never have too much fiber

As we move to distributed access architectures (DAA), whether remote PHY or remote MACPHY, we need fiber to reach them. We also need fiber for many business customers. I have never heard anyone say “boy, I wish we didn’t install all that extra fiber.” On the contrary, I have heard the opposite and we berate ourselves for not taking advantage of having construction crews out there and using that opportunity to deploy more fiber than we currently need in the two to three-year window.

Fiber is a long-term strategy. I love cable. It has been my work life for the past 32 years. However, I know the value of fiber for enabling future technologies and positioning ourselves for upgrades in our network as strategies change. As I like to say, cable has a long useful life. However, we still need to prepare every chance we get for the next thing, which means a lot of fiber deployed regardless of which path we go down.

The highway analogy works well here. Who has not seen the amount of mind-boggling traffic created as highway crews disrupt the already congested flow of cars to work safely adding new lanes? Imagine if they had thought years ahead and had those lanes ready before they were needed. Some folks would complain about wasteful spending, but they are the same ones who likely would complain even louder later for the roads not being prepared. As John Lydgate, 15th century Benedictine monk from the monastery of Bury St. Edmunds, famously said (or at least it is the earliest attribution I can find):

“You can please some of the people all of the time, you can please all of the people some of the time, but you can’t please all of the people all of the time.”

The lesson learned is sometimes you have to break a few eggs to get things in place for future growth. Yes, we may possibly tick off a few folks in the process, but it falls into the “non-regrettable spend” category for the long-term health of our networks.

Bottom line: Pull as much fiber as you reasonably can when the opportunities present themselves.

Learning 2:

DOCSIS 3.1 and mid/high-split is our friend

Every major version of DOCSIS to date has taken quite a bit of time (from five to 10 years) to go from ideation to deployment. Many of us who drive these technologies know that all too well. However, we can take advantage of preparedness for the next major release by taking a macro view of the needs of the new technologies.

Let’s use DOCSIS 3.0 as an example. We knew years in advance that silicon was coming that would allow up to 32 SC-QAM channels in a bonding group for the CMTS. We began getting CPE early on that only had enough downstream to use bonding groups of four, then eight, 16, 24, and finally 32 channels. We did not wait for the final product, but instead began “seeding the market” with products that had features we knew we would not, and could not, use until we had the CMTS to support bonding up to 32 channels. While this did prove to be a good choice, it did introduce complexities in bonding group configurations that we are still dealing with today. It is not unusual to have a 32-channel bonding group which is overlaid with two 16-channel bonding groups, which is then overlaid with four eight-channel bonding groups. It adds to the complexity of scheduling, but we are fortunate to have some very smart developers who have managed to hide the complexity from us.

If we would have waited until we had the final form CPE product before deploying, we would have been far behind many of our competitors. We made the strategic investment of seeding multiple versions of CPE when we managed to maintain a technology leadership over many of the competition and more importantly keeping our customers happy.

Bottom line: Getting ahead of the technology curve and deploying portions of the solution makes the path easier by taking it one step at a time.

Learning 3:

Don’t be afraid of them golden stairs

What I mean here is that making opportunistic and strategic investments simplifies deploying complex technology ecosystems. Look at DOCSIS 4.0. It is the most complex version of DOCSIS by far. We went through a period of fog where we were running in multiple directions simultaneously. Thanks in large part to operators working together we have a singular clarity of vision and have unified behind two technologies. First is what I originally called extended spectrum DOCSIS but is now called frequency division duplex (FDD) DOCSIS. The second is what we call full duplex (FDX) DOCSIS.

There is industry support for both paths of DOCSIS 4.0 with the key difference being plant upgrades needed for one or the other. Some of us prefer a 1.8 GHz downstream and up to 684 MHz upstream direction), and others prefer a 1.2 GHz downstream and 684 MHz upstream configuration with the 108 MHz to 684 MHz region being the FDX band. Both are great solutions and the decision point is around how companies plan on upgrading their plant.

DOCSIS 3.1 has a lot of life left to it, but we need to begin the process of upgrading plant for whichever DOCSIS 4.0 path you want to take. Back to my highway analogy, step 1 is the plant upgrades are creating the lanes for traffic that will not be seen for a while. Step 2 is we then need distributed access architecture devices (RPD and RMD) to direct the traffic into the proper lanes on the highway. Finally, in step 3 we need CPE to drive on the highway taking advantage of the lanes of traffic we created to optimally take advantage of the new roads we have built.

Bottom line: Making strategic investments in our network now will reap huge rewards later. Every dollar spent must be done so to position us for future technology paths.


As we say at Cox:

Act Now: Pull as much fiber as you can, and more if possible

Be Bold: Get ahead of technology by seeding the market with passives, actives, amps, CPE, and DAA devices

Stay True: Keep your eye on your North Star and drive technology towards that goal

Until next time, stay safe!


Jeff FinkelsteinJeff Finkelstein

Executive Director of Advanced Technology
Cox Communications

Jeff Finkelstein is the Executive Director of Advanced Technology at Cox Communications in Atlanta, Georgia. He has been a key contributor to the engineering organization at Cox since 2002, and led the team responsible for the deployment of DOCSIS® technologies, from DOCSIS 1.0 to DOCSIS 3.0. He was the initial innovator of advanced technologies including Proactive Network Maintenance, Active Queue Management and DOCSIS 3.1. His current responsibilities include defining the future cable network vision and teaching innovation at Cox.  Jeff has over 43 patents issued or pending. His hobbies include Irish Traditional Music and stand-up comedy.