DOCSIS – Change is Coming and it is Good
Everyone is talking about it. Everyone in cable will be impacted by it. You may not realize it but DOCSIS 3.1 will change the cable industry — whether cable operator, vendor or consumer. Even if you don’t plan to upgrade to DOCSIS 3.1 you will be impacted. There are two fundamental technologies that DOCSIS 3.1 has me excited about. First, DOCSIS 3.1 enables gigabit speeds, which allows us to extend the life of coax cable without the expense of pushing fiber closer to the home. Second, DOCSIS 3.1 enables test capabilities like we have never imagined possible before in cable modems (CM) and CMTSs. Conceivably we can automate the detection and location of RF impairments in the coaxial network with unprecedented accuracy — even better than the amazing capabilities of today’s proactive network maintenance (PNM) tools and far superior to most of todays conventional test equipment. I will briefly touch on the speed aspect of DOCSIS 3.1, more importantly what I consider the trickle down effect, but my main topic of this article is the test capabilities built in to DOCSIS 3.1.
Speed Trickles Down
Other articles will cover DOCSIS 3.1 speed, so it is sufficient to not get into details. What may be more relevant is that as large cable operators upgrade their DOCSIS 3.0 CMTSs to DOCSIS 3.1 we can anticipate surplus equipment. This is great news for Tier 2 and Tier 3 cable operators who are struggling to meet subscriber’s needs on older DOCSIS 2.0 and small DOCSIS 3.0 systems. Over the next couple of years we should see a rapid decline in DOCSIS 3.0 CMTS prices. This means smaller operators can afford upgrades in regions that have generally been underserved. Subscribers in those areas will see some much needed relief. All ships rise with DOCSIS 3.1.
Testing
Testing in a DOCSIS 3.1 environment is the real game changing opportunity for cable operators. As an industry we have learned a great deal about how to utilize the power of CM and CMTSs to do troubleshooting with DOCSIS 2.0 and 3.0 devices. This includes upstream and downstream spectrum analysis, detecting and finding impairments in coax cable, identifying intermittent connections that are likely leaking ingress into the upstream plant and more. At the same time we realized there was a lot we could not do because CM and CMTSs were not designed to be test equipment. This changes in DOCSIS 3.1. As the DOCSIS 3.1 specification was being drafted we included as many testing features as we could think of that we could not currently do with pre- DOCSIS 3.1 devices. In my previous article I mentioned laser clipping detection and spectrum analysis. This was just the tip of the iceberg. The way I believe it has been best described is that DOCSIS 3.1 CMTSs and CM is close to having an $80k vector signal analyzer in every subscriber’s house and headend. This is a pretty bold statement. However, the reason a cable operator would use such an expensive piece of test equipment is to capture and identify transient RF events, like burst noise. DOCSIS 3.1 equipment can do this. Let’s not stop at just a vector signal analyzer. In the first sentence of section 9.2 of the DOCSIS 3.1 PHY specification it reads “…the CMTS and CM contain test points which include essential functions of a spectrum analyzer, vector signal analyzer (VSA), and network analyzer…”. I really love that statement. It sets the ambitious tone the authors had for the entirety of section 9, called “Proactive Network Maintenance” (PNM). You see this was the first DOCSIS specification to ever have PNM written into the specification. The authors really understand the challenges of cable operators and cable plants, and therefore created a goal of integrating very powerful test tools into the CMTS and CMs. Having a test & measurement background, this makes DOCSIS 3.1 equally as powerful as its speed capacity.
Spectrum Analyzer
Let’s expand upon that sentence of section 9.2 in more detail to understand the power that we should expect from DOCSIS 3.1 CMTSs and modems. Today we already have upstream spectrum analysis performed by the CMTS and downstream spectrum analysis by DOCSIS 3.0 CM. In DOCSIS 3.1 we add noise power ratio notch (NPR) testing at the CM and triggered spectrum analysis at the CMTS. You’ve likely heard of NPR testing for characterizing return path lasers. Having this ability in the downstream is really cool in a DOCSIS 3.1 world because we will be able to see impairments in an OFDM DOCSIS 3.1 channel such as LTE interference, intermodulation products or any type of ingress that could be causing our 4096- QAM channels to be dropping to a lower modulation, all without having to turn off the entire DOCSIS channel. With triggered upstream spectrum analysis at the CMTS we will have much more flexibility to identify and capture upstream impairments. When we couple this with new CM that also capture upstream RF noise from the home we can correlate the source of ingress noise (in the home) and know exactly which homes to send techs in order to eliminate ingress noise. Gone are the days of pulling pads in amps at 3AM.
Vector Signal Analyzer (VSA)
The vector signal analyzer (VSA) is the second piece of test equipment mentioned in the sentence. If you have ever used a VSA you know that it is incredibly powerful for RF measurements and also incredibly difficult to setup and use. The latter makes it a challenge in cable networks in addition to its $80k price tag. What’s really amazing is that the bulk of the VSA features end up in the CM. These include: • Upstream pre-eq coefficients (also have pre-eq in the CMTS) • Downstream channel estimates • Downstream constellation display • Modulation Error Ratio (MER) per sub-carrier Pre-equalization coefficients are nothing new. Most of you are likely familiar with these by now from using PNM tools and have seen 8-tap and 24- tap equalizers. A DOCSIS 2.0 and 3.0 modem has these taps spaced over the 3.2 MHz or 6.4 MHz upstream. For a 6.4 MHz upstream that is one post-main tap per 400 kHz. With DOCSIS 3.1 things are a bit different. When in OFDMA mode each cable modem upstream subcarrier will only have a single complex pre-equalizer value vs. the 24 complex values of a DOCSIS 2.0 or DOCSIS 3.0 modem. However this is actually better because each sub-carrier will have a spacing of 25 or 50 kHz. Instead of one complex tap per 400 kHz we will have one complex tap 50 kHz minimum. Therefore our tap resolution will go up by at least eight times. This means impairment location accuracy will improve dramatically. Similarly, downstream channel estimates allow for the CM to provide estimates of the downstream performance back to the CMTS. Downstream constellation display and modulation error ratio (MER) are fairly self explanatory, but consider that we are getting these from a CM and not an expensive test meter. In a recent project I worked on we determined that a DOCSIS 3.1 CM with a battery pack and an iPhone could do everything and more than competing hand-held test equipment. This included tasks such as sweeping the network (assuming you only swept the 192 MHz DOCSIS band). With respect to sweep, however, there are some techniques in the upstream that are being finalized which allow complete RF characterization, so it is likely upstream sweep is complete. I am optimistic that we will be able to accomplish the same in the downstream once DOCSIS 3.1 equipment is more readily available.
Network Analyzer
The network analyzer functionality in the both the CMTS and CM is something that I believe will be heavily used once DOCSIS 3.1 systems are deployed. The concept is simple. Downstream and upstream channel estimates are sweet features and can be best envisioned as a sweep system performed by the CMTS and the CM. This is accomplished because the CMTS knows every symbol it transmits, which is stored in memory. The CM then receives data from the CMTS and reports this set of channels (i.e. channel estimate) back to the CMTS upon request. A comparison of the transmitted data by the CMTS and received data by the CM is performed which results in a channel response. An example channel response is shown in figure 1. This is similar to sweeping the cable plant because it shows the frequency response of the cable plant, however it is much more powerful because it is a complex frequency response which shows both magnitude and phase impairments along with any impaired symbols. This process also works in the upstream, which is very powerful when using OFDMA in the upstream. In this case sweeping the return would not be possible. In a DOCSIS 3.1 environment one could sweep the return from any subscriber’s home 24×7 without ever leaving the headend. This is powerful technology and I’m really excited about it as you should be too.
Summary
The impact for cable operators is extraordinary. (PNM) tools than ever before. PNM is very powerful, effective, and desired by cable operators today without question. I am even more excited to see PNM in a DOCSIS 3.1 network! Change is coming and it is good.
As seen in BROADBAND LIBRARY
By Brady Volpe
Previous events can be seen under the blog.
- If you are watching this on youtube please hit the subscribe button!
- Let us know what you think and remember to share!
- You can find slides at the bottom of the page and some on slideshare.
- Find out about events or articles by following us on Twitter, LinkedIn or Facebook too.
- We have stuff. Interested in buying some clothes with cable stuff. Click here
Also available on iTunes, Google Podcasts, Spotify, vurbl see podcasts “get your tech on”.
Hi Brady,
Up in the white north we have some ftth & ftts implementations and are using rfog to the suites. We currently have some tier two tools we utilize to see noise in our cable plant, and I have seen noise in all-fiber buildings (presumably some form of light interference getting into the fiber? or the noise is so high it passes past the onu?) Do you forsee any possible way to utilize docsis 3.1 and pnm to help track down the source of these disturbances too?
Hi Matthew,
The problem with RFoG (DPON) is the ONU will turn on for any energy whether it’s DOCSIS or not. Issues could be uncorr FEC not from OBI, but optical link misconfig/wrong levels/setup. Other issue is lasers turn on from any energy (CB, noise, docsis, STB, impulse noise, etc). Not much we can do about it. I know our own ONU has a squelch feature to turn off the laser if on continuously more than about 20 msec and logarithmically increases that turn off when it happens more often. It is difficult to pinpoint a bad house in an RFoG environment since the noise (if low enough) will not activate the ONU/laser and you would not “see” it in the HE. PNM’s base feature was to look at pre-eq info which is not the same as noise viewing. Maybe using new CMs with US ingress viewing would be fastest and easiest way to identify the culprits.
On my opinion this wold be a very complicated decision for the operator. I do work as HFC manager close to 3play service provider and for the moment we are benefiting a lot jumping from D2 to D3. Only cmts changed and modem d2 and d3 coexisting together, no active or passive to be changed as far as we are almost in the same bandwidth frequencies. Here the 3.1 is a huge jump and a lot need to be change passives and actives also. Don’t know if it is going to worth for a hundred of MHz of bandwidth more?! Huge investment are needed. Personally in this moment I would go for a deeper fiber segmentation that means at least 2 time same bandwidth for the same customers (frequencies reuse). Same equipment, better SNR that means highest modulation scheme can be easily used. It is just a couple of fiber and you easily double your spectrum, no need for new transmitters, new node & amplifiers, new passives and the last not least the proper tuning of the network, which I do not believe that will be a simple equipment change. To fit the same cable distances higher gain amplifier are necessary needed and on the other hand the new equipment might have at list same noise figure, CSO and CTB for a bit more channel load if you are going to use all the bandwidth that means higher equipment power consumption and in the end do not forget that cables that are getting older year by year. As I said it will be a very complicated decision.
Hi Ervinbegu,
You make some great points and there is a lot of debate among cable operators as well as vendors on DOCSIS 3.1. Many say why not just go fiber to the home, while others recognize that extending the live of the invested coaxial infrastructure has significant merits. DOCSIS 3.1 does not require any changes to the outside plant – this is optional if you want to really experience the full potential of DOCSIS 3.1.
You could drop a DOCSIS 3.1 CMTS and cable modem into you existing plant and see significantly increased downstream and upstream spectral efficiencies without any plant changes due to OFDM and the LDPC error correction. This is one big reason major cable operators are looking to invest in D3.1 to extend the life span of their existing HFC plants with minimal investment in the near term. Further the built-in testing capabilities of D3.1 are quite powerful – this is the PNM applications.
Longer term cable operators will likely need to change the split (85 MHz or 200 MHz returns), which means a lot more investment. This does bring up the question of “when does more fiber make sense?” I think that question will be dependent on a case-by-case analysis.
What do you as readers think about DOCSIS 3.1 vs. PON architectures?