Besides the capacity and robustness improvement, there’s something else about DOCSIS 3.1 and its use of orthogonal frequency division multiplexing (OFDM) that excites me: the network intelligence it enables. We already use DOCSIS devices as network health probes, but OFDM will essentially turn our equipment into multifunction test generators that can be used to measure most of the key RF performance parameters and diagnose the HFC network automatically in order to optimize performance and maximize capacity.
Let’s start with what we can do now. The DOCSIS specification instructs us to calculate the codeword error rate RC as a proxy for post-forward error correction (FEC) bit error ratio (BER) if the latter cannot be measured directly: RC equals the number of uncorrected codewords divided by the total number of codewords received over some interval of time. For those of you preparing to take the SCTE’s upcoming DOCSIS certification exam, I just helped you prepare for it.
But more recently, we’ve gone beyond BER-type measurements and can now do the equivalent of time domain reflectometry (TDR) using the adaptive equalization process built into DOCSIS. For engineers, you get essentially a discrete form of the impulse response of the cable network going to individual homes. This measurement can then be triangulated with geographical information to identify the magnitude, type and location of typical cable network impairments, such as cracks/breaks in the coax and the connectors, water flooding, and so on. For technicians, think TDR measurements every 20 seconds or so with cable modems that used to require field equipment and truck rolls.
CableLabs calls this the Proactive Network Maintenance tool, Comcast calls it Scout Flux, but the point is that MSOs are rolling this out throughout the world. If you want a good tutorial on the process, an appendix in the recently published “SCTE Measurement Recommended Practices for Cable Systems” explains it very well.
The latest addition, which CableLabs calls InGeNeOs, adds spectrum analyzer-type functions using the built-in sampling and FFT capabilities of modern cable modem chips. Now we can do a mini-sweep and TDR from the headend! Due to its narrow, multi-tone nature, OFDM in DOCSIS 3.1 will allow the same kind of measurements, only with greater precision, accuracy and flexibility in measurement configuration.
But much more is also possible. To characterize the non-linearity in the actives, SCTE 115 is the age-old two-tone intermodulation distortion (IMD) measurement for cable networks, where two test tones are injected into the network, and the IMD products at sums and differences of the test tone frequencies are examined. Last year, the SCTE began exploring the use of the multicarrier capabilities of DOCSIS 3.0 modems to effect the same measurement with up to six tones, but it would require customized diagnostic cable modems, and only one vendor had them.
But with OFDM in DOCSIS 3.1, we can generate up to thousands of tones to probe the network for IMD and can look for IMD products just about anywhere we want to, on both upstream and downstream. Further, we can conceivably perform the more rigorous noise power ratio (NPR)-type measurement (also described in the SCTE Measurement Recommended Practices), where an RF channel is loaded fully except for a notched-out portion of the spectrum where the IMD is measured.
And that’s not all. Back in the analog carrier days, we looked for common path distortion (CPD) primary tones at multiples of 6 MHz on the upstream, or if it was strong enough, we could see it in the triplet pattern of tones around those primary tones. With OFDM, we could use pilot tones in the downstream to emulate the old analog carriers, including locking the distributed pilot tones in phase to coherently add to enhance the effect. Notching out a configurable portion of the spectrum in OFDM allows us to measure the CPD. It’s a test engineer’s dream setup, and it will allow us to characterize the RF plant health long before it degrades service and fix those problems with surgical precision.
What about leakage measurement? There are several new RF leakage detection systems designed to increase the sensitivity for detecting RF leakage in all digital plants – either by much higher antenna gain or by time domain processing gain via correlation with the downstream QAM signal – and another more traditional method that uses markers.
The innovation is required because QAM signals have flat, lower-power spectral densities and also lower transmit power levels than analog carriers. But we could also use OFDM pilot tones as unmodulated carriers to look for leakage using traditional low-cost techniques.
Or combine this with the higher sensitivity innovations for a double enhancement.
We may need it when even small amounts of LTE signal ingress into our networks could degrade our network capacity.
So start thinking now about how you could innovate with OFDM in your systems to proactively maintain the health and maximum capacity of your HFC networks.
For a deeper explanation on this, please visit https://www.scte.org/OFDM to review numerous resources on the topic.
dhoward@scte.org
