By Ron Hranac, Cisco Systems; James Medlock, Akleza,Inc.; Bruce Currivan, JJP Development; Roger Fish & Tom Kolze, Broadcom; Jason Rupe & Tom Williams, CableLabs; Larry Wolcott, Comcast
Receive modulation error ratio (RxMER) has long been a powerful metric for cable network maintenance and troubleshooting. A limitation to single carrier quadrature amplitude modulation (SC-QAM) RxMER is that the reported value doesn’t give an indication of why that value is what it is, or what kind of impairment might exist.
The DOCSIS® 3.1 specifications define several operational measurements that can be reported by the cable modem and cable modem termination system (CMTS) or converged cable access platform (CCAP).
One important modem performance parameter is orthogonal frequency division multiplexing (OFDM) RxMER per subcarrier, which can be plotted to show a graph of all subcarriers’ RxMER performance.
Based on real-world observations of data from production cable networks and subsequent lab testing to recreate and validate the observations, a number of specific impairments can be identified that point to faults in the underlying network. Not only does the identification of these problems assist with maintenance and troubleshooting of the network, but various impairments identifiable in the RxMER per subcarrier plots can impact subscriber service and result in lower throughput and performance than expected. Plus, due to the sensitivity of the RxMER per subcarrier measurement, it can find impairments in the network before they adversely impact customer service, and before repairs become costly.
This paper includes discussions about a number of impairments that have been observed, describes the findings when recreated in a laboratory environment, and explains how the observed results point to potential cable network faults. Examples include:
- Amplitude ripple in the channel in the frequency domain can under certain conditions cause amplitude ripple in RxMER per subcarrier graphs.
- Interference caused by long term evolution (LTE) and other ingress can be correlated to specific frequencies by observing the impact on the RxMER per subcarrier graphs.
- SC-QAM signals adjacent to OFDM signals can cause a rolloff at the edges of the RxMER per subcarrier graph.
Production network examples are presented that show how analysis of RxMER data collected from cable modems can be used to identify and locate specific cable network impairments, resulting in an improved subscriber service performance and experience.