<![CDATA[ A Technical Paper Prepared By
Diego R. Ambühl
Broadband Product Manager
Campillo 2541 – Buenos Aires – Argentina
+54 (911) 6040-4106
Cable operators are demanding devices that support an increasing number of simultaneous channels in order to provide consumers with new high-bandwidth service. Broadcom has introduced a new technology, Full-Band Capture Digital Tuning, that overcomes these limitations and delivers new advantages and features. From the NOC perspective, one of the most important features that Full-Band Capture delivers, is the ability to retrieve real time RF spectrum analysis that enables operators to remotely monitor and optimize cable plants.
The objective of this document is to present how the Full Band Spectrum capability, available in the new chipsets of DOCSIS cable modems, can provide information to diagnose and troubleshoot the TV Service in HFC Networks, and to what extent this ability can open a new outlook for operators, marked by substantial savings, significant improvements in the level of service and productivity, and the possibility of retaining their current customers and optimizing their satisfaction.
Cable Networks transmit downstream signals (from the headend to the customer premises) using frequencies between 42 and 1000 Mhz, depending on the equipment deployed in the field.
The Cable NOC usually diagnoses cable modems either on demand or on a regular basis, thanks to the ability to remotely access system variables that provide visibility to HSD working parameters. This way, parameters such as SNR, Power Level, Error rate, etc. are used to diagnose and troubleshoot HSD service.
These parameters provide visibility to those frequencies used for HSD transmission (usually up to 32 channels with DOCSIS 3.0 cable modems). Each downstream channel has 6 Mhz of bandwidth, which means that in the best case, the visibility range of the Downstream Spectrum achieved by remote monitoring tools is 192Mhz.
We mentioned before that the Downstream Spectrum is 1Ghz (1000Mhz) wide. This means that the NOC is leaving around 800Mhz out of the range of the HSD remote diagnostics tools.
DOCSIS 3.1 enlarges the spectrum up to 1.8Ghz (1800Mhz). At the moment this is a specification parameter, but equipment manufacturers are working to expand the spectrum from 1Ghz to at least 1.2Ghz. This spectrum expansion will make the ability to remotely diagnose the full spectrum more critical.
On those frequencies where HSD data is not transmitted, TV services such as analog TV, digital TV, VOD (Video on demand), IPTV, OTT, etc. are transmitted to the network.
If a NOC user wants to diagnose or troubleshoot a TV service issue, they will do so only after a customer has already complained, and they will need to send a technician to the customer’s premises, disconnect the service and connect a piece of specialized equipment (portable spectrum analyzer) to be able to diagnose and troubleshoot the problem. Even worse, if the problem is being caused by an impairment near the fiber node, they will need to go up to the passive elements connecting the spectrum analyzer until they get to the point of the node where the problem isn’t present, and then move down until they find the root cause of the impairment.
A New Paradigm
With the introduction of the Full Band Capture (FBC) functionality in DOCSIS, remote measurement of the downstream spectrum is now available.
Broadcom’s Full-Band Capture Remote Diagnostics technology can be implemented on the company’s entire cable set-top box, gateway and modem platforms using Full-Band Capture digital tuning technology chipsets.
By remotely capturing the downstream spectrum measurements from a cable modem, the NOC can quickly diagnose spectrum problems during a customer call (either on-demand or on a regular basis) and what’s more, by analyzing neighbor devices and by correlating the information they can pinpoint the problem from the NOC. That is to say that operators are able to anticipate and prevent problems through proactive analysis at regular intervals, can generate alarms when issues affecting the spectrum arise, and map any equipment or affected areas to prioritize the attention. They can also diagnose the situation and the use of certain frequencies before migrations, and correct problems before they impact service. This may assure customers’ full availability and avoid the huge expenditure of attending calls and complaints, as well as the subsequent truck rolls. Finally, it can also prevent the loss of customers dissatisfied with service interruptions.
Fig. 1: Multiple modems comparison
Use Cases: On-Demand vs. Scheduled Data Retrieval
Use cases are the key to understanding a variety of scenarios where the same information can be obtained, processed and presented in different manners.
For example, when a user is diagnosing a specific customer or working to solve an existing issue on the network, they might need to get the spectrum analysis measurement from a cable modem on-demand. This information is used to make an initial diagnostic, or even to update a measurement after an impairment source was found on the network to verify that the problem has been solved.
Fig 2: Single customer troubleshooting
A different use case is related to the ability to retrieve spectrum data at regular intervals, to apply analytics, correlation and generate alarms based on the analysis of the scanned information. The objective now is to address network issues before they even affect customers, to be able to pinpoint the closest point of failure, and even catalog the source of the impairment automatically.
Fig 3: Per CM severity classification
An additional use case takes place when the NOC wants to generate reports based on the Downstream Spectrum information to diagnose the usage of a specific frequency before or after a channel migration. Using the processed information about the Downstream Spectrum for each device, a report about a specific frequency can be generated to find existing traps that shouldn’t be present in the spectrum.
A fourth use case consists of counting with a high level dashboard that represents the distribution of the affected devices and the ability to navigate into the topology to understand which Market, Hub, CMTS, Node or Sub-node has the most affected devices and decide, based on this information, how to prioritize which sections of the Network should be attended to first.
Figure 4: Topological dashboard
Integration, Correlation and Analytics
In a scenario where certain integrations are made to include logical network topology information (passive elements hierachy) and/or an integration to a TV Channel map, valuable information can be processed to expand the downstream spectrum remote capture capabilities.
TV Channel map
When the TV Channel map is integrated to the Downstream Spectrum data, the NOC can complete tasks such as:
- Validate that the TV Channel is updated (by finding traps not present in the channel map)
- Select a group of ViP TV Channels to make a specific monitoring of these channels and detect issues affecting them
- Search for specific channels when a customer calls because they are having problems with a specific signal or group of signals
- Report frequency usage before a channel migration
- Confirm a channel migration is complete in all markets by analyzing that the previously used frequency is free
HSD service diagnostics
Monitoring tools have the ability to detect what frequencies a DOCSIS device is using in downstream. By mapping those frequencies to the full band spectrum capture, the NOC user can:
- Determine expected power level at each HSD frequency and generate alarms if power level is below the expected values
- Understand the health of a frequency before adding new HSD channels to the channel map
- Find undetected frequencies not registered in the channel map, avoiding problems caused by turning on a new channel overlapping with an existing HSD frequency
Full band diagnostics
Full band impairments are well-known, but automation of the identification is not a simple task because these impairments usually affect more than one channel, and the depth or variation of the power level can be very broad.
Full band impairments can affect a specific channel, which can be detected by comparing channels to a defined spectrum profile, one by one. Spectrum profiles are expected signatures of the full band spectrum that are used as reference to detect issues affecting the spectrum and generate alarms based on those analyses.
When full band impairments affect a wider portion of the spectrum, more complex analytics must be applied to classify the impairment, but this information can provide valuable information that is usually only available through visual analysis.
Well-known full band spectrum impairments provide valuable information for automating the detection of the root cause that is affecting a node.
Some examples of full band spectrum impairments are: FM and LTE ingress, Adjacency, suckout (notch), standing waves (amplitude ripples) due to impedance mismatches, presence of filters, roll off, tilt, resonant peaking, etc.
Fig 5: Several full band spectrum impairments example
Conclusion: The Whole Picture
If the NOC can count on a tool that can retrieve downstream spectrum information from DOCSIS devices either on-demand or at regular intervals, process that information in order to compare it with expected values or well-known impairments, determine severity and generate alarms based on severity thresholds in conjuction with network logic information, customer geolocalization and TV channel map, a blinded side of the network will become visible for the NOC to improve customer satisfaction by correlating information, detecting, classifying and pinpointing downstream spectrum impairments affecting HSD or TV service. This visibility allows the operators to act proactively and avoid the cost of handling complaints and calls to contact centers. It also allows them to plan and prioritize maintenance and reparation tasks, as well as reduce the cost of countless truck rolls that do not have prior analysis and diagnosis.
Bibliography DOCSIS® 3.1 Operations Support System Interface Specification, Cablelabs
 DOCSIS® 3.1 Physical Layer Interface Specification, Cablelabs
 DOCSIS® 3.1 MULPI Interface Specification, Cablelabs
 PNM Best Practices: HFC Networks (DOCSIS 3.0) Specification, Cablelabs
Abbreviations & Acronyms
NOC Network Operations Center
RF Radio Frequency
CM Cable Modem
CMTS Cable Modem Termination System
DOCSIS Data Over Cable Service Interface Specification
FBC Full Band Capture
HFC Hybrid Fiber Coax
HSD High Speed Data
IPTV Internet Protocol TV
LTE Long Term Evolution
OTT Over the top
SNMP Simple Network Management Protocol
SNR Signal to Noise Ratio]]>