Is your DOCSIS Network Ready for VoIP?

The $50+ billion yearly cable industry is positioned to directly compete with the $200 billion yearly telephone industry. To do this, cable operators will use voice over Internet protocol (VoIP) and additional infrastructure, such as a number of call provisioning elements, multimedia terminal adapters (MTAs) and their existing Data Over Cable Service Interface Specification (DOCSIS) networks. The result will be a bigger piece of the $200 billion pie for the cable operator. For this to be successful, DOCSIS networks require improved reliability. In order to obtain the revenue generating opportunities of VoIP, it is essential to understand and troubleshoot a DOCSIS network. If all of the elements of the DOCSIS network are working properly, from RF to IP, then VoIP should be elementary. However, if there are any RF impairments, DOCSIS protocol impairments or Internet protocol (IP) data routing anomalies, then VoIP deployments can be disastrous. Understanding and identifying these impairments requires new technologies and the understanding of new concepts. DOCSIS working model Developed by CableLabs and a consortium of cable operators, DOCSIS provides a standard for bridging Ethernet data over the hybrid fiber/coax (HFC) network. DOCSIS defines the method by which DOCSIS-based devices operate on an RF plant. Fundamentally, there are three layers of communication that must be understood in order to analyze a DOCSIS network. This model is best illustrated in Figure 1. The base of the pyramid represents the physical layer of DOCSIS. This is where data is modulated and upconverted to an RF carrier for transport across the HFC plant. The middle of the pyramid is the DOCSIS media access control (MAC) layer. This layer represents the intercommunications between DOCSIS-based devices that enable cable modems and MTAs to communicate with the cable modem termination system (CMTS). Finally, the top of the pyramid represents the layer at which IP data is transported. This layer is used both by subscribers for data access and by cable modems and MTAs during registration (explained later). Troubleshooting DOCSIS In assessing and improving DOCSIS network reliability, it is critical that all three layers of the DOCSIS working model be analyzed. Impairments that occur in the RF plant may appear to be DOCSIS- or IP-related problems; similarly, problems in the DOCSIS MAC or IP layers may have symptoms similar to problems caused by RF impairments. This creates "finger pointing," which results in loss of time and money by attempting to resolve problems that are not understood. In order to solve this problem, we will first describe the origin of many common DOCSIS network problems. We then troubleshoot the network by working our way up the DOCSIS pyramid. RF plant impairments The majority of DOCSIS network problems occur at the physical or RF plant level. This stems from the fact that the RF network, particularly in the upstream, is inherently rich with mechanisms that may prevent or impair DOCSIS communications. Some very common and well-known RF impairments are: Individually, these impairments may impact DOCSIS reliability and can be difficult to identify using conventional test equipment. For example, carrier-to-noise ratio (C/N) and signal-to-noise ratio (S/N) do not identify all RF impairments; however, modulation error ratio (MER) does reflect all RF impairments, although MER by itself doesn’t identify the specific problem. Collectively, these impairments create an environment in which it is nearly impossible to identify the source of network failures. Analyzing individual cable modem performance requires a new breed of test equipment called DOCSIS protocol analyzers. These analyzers demodulate individual cable modem transmissions and provide both quantitative measurement of the signal’s quality and information designating the source of the transmission. This is critical in DOCSIS networks because there are hundreds of individual cable modems transmitting data. A DOCSIS protocol analyzer can filter on a single cable modem and provide signal quality measurements as shown in Figure 2. Avoiding pointless truck rolls Equipped with this analyzer, a cable technician can look at a single cable modem experiencing performance issues and determine if performance issues stem from RF impairments, without leaving the headend or hub site. Figure 2 depicts a cable modem transmitting in 16-QAM (quadrature amplitude modulation), with MAC address 00:E0:6F:29:43:FE, and exhibiting a 27.7 dB MER. Since 16-QAM transmissions require 18 dB plus MER, the technician can be certain that this modem is operating properly in the RF plant. However, if the MER of this particular modem were less than 18 dB, then the technician would know that the modem is experiencing impairments from the RF plant. With this information, in a matter of moments a headend technician can determine if a customer’s modem, or MTA for voice communications, is experiencing problems from RF impairments from the physical plant. If plant related impairments are present, then a truck roll is likely necessary; however, if MER is good, the problem is unlikely to stem from the RF plant (one exception might be transient impairments such as impulse noise, which may not show up in an MER reading—Ed.), and the problem lies within the DOCSIS or IP network, which involves resources other than field technicians—a pointless truck roll stopped. DOCSIS protocol impairments When DOCSIS-based devices violate the DOCSIS protocol, network failures are sure to follow. Using conventional RF test equipment to identify DOCSIS protocol violations is not possible. The CMTS provides some diagnostic information, but this usually only verifies that a problem exists without showing the root cause of the problem. In order to identify and isolate DOCSIS protocol violations, you must analyze the protocol in its native format—as RF modulated carriers in the RF plant. Additionally, DOCSIS commands have a format unlike any other in the communications industry. Cable operators must again turn to DOCSIS protocol analyzers in order to troubleshoot DOCSIS protocol related problems. This is the same way the DOCSIS vendor community develops and troubleshoots DOCSIS-based CMTSs and cable modems. A DOCSIS protocol analyzer provides a user-friendly graphical user interface (GUI) to capture DOCSIS protocol and a viewer that presents the DOCSIS protocol in a format that is readable by the human eye. (See Figure 3.) This output is a powerful tool for both identifying the presence of DOCSIS violations and for settling disputes with vendors providing equipment that violates DOCSIS. IP layer impairments IP layer impairments impact both the subscriber and the DOCSIS network. It is obvious that for a cable modem and MTA to register with a CMTS, both the RF plant and the DOCSIS MAC must be functioning properly. But the IP layer is also critical for cable modem and MTA registration. First, the cable modem must receive a unique network address from a dynamic host configuration protocol (DHCP) server. Additionally, the cable modem must download a configuration file from a trivial file transfer protocol (TFTP) server. Both the DHCP and TFTP servers transmit data over the IP network. After cable modem registration, the IP layer is used by subscribers to transfer and receive IP data such as email, Web pages and gaming sessions, and MTAs for VoIP calls. IP layer impairments such as server failures, network congestion, data delay and jitter will provide a less than satisfactory user experience or may prevent the MTA from supporting reliable VoIP services. The IP layer also suffers from any RF- or DOCSIS-related impairments. For this reason, the IP layer resides at the top of the DOCSIS working model pyramid. Failures at any sublayer directly translate to apparent failures at the IP layer. Failures at the IP layer translate to network reliability problems and subscriber complaints. Achieving reliability DOCSIS protocol analyzers, when not being used for tactical purposes, can monitor downstream and upstream data communications of all DOCSIS devices. This information is archived to a database, which provides trend analysis information on cable modem/MTA MER, power, frequency, timing and equalization adjustments. User definable thresholds can be set that automatically notify appropriate personnel when device operation becomes critical because of any number of impairments in the DOCSIS pyramid. Consider a system with 1,000 cable modems. If just 3 percent of those modems are offline on average because of difficult-to-resolve problems, the result is $18,000 in lost revenue per year at $49/month/customer. Note that this does not include the troubleshooting costs associated with the multiple truck rolls and repeat customer complaints, nor does it consider the cost of possibly losing the customer to a competing service such as digital subscriber line (DSL). DOCSIS protocol analyzers are the cutting edge of resolving hard-to-find problems in DOCSIS networks. As with any new technology, an industry must learn new techniques and adopt new tools in order to make the technology successful. Improving DOCSIS network reliability requires new skills and new test equipment, in addition to the skills and test equipment we possess as an industry today. The tools and training for implementing and stabilizing DOCSIS networks are available. It is critical that cable operators invest in these tools to fully realize profitability from DOCSIS networks—video, voice and data. Achieving the triple play DOCSIS protocol analyzers serve as a critical element to achieving network reliability. Network reliability is the single greatest instrument to obtaining and retaining paying customers. It is the single greatest instrument to increasing revenue by decreasing cost associated with DOCSIS-related issues. It is the metric that will determine the successful deployment of advanced IP services beyond simple email and Web browsing. In order to compete with the telephone industry, it is no longer an option or a luxury—network reliability is a necessity for the cable industry. A reliable network translates into achieving the triple play. Brady Volpe is director of technical marketing at Filtronic Sigtek. Reach him at [email protected]. Figure3