As operators move toward the vision of all digital networks and convergence of voice, video and data, the network edge becomes even more critical. The move to provide a plethora of everything on demand (EOD)-type services not only creates additional data capacity needs, but puts more emphasis on the necessity of high availability at the edge. Cable modem termination systems (CMTSs) will need to continue their evolution to meet these needs.

The current generation CMTS utilizes distributed processing combined with additional technologies to focus on density, high availability and observability. The next evolution must build upon this foundation to replace today’s labor-intensive operations with more advanced automated intelligence for subscriber and network management. This could come in the form of peer-to-peer (P2P) traffic management, packet inspection for congestion mitigation and configuration flexibility.

Future CMTSs also must be capable of supporting a large list of new services, including the following: PacketCable voice over IP (VoIP); PacketCable-Multimedia gaming, video over IP and content downloading; advanced data capacity management to provide guaranteed throughput to subscribers; and even combined Moving Picture Experts Group (MPEG) video transport. In addition to these services, it is also quite apparent that future cable subscribers will demand more throughput for higher definition video applications and higher-speed data downloads.

As the CMTS continues to take on new services and demands, it also is imperative that it have the ability to process the advanced routing protocols needed in tomorrow’s highly complex networks. Whether an operator chooses multiprotocol label switching (MPLS), open shortest path first (OSPF), intermediate system-intermediate system (IS-IS) or even border gateway protocol (BGP) in the metro network, the CMTS must be flexible enough to evolve with the industry at what seems like breakneck speeds.

In addition, future CMTS architectures must be extremely flexible to accommodate rapidly evolving standards. As an example, the recently released DOCSIS Set-Top Gateway (DSG) specification may lead ultimately to dramatic changes in the number of upstream service flows that must be supported in a single media access control (MAC) domain. This will require architectural plasticity within the CMTS to support these and other growing demands of the industry.

Future CMTS capabilities

Moore’s Law (which predicts that the density of gates and memory in silicon integrated circuitry will double every 18 months) essentially guarantees that technology always improves over time. This phenomenon permits the integration of many new functions into a single subsystem even though the functions previously required separate subsystems. This integration of both analog and digital subsystems will have a profound effect on future CMTSs, and several CMTS features are likely to change as a result of this integration. These changes are likely to occur in several different areas, as outlined below. Upstream/downstream data capacities Current chassis-based CMTS products typically support less than 40 downstream channels and less than 200 upstream channels. Increased integration and advanced chipsets that support many transmitters/receivers likely will permit much higher per-chassis channel counts in the future. The actual bit-rate supported on each downstream and upstream channel also is likely to increase through the application of more advanced PHY technologies. The integration that permits this increase in density also will lead to decreases in per-downstream CMTS prices. In addition, current CMTS products typically support cards with a fixed number of downstream and upstream channels, and the channels within a single MAC domain are pre-defined by the cards. Future CMTS products will permit much more flexibility on the per-chassis upstream and downstream channel counts. In addition, software configuration will be able to dynamically define the number of channels that are bound together within a single MAC domain. This flexibility is important because different types of service offerings are likely to demand very different bandwidth mixes on the future HFC plant. For example, in the past few years, high-speed data services transitioned from requiring asymmetric (higher downstream) channel bandwidth and data throughput to requiring more symmetric performance as a result of the ubiquity of P2P file transfers. (See Figure 1) However, the future arrival of DSG functionality and PacketCable-Multimedia-based IP video applications is likely to push these requirements back toward the asymmetric bandwidth conditions of the past. In fact, the upstream:downstream ratios of today typically are greater than 1.1 (4:1, 6:1, or 8:1 are typical), but the future services actually may force the upstream:downstream ratios to fall below 1.1. As a result, future CMTSs must be capable of supporting this wide range of requirements. Data management Future CMTS products undoubtedly will add an abundance of data capacity and throughput management capabilities. For example, most CMTSs likely will capitalize on the DOCSIS dynamic channel change (DCC) feature for intelligent load balancing and channel noise avoidance. Another throughput-preserving benefit of DCC will be realized by some CMTSs that add the ability to move subscribers to downstream channels that are carrying dedicated broadcast content streams, eliminating the wasteful replication of multicast flows on multiple downstream channels. To help minimize the rate at which cable operators perform node-splits, future CMTSs may very well support complex modulation formats that go beyond the capabilities of today to provide more data throughput to each fiber node. For example, 512 QAM or 1024 QAM may find useful applications as higher-speed downstream modulation types within future CMTSs. CMTS data management will see many other innovations in the future. For example, future CMTSs are likely to include features that permit monitoring, identification and control of P2P and other bandwidth-intensive applications. Policing of HFC bandwidth used by different PacketCable-Multimedia content providers also will be relegated to the future CMTS. Operations support Future CMTS products will be called upon to operate in more complicated networking environments, so CMTS vendors undoubtedly will add an array of tools to help operators monitor and control their systems. The development of these tools may require CMTS vendors to work more closely with operations support system (OSS) vendors to ensure the tool sets stay ahead of industry needs. These systems may include channel analysis software that guides cable operators to the optimum configuration for the given HFC plant characteristics. Another set of tools continually will monitor the user traffic flows and traffic losses and compare those readings with HFC topology information to help identify potential HFC problems. Other tools likely will be added to help operators identify problems by probing the connections to various network elements using network-level diagnostics. In general, bundling these OSS features into a single box will not only help to decrease growing operational expenses, but also will reduce dramatically mean time to repair (MTTR) providing customers with increased availability. Increased integration Because of the power of Moore’s Law, future CMTS products also may have the processing power to integrate many of the functional subsystems that currently surround them. Economics and operational impact ultimately will determine which subsystems are included, but possible candidates include the combiners, the optical lasers, the edge-QAM modulators and their associated upconverters, and the provisioning servers (time of day (TOD), dynamic host configuration protocol (DHCP), trivial file transfer protocol (TFTP), etc.). It is even possible that vendors may choose to combine PacketCable and PacketCable-Multimedia subsystems into the CMTS. Candidate subsystems include call management systems, signaling gateways, media gateways, application managers and policy servers. Conclusions Future CMTS designs will support a plethora of functions that were unheard of only a few years ago. These functions will be required by cable operators as they begin to provide converged services in the future. In general, operators can look forward to more functionality and lower per-subscriber costs from their CMTS vendors as the evolving technologies begin to appear in more CMTS product designs. Tom Cloonan is CTO of ARRIS. Email him at
John Treece is director of new product technology development Comcast IP Services. Reach him at Did this article help you? Email comments to Bottom Line: CMTS Evolution
As operators move toward converged services, the network edge becomes even more critical. This move creates the need for additional data capacity and high availability at the edge. Cable modem termination system (CMTS) products must continue to evolve to meet these needs. The CMTS of the future could list many new functional items on its resume, including the following: Figure 1: Trends in CMTS Data Throughput Requirements

The Daily


Representation Matters: Fewer Women, People of Color on TV

Nielsen released its first-ever report of the television media landscape’s progress and gaps in on-screen inclusion.

Read the Full Issue
The Skinny is delivered on Tuesday and focuses on the cable profession. You'll stay in the know on the headlines, topics and special issues you value most. Sign Up