Modern cable networks provide dynamic digital video transport that is a mix of linear, switched, and on-demand MPEG2 transport streams, as well as multicast, narrowcast and unicast IP video streams carried over DOCSIS, Wi-Fi, or carrier Ethernet. The customer quality of experience is much more diverse for IP video, and it has the potential to compete with conventional web browsing and other high speed data services. But it also offers lots of benefits to cable operators–as long as the right technology infrastructure is in place and the workforce knows how to use it.

The business reasons given for this evolution of video transport infrastructure include: delivering content to laptops, tablets and smartphones; greater transport efficiency via variable bit rate (VBR) encoding schemes; advertising revenue growth; improved rights management and reduced theft of service; and of course moving to an all-IP network infrastructure to improve and integrate back office operations and support systems, and ultimately to enable a 360 degree view of the customer for continuous improvements in customer experience. 

The IP video infrastructure is complex, but the industry has been working on this for many years now. Unicast IP delivery of video on demand (VOD), multicast IP delivery of broadcast video, video telephony/telepresence and the need for improved quality of service (QoS) mechanisms for IP video were all key drivers of DOCSIS 3.0 and PacketCable 2.0. And concerns over scale issues for IP video through CMTS equipment led to proposals for bypass and modular CMTS architectures and to the converged cable access platform (CCAP) specification. 

While 3.0 has been out for a while, many MSOs are still either in the process of, or are planning, their migrations to IP video transport. New DOCSIS Engineering Professional (DEP) and Digital Video Engineering Professional (DVEP) courses from SCTE can help ensure that DOCSIS and digital video engineers understand the fundamentals of IP video transport as it is currently implemented so they can leverage things like IP statistical multiplexing with variable bit rate transport and utilize existing CMTS capabilities to deliver more video streams in less bandwidth. 

They need to understand how dynamic bandwidth sharing allows them to increase bandwidth utilization for all services; how improved admission control and QoS in DOCSIS 3.0 allows them to reserve DOCSIS bandwidth and ensure that overall customer quality of experience is maintained across the diverse content sources and network paths to the customer; and how cable modem load balancing can be used to support multiple bonding groups per service group as IP video traffic grows.

These IP video transport features just mentioned were added to the DOCSIS 3.0 specification. In 3.0, popular channels can be statically RF spanned and sent as static multicast flows for broadcast of linear TV channels to IP devices, while longtail content that was previously sent via switched digital video (SDV) can be transported via dynamic narrowcasting to active viewers only. Other multicast IP video transport enhancements include: support for source specific multicast (SSM), which reduces demands on the network and improves security; support for version 3 of the Internet Group Management Protocol (IGMPv3); IPv6 multicast support (pre and post registration); support for multicast QoS to improve the video quality of experience on IP devices like laptops, tablets, and smartphones; support for bonded and non-IGMP based multicast; support for multicast authorization and multicast encryption; and explicit tracking of multicast listeners. 

The PacketCable Multimedia (PCMM) framework and PacketCable 2.0 specification added an improved way to obtain QoS for video over DOCSIS that was agnostic for signaling and even more generic. MSOs can generate policy controls on the flows, and PCMM provides a generic application framework that can be used to enable video QoS even on devices that are not QoS aware, such as soft-phones and gaming consoles. This is important since gaming applications increasingly require high definition video as well as low-latency.

In the DOCSIS 3.1 specification, CableLabs chose an implementation of orthogonal frequency division multiplexing (OFDM) that offered lower latency for interactive video and high speed data applications than the OFDM used in DVB-C2 for example. It also added active queue management (AQM) to alleviate the impact that buffering has on latency-sensitive traffic, while preserving bulk throughput performance. Special MAC improvements for video over DOCSIS 3.1 were added such as the ability to rate limit to an aggregate limit all service flows (SFs) of a specific service type across all cable modems on a CMTS. Cable operators will be able to specify that the set of all video SFs on a MAC domain do not consume bandwidth over a set limit so that they do not crowd out the non-video high speed data SFs.

Modern digital video and DOCSIS engineers must know about these features as they transition their networks to some- and ultimately all-IP video transport, SCTE’s new online courses for DEP and DVEP, as well as the existing Overview of IMS and SIP - PacketCable 2.0 SCTE course, can provide the training engineers need to get the most from modern cable networks.