Seamless ad insertion
Aug 1, 2008 12:00 PM, By Yaniv Ben-Shushan
Target ads to a specific audience with DPI.
Figure 1. DPI architecture
Select image to enlarge
Over the past several years, digital program insertion (DPI) has enabled seamless commercial ad insertion into an MPEG stream. Service providers realize the business potential that lies in this technology, as well as the quick return on investment, and are expanding and investing in more digital splicing channels.
To date, the technology for regional ad insertion has mainly been based on MPEG-2 codecs — a mature technology that is widely available. However, digital video solution providers are developing products that will enable digital ad insertion not only on SD MPEG-2 services, but also on HD and advanced video codec (MPEG-4 AVC or H.264) services.
The transition to the new codec — in conjunction with the introduction of IP networking technology for video delivery — has provided an opportunity for enhanced and updated ad insertion solutions, resulting in further advancements in this space.
Breaking it down
The overall ad insertion system consists of many components that must interact according to certain expectations. (See Figure 1.) The typical system components are traffic and billing, automation, digital compression, a video server, and a digital splicer.
Many years ago, companies produced general-purpose splicers that functioned without decompressing the MPEG-2 signal. Interestingly, they performed splicing in multiple-program transport streams (MPTS), as well as in single-program transport streams (SPTS). When it became clear this technology functioned well, broadcasters' interest in it grew rapidly. They realized that if the splicing triggers were carried in digital form, new and expanded capabilities would also become possible.
Select image to enlarge
A set of standard communication protocols between the automation system and the compression system, the compression system and the digital splicer, and the ad server and the digital splicer were established. The protocols make sure that all the components can easily be integrated and work together, no matter which vendor's equipment is being used.
Drafting of the standards began several years ago, focusing initially on the splicer/server portion of the system and permitting continued use of the existing contact closure mechanisms. The standards for the in-band signaling between the compression system and the splicer/server were devised shortly thereafter.
The Society of Cable Telecommunications Engineers (SCTE) created the following standardization protocols:
SCTE 35 defines the cue messages that are embedded in the transport stream (TS). It also defines the upcoming splice points and other timing information.
SCTE 30 defines the communications between a splicer and the ad server in response to the messages defined in SCTE 35.
SCTE 104 defines the communications between an automation system and the compression system that will insert SCTE 35 private sections into the outgoing transport stream.
Integration
Now let's try to understand how the system works and how all the parts come together. The signal flow begins with the compression system that receives live content and compresses it in an MPEG stream. The automation server triggers the compression system to place SCTE 35 messages in the TS. The positions of advertising avails are signaled by SCTE 104 from the automation server to the compression system.
The digital splicer extracts the SCTE 35 message from the TS and then passes the splice request to the ad server over a SCTE 30 message. The ad server receives the schedules from the traffic and billing system. The schedule contains the information that specifies the channel and which ad to insert at which time. The ad server follows these instructions when the splicer requests an ad. The splicer inserts material from the server in place of the incoming transport stream in response to SCTE 35 commands embedded in the TS.
Customer expectations of a fully digital ad insertion system's performance are fairly high. Moving from a traditional ad insertion architecture, using baseband, to a fully digital system introduces the need for a more seamless splice because of the improved video quality. It is worth noting that the delivery of clean switches is the responsibility of the splicer involved, rather than the encoder or server.
Targeted ad insertion
The existing MPEG-2 ad insertion technology was designed and developed using rate shaping (requantization or re-encoding) as its starting point. Rate shaping is widely used by service providers who want to groom a set of channels into a fixed bandwidth pipe. It is therefore no surprise that the majority of ad insertion deployments to date have been within the cable industry, as its hybrid fiber-coaxial (HFC) architecture requires delivering video in fixed bandwidth pipes known as quadrature amplitude modulations (QAMs).
The fact that more cable networks are segmented is one of the biggest advantages with respect to digital ad insertion, or to be more specific, regional ad insertion. In fact, the first motivation for segmenting the cable network was to offer on-demand services, but soon after that, operators discovered that they could use this architecture for ad zones. A large cable operator has the ability to insert different ads in various geographical locations on the same program.
This network topology enables pay-TV operators to insert the right ad to the right target channel at the right time to gain the maximum impact, which enables them to charge a premium for the ad insertion service. However, this kind of service needs to be cost-effective in order to be successful. The ad servers need more capacity, more streaming throughput and the right pricing structure. The same goes for the digital splicers, which require a dense architecture and more throughputs. This is the first step toward targeted ad insertion.
While DPI technology was rapidly adopted by the North American cable market, the international market is moving toward this technology at a much slower pace. The main reason is the number of channels the MSOs are allowed to splice into, which is increasing and will soon make it worthwhile for those operators to invest in a DPI solution. However, longer term, the industry is likely to transition to a fully unicast IP-based environment that will incorporate targeted ad insertion.
Yaniv Ben-Shushan is product marketing manager for Harmonic.
| Want to use this article? Click here for options! |  |
