Mobile DTV

The U.S. television broadcast world will undergo major changes in the months ahead with the end of analog broadcasting and the beginning of mobile DTV service by some broadcasters. The industry has been abuzz for the past year, with demonstrations of mobile technology from several companies. The ATSC Technical Standards Group (TSG) has been busy, along with the Open Mobile Video Coalition (OMVC), with the selection of technology and documentation leading up to an ATSC Mobile/Handheld (M/H) standard.

Several years ago, broadcasters and equipment manufacturers began to realize that over-the-air broadcast television would soon see competition from new players who wanted to reach an ever-growing audience of on-the-go viewers. The competition was coming from companies that planned to use the 700MHz spectrum that was recently auctioned off by the FCC and from 3G wireless operators. Broadcasters realized that they needed to compete, but there was one problem: The ATSC DTV system was not designed to reach mobile devices.

In the mid-1990s, when the ATSC DTV system was developed, the main focus was to achieve as much digital payload as possible within the 6MHz of allotted RF spectrum. This was necessary to allow for the then-nascent MPEG-2 HDTV encoding. As technology advanced, the bit rate necessary for HD encoding decreased, allowing broadcasters to dedicate bits for additional services, such as multichannel or mobile.

ATSC-M/H technology overview

The ATSC-M/H system is based on technology developed by Harris, LG and Zenith in their MPH system, blended with some parts of the A-VSB technology developed by Samsung. The system technology is made up of three layers of activity:

  • the physical layer, which is all about getting the bits from the station to the handheld and mobile receiving devices;
  • the management layer, which includes signaling and announcement information on services, conditional-access system (CAS) information, digital rights management and electronic service guide (ESG) information; and
  • the presentation layer, which includes audio and video encoding, closed captioning and interactive applications.

The mobile services are based on IP transport rather than legacy MPEG. However, the transport of the M/H data from the integration point with the conventional ATSC signal to the transmitter exciter requires that the IP datagrams be encapsulated into the MPEG-2 transport. The IP-based transport methodology allows for an easy integration of both real-time and non-real-time services. It also offers support of multiple program streams in each M/H channel or “parade.”

The magic is in the physical layer

Each M/H channel can typically carry about 600Kb/s of payload. Depending on the level of robustness selected by the system operator, each M/H channel will occupy about 2Mb/s of the main ATSC stream due to the additional channel processing required to make the mobile data withstand the rigors and impairments found in mobile reception.

Program content is encoded using H.264 v1.3 for video and HE-AAC for audio. Video resolution is scalable up to 416 × 240 supporting 16:9 aspect ratio presentations, and the audio is stereo with future capability for surround. All real-time streams are encapsulated using RTP/RTCP, while non-real-time content is encapsulated using File Delivery over Unidirectional Transport (FLUTE) protocol.

Rather than create all services from scratch, the ATSC TSG S4 working group has selected elements of services, such as the ESG from existing mobile standards like the Open Mobile Alliance Mobile Broadcast Services Enable Suite (OMA BCAST). More details on the various services will be revealed by the ATSC in the near future.

The M/H data needs to be specially coded to serve the M/H receiver for reception under rapidly varying signal conditions, and at the same time appear the same as ordinary 8-VSB data to legacy receivers so that they are not disturbed. This is accomplished by precoding the M/H data and then passing it through the legacy ATSC process in the form of normal-appearing data packets.

ATSC-M/H system architecture

The precoding is done at two levels, first as data bytes and then as channel symbols. M/H data bytes are cross-interleaved and coded with both Reed-Solomon and cyclic redundancy check (CRC) codes. The data and its code bytes are packaged into normal ATSC data packets, which then are processed in the usual legacy process, including interleaving. This changes the solid block (or “group”) of M/H data to a sawtooth-shape region for transmission. In the second level of coding, normal ATSC trellis coding is augmented with a serial concatenated convolutional code (SCCC) for the M/H data. (See Figure 1.)

The group of M/H packets also contains known data sequences at regular intervals. These training sequences allow the M/H receiver to make accurate and frequent estimates of the channel multipath conditions. In fact, the M/H receiver can successfully receive a single, isolated group because of its instantaneous measurement of the channel.

The ability for instantaneous turn-on and acquisition allows power saving in handheld, battery-powered devices. Various programs can be carried in separate parades of groups, and the receiver RF front end only needs to draw power during relevant times. (See Figure 2.)

ATSC-M/H system architecture

Integrating ATSC-M/H systems into an existing ATSC broadcast chain is relatively easy, as was demonstrated during the OMVC Initial Demonstration of Viability (IDOV) testing that took place earlier this year.

At the studio end of the system, an M/H multiplexer is integrated into the ATSC stream after the existing ATSC multiplexer. Depending on the manufacturer, the multiplexer could include an integrated IP encapsulation system and also integrate operational and maintenance data packets to control the various M/H modes created in the M/H exciter. The multiplexer may also perform some packet timing functions to prevent buffer overload at the receiving end. The multiplexer output format is a conventional 19.4Mb/s stream based on either ASI or 310M format. No changes are required for the station's STL.

Each real-time M/H program stream will require an IP-based mobile program encoder that will send the mobile signals to the encapsulator and multiplexer. Non-real-time (NRT) content requires an authoring station and server for storage and a management system to schedule and release the NRT content. Because the connectivity between these content sources and the encapsulator/multiplexer is IP-based, a reliable IP link can allow the content sources to be separated from the transmission system. The ESG system, CAS and other advanced applications systems all connect to the M/H encapsulator/multiplexer using IP as the transport. (See Figure 3.)

At the transmitter, the ATSC exciter needs to be replaced with one that can support M/H. In recent months, manufacturers have introduced new software-based exciters that can be upgraded to ATSC-M/H service. Older exciters are not upgradable because of the amount of processing power required to create the M/H signal.

There are several major RF considerations that a station must review when entering the mobile business. ATSC-M/H is an RF-delivered wireless business with no delivery help from cable or satellite. If the transmission system fails, the signal is off-air, making redundancy crucial. Redundancy should cover the entire chain from encoders through to the transmitter.

While UHF works best for mobile DTV, high-band VHF has also been tested and works quite well. There are receiving antenna shortcomings at VHF that introduce some reception limitations.

Transmit antenna polarization may also be a factor because most portable and mobile receiving devices have vertical, polarized receiving antennas. At least one broadcast antenna manufacturer reports that there has been a recent increase in interest by broadcasters looking to add vertical polarization to their signal.

Finally, broadcasters should consider maximizing their power and providing as much signal coverage as possible under the FCC rules. During the recent FCC window of opportunity to maximize, more than 600 stations applied for maximization.

With the ATSC hoping to complete the standards process by early 2009, new broadcast equipment coming this fall and consumer devices planned for later in 2009, this coming year will be busy for broadcasters looking to enter the mobile DTV business.

Jay C. Adrick is vice president of broadcast technology for Harris, and Wayne E. Bretl is senior principal engineer for Zenith.