Mobile television: strong, weak, or zero reception?
( 01 May 2008 )
By Brian Dipert, Senior Technical Editor, EDN
I am a firm believer in the theory that science fiction not only forecasts, but also shapes the future. Those of us who grew up watching The Jetsons may well remember the scenes of the family watching video on portable devices and in their folds-into-a-suitcase flying car. Granted, the Star Trek: The Original Series communicators were audio-only. However, communication and monitoring devices that follow-on Star Trek: The Next Generation episodes showcased, as well as those in other outer-space-themed television shows, such as Battlestar Galactica, and in science-fiction movies of the era, touted full audio-plus-video capability. And don’t forget about Dick Tracy’s 1964 two-way wrist TV, whose wrist-radio precursor foreshadowed the cellular phone.
If my theory is correct, consumers’ eventual widespread embrace of watching television while on the move should be a foregone conclusion—old news, at least in a sense. Sony in 1982 introduced its first Watchman, the monochrome model FD-210; the company had 22 years earlier developed its first “luggable” TV, 1960’s TV8-301 (Figure 1). Many of you have seen “boomerang” analog-television antennas atop limousines; unfortunately, the current analog-to-digital-TV transition complicates this usage scenario. In the analog age, multipath distortion and weak signals may have degraded broadcast images, but, in the binary era, those compromised images will go completely blank. And DirecTV and Dish Network subscribers may be aware of low-profile vehicle antennas that enable satellite-television reception not only when you’re sitting at a campground, but also when you’re traveling down the highway.
To label the current situation “widespread” would, however, be a stretch, despite the fact that I know someone who’s not a technology early adopter, but who uses a portable-analog television to catch up on a soap opera during lunch hour. Migrating from today’s early-adopter stage to one of broad adoption requires that service providers, their systems partners, and those partners’ silicon- and software-building-block partners address a lengthy, intertwined, and—at least at first glance—often-contradictory set of questions:
What systems will contain “third-screen,” or mobile, TVs, and what size displays will they have? (The third-screen device gets its name from the fact that it comes after the traditional TV, or “first screen,” and the computer, or “second screen,” in the time frame of consumer adoption.)
Will users risk a drained battery and consequent missed calls to receive television broadcasts on mobile phones or other multifunction devices, or is a dedicated-function product more appropriate?
Will the dominant human-transportation scheme be individuals in cars, in carpools, or using mass transit?
How long will potential customers be willing to wait to access a channel or to switch between channels?
Will consumers prefer short- or long-playing content?
Do consumers require real-time content, particularly as they grow more comfortable with delayed viewing courtesy of DVD sets, personal video recorders, Internet downloads from iTunes and other services, and network-Web-site-based rebroadcasts? Do users need individualized content? Does mobile-unique material hold appeal?
Will consumers expect to pay for this content on a per-minute, per-megabyte, per-episode, per-series, or per-month basis?
The answers to these questions will likely vary depending on the cultural and other norms of each target geography’s and country’s consumers. The last time EDN covered mobile television in this degree of detail, In-Stat analyst Michelle Abraham predicted that US subscribers to mobile-video services would cross the 5 million mark in the 2006 to 2007 time frame (Reference 1). Although accurate actual numbers are difficult to acquire, this forecast likely overstated reality by an ample margin. Nonetheless, the long-term prognosis is still optimistic. “By the end of 2010, mobile-TV-broadcast subscribers worldwide will have reached 102 million, a giant leap from 3.4 million in 2006,” Abraham now predicts (Reference 2). Near-term trends outside the United States are encouraging in this regard, with analysts estimating sales of 7.3 million mobile-broadcast-TV handsets in Japan through the first quarter of 2007 and more than 6 million subscribers in South Korea as of August 2007.
Dollars and sense An article covering mobile-television broadcast and reception must also cover cellular-data-standards evolution. As with wired-Internet access, including cable, DSL (digital-subscriber line), and fiber, video is a key driver of bandwidth, latency, and QOS (quality-of-service) improvements in wireless-communications protocols. In early 2005, US CDMA (code-division/multiple-access) cellular providers Sprint and Verizon offered only 2.5G CDMA 2000, or 1×RTT (one-times-radio-transmission-technology) service. Their GSM (global-system-for-mobile)-communications counterparts were primarily at the GPRS (general-packet-radio-service) level, and EDGE (enhanced-data-for-GSM-evolution) deployments were just beginning. Video-over-cellular per-frame resolutions were 176×220 pixels or less, and they accompanied “jerky,” less-than-10-frame/sec best-case frame rates.
A lot has changed in three years. Sprint and Verizon have both blanketed a substantial chunk of the United States with Revision A of EVDO (evolution-data optimized), a two-generations-faster cellular-data approach. Both AT&T and T-Mobile offer EDGE capability, and both carriers are ramping up next-generation UMTS (universal-mobile-telecommunications-system) networks that the companies base on HSDPA (high-speed downlink-packet access) and HSUPA (high-speed uplink-packet access). HSUPA has faster upstream-data rates than HSDPA, but those faster rates don’t notably apply to download-centric video-streaming applications. Given the bandwidth potential of these advanced protocols, are they now sufficient to reliably carry video streams in addition to voice traffic, SMS (short-message-service) blurbs, e-mail and Web-surfing bits, photographs, music clips, ring tones, and all of the other services that cellular providers now offer?
The answer is “probably not,” at least to the degree that the video streams match most potential customers’ capability and cost expectations, although future bandwidth increases may change this response. The bottom line is that the service providers are in the business of turning a profit to satisfy their shareholders and to cultivate funds for future R&D investments. Except for Verizon’s VCast TV, today’s mobile-video services, like traditional cellular-voice capabilities, are primarily unicast—that is, one-to-one technologies. Verizon bases VCast Mobile TV on Qualcomm’s MediaFlo technology. A table (in the Web version of this article, at www.edn.com/080207cs, summarizes the variable-cost challenge that service providers face and the financial underpinnings of the current service characteristics.
The financial model that drives the data that the table shows, along with the information in Figure 2, comes from the Mobile DTV (digital-TV) Alliance, an advocacy group that promotes mobile broadcast using the DVB-H (digital-video-broadcasting-handheld) noncellular scheme. Consider the source when interpreting the results. It uses the assumptions that a network operation’s variable costs are approximately 20 cents per megabit and that network operators charge consumers 99 cents—or, following the iTunes model, $1.99 for long-playing material—per episode to view video content. In analyzing the data, look first at the endpoints of the cost-versus-revenue spectrum. An average SMS message is 200 bits in size, translating to an average cost to the carrier of 0.004 cents. That same carrier charges its customer 7 cents per message on average, leading to roughly $375 of revenue per megabit of network-resource usage at a substantial profit margin. At the other extreme, consider a 30-minute, unicast, high-resolution, high-frame-rate video stream, comprising a 384-kbps average bit rate. The total payload is larger than 691 Mbits, translating to a highly unprofitable variable cost of greater than $138 versus $1.99 in revenue. The break-even unicast-video threshold, given the above assumptions, is a six-minute, 128-kbps video stream, whose low-bit-rate characteristics render it unable to deliver smooth frame rates of 20 or more frames/sec or resolutions that provide reasonably artifact-free presentations on large-screen devices, such as Apple’s iPhone. This fact is true even when you couple the unicast video with modern video codecs, such as H.264 and VC-1. But remember: The carriers aren’t in the business to just break even. That’s why two-minute-maximum, low-bit-rate unicast presentations are now the most common.
So far, I’ve focused only on variable costs. Cellular-network-infrastructure-construction-and-management fixed costs are equally important, and, in this regard, today’s cellular-data protocols are also not up to the task. Each unicast-video stream consumes network resources equivalent to those you would use in dozens or hundreds of voice calls. Cobble together insufficient network resources, and your customers will have unsatisfactory video experiences. Conversely, overbuild your network to handle video-“pull” peaks, and you’ll end up with substantial amounts of unused, cost-consuming resources during more typical and predominant demand times.
Bandwidth savior? Given these eye-opening economic realities, you can probably see why the DVB-H and MediaFlo camps are successfully advocating their alternative DTV schemes to cellular providers. Both DVB-H and MediaFlo unburden the cellular-data channel by shipping their video bits at alternative frequencies, often within the 700-MHz UHF (ultra-high-frequency) band, which, unlike cellular, can easily penetrate premises from external antennas many miles away (see box “Surf for CES and 700-MHz updates”).
“The 700-MHz range carries data farther at lower power than cell frequencies now in use,” says Neal Weinstock, chairman of system integrator Be Media and principal analyst at Weinstock Media. Weather, steel barriers, and other impediments less frequently interfere in this range. “It’s not perfect, and we know most—if not all—of its faults, but it is so much better than what the telephone companies have to work with now that applications riding on it will have a leg up in the marketplace,” he says (Reference 3). DVB-H and MediaFlo also migrate from unicast to more bandwidth-efficient multicast distribution.
“Ironically, not a lot has changed in MediaFlo and DVB-H,” EDN Editorial Director Maury Wright stated in November 2007 during an e-mail-based article-topic-brainstorming session and in reference to his Feb 3, 2005, article on mobile TV (Reference 1). Although he may have been correct from a technical standpoint, the US business dynamics of the two competitive low-power broadcast-television technologies have radically shifted in the past three years. This swing has generally been in MediaFlo’s direction, at least for the short term, although the jury’s still out on the long-term outcome. Consider that, in February 2007, AT&T announced its intention to launch a MediaFlo-based mobile-television service by the end of the year. As it turns out, the company did not meet its self-imposed deadline. Also, consider that Verizon launched the MediaFlo-based VCast Mobile TV service in March 2007 in more than 20 US markets at an incremental markup of $13 to $25 per month on top of cellular-service fees (Figure 3).
After a press demonstration in New York, which “coincidentally” coincided with the VCast Mobile TV rollout, Crown Castle in late July pulled the plug on subsidiary Modeo’s DVB-H aspirations, citing an inability to recruit a cellular carrier to align with it. This scenario occurred despite the media’s generally positive hands-on evaluations in the weeks following the demonstration. Instead, Crown Castle leased its 700-MHz-band spectrum to a joint venture for $13 million for unspecified applications. Ironically, this shutdown came after Crown Castle had persuaded the FCC (Federal Communications Commission), in late February 2007, to allow Modeo to provide a tenfold boost in broadcast-signal power in urban markets and a 20-fold boost in rural areas for its top 30 markets. Modeo competitor, albeit fellow DVB-H advocate, Hiwire, a subsidiary of Aloha Partners, began technical trials of its service in Las Vegas in December 2006, erecting the first towers during the week before Christmas. Hiwire, in partnership with handset developer LG Electronics and service provider T-Mobile, completed the network buildup by February 2007 and publicly demonstrated it at the April 2007 NAB (National Association of Broadcasters) conference. However, in early October, Aloha Partners sold $2.5 billion worth of its 700-MHz-spectrum licenses, along with Hiwire’s assets, to AT&T (Reference 4).
Although DVB-H’s fortunes might at first glance seem grim, the October AT&T acquisition of Aloha/Hiwire’s equipment and spectrum assets, along with the delay in AT&T’s MediaFlo-based service rollout, potentially suggests otherwise. At the April 2007 NAB conference, Stuart Pekowsky, DVB-silicon supplier DiBcom’s vice president of strategic partnerships, and Scott Wills, president and chief operating officer of Hiwire, pointed out that AT&T’s public MediaFlo plans at the time were limited in scope, with no long-term, widespread program commitment.
At the NAB conference, Wills provided data comparing his company’s network characteristics with those of his then-competitors (Table 1). The spectrum licenses that AT&T acquired in October can reach an estimated 196 million consumers, and they cover 72 of the top 100 US service markets. Why did AT&T make a deal with Aloha? Perhaps the company generally believed that the $2.5 billion it paid represented a better deal than it could get through the anticipated January-through-March-2008 FCC-run 700-MHz-spectrum auction, regardless of how AT&T ends up using the spectrum. Alternatively, AT&T may be instead planning a nationwide DVB-H rollout.
Note, too, that T-Mobile partnered with Hiwire to evaluate DVB-H in the Las Vegas network installation. T-Mobile hasn’t yet revealed its mobile-television plans. But, as a subsidiary of Deutsche Telekom, it may have a preference for DVB-H, which is a low-power, mobility-focused derivative of the DVB-T (digital-video-broadcasting-terrestrial) European DTV standard. Then again, it is also conceivable that, given T-Mobile’s German pedigree, Hiwire might select the alternative T-DMB (terrestrial-digital-multimedia-broadcasting) technology, whose developers based it on the DAB (digital-audio-broadcasting) scheme, and which is locally strong in portions of Europe, including Germany and the United Kingdom.
Despite regional opposition, however, the eventual pan-European acceptance of DVB-H appears inevitable. In November, the EU (European Union) mandated DVB-H as an official standard, a move that followed the organization’s formal endorsement in July of DVB-H and is reminiscent of Europe’s sweeping adoption of GSM digital-cellular technology. The group officially stated, “As a result, all EU Member States will have to support and encourage the use of DVB-H for the launch of mobile-TV services, thus avoiding market fragmentation and allowing economies of scale and accordingly affordable services and devices” (Reference 5). Even though DVB-H now has the EU’s blessing, it remains questionable whether DVB-H can achieve widespread adoption.
DVB-H employs unique broadcast spectrum and unique protocols, and today’s comparably pervasive DVB-T-only equipment cannot, therefore, tune in DVB-H. Its feature advancements over DVB-T include time slicing for power reduction. IP (Internet Protocol) data grams transmit in small-time-slot data bursts as large as 2 Mbits, and the DVB-H receiver is correspondingly in fully on mode only during these time slots. Over the several years that it took to develop DVB-H, however, its DVB-T predecessor achieved power-consumption improvements of its own by virtue of evolutionary circuit-design optimizations and Moore’s Law-fueled lithography reductions. Do DVB-H’s lingering power benefits justify its survival in the face of the DVB-T juggernaut? That’s a question that only market dynamics can address.
T-DMB may be losing momentum in Europe, but it’s the DTV technology of choice in South Korea, where it has experienced enthusiastic adoption, as its subscription numbers show. The ITU (International Telecommunication Union) also in December formally recognized T-DMB as a global standard, along with MediaFlo, DVB-H, and ISDB-T (integrated-services digital broadcasting-terrestrial). Mobile television’s acceptance in both South Korea and elsewhere in Asia reflects a combination of factors: widespread use of public transportation, lengthy daily round-trip commutes, and the sociological encouragement for users to cocoon themselves in an LCD screen and set of headphones while on public transit. China’s mobile-video plans are still in some degree of flux. Although the homegrown CMMB (China-multimedia-mobile-broadcasting) technology currently appears to have the inside track, DVB-H and T-DMB backers continue to lobby for their approaches, as do the developers of other China-proprietary approaches.
The name of the mobile variant of Japan’s DTV-broadcast technology, 1seg, reflects the fact that each ISDB-T channel subdivides into 13 segments, along with an incremental segment to separate channels. Japanese HDTV (high-definition-TV) broadcasts use 12 of the 13 segments; the remaining segment finds use for mobile reception. As you might expect, 1seg-based cell phones are increasingly commonplace in Japan, but plenty of other mobile-DTV options exist, including stand-alone receivers and portable-game consoles (Figure 4). Sony’s second-generation PlayStation Portable even supports video recording in conjunction with its companion 1seg adapter.
4G resurgence, etc Although cellular-service providers would prefer to focus their customers’ bandwidth usage on lucrative revenue and profit pursuits, a practical limit exists to how many e-mails, SMS messages, and photos even the most prolific smartphone addict can send and receive. Higher bandwidth 4G cellular-data services, such as GSM’s LTE (long-term-evolution) and Qualcomm’s UMA (unlicensed-mobile-access) technologies, are emerging, and their maturation drastically reduces cost per megabit for service providers and, therefore, for consumers. As a result, television services running on them may become more economically feasible than is the case with today’s 3G approaches. This reacceptance of video over cellular data is particularly palatable if 4G services encompass QOS support, a feature that today’s 3G technologies lack, along with provisions for bandwidth-thrifty multicast (one-to-many) protocols. At a minimum, cellular-data-based unicast may act as a value-added supplement to broadcast-digital-video-overlay networks, such as DVB-H and MediaFlo.
In August 2006, Sprint revealed plans to spend billions of dollars on a nationwide, unified mobile-WiMax-based network, which Sprint later branded Xohm, for voice, video, and data services. The company plans to launch that network in the second quarter of this year. Clearwire, another formidable US WiMax-services supplier, currently has an 802.16d-based network in 16 states as well as in Europe and Mexico, which the company plans to upgrade to fully mobile-capable 802.16e technology. South Korea’s branded WiBro is another notable WiMax deployment, with various companies’ and countries’ telecom providers performing additional WiMax trials worldwide. And you cannot understate mighty Intel’s long-standing support for WiMax.
Many countries’ and geographies’ COFDM (coded-orthogonal-frequency-division-multiplexing)-based digital-broadcast-TV standards are inherently mobile-capable. Conversely, the 8VSB (eight-level-vestigial-sideband)-modulation-based ATSC (Advanced Television Systems Committee) approach, which North America and a few countries on other continents use, is not inherently mobile-capable. Its developers defined it predominantly for use with home-based receivers coupled to 30-ft roof-mount antennas (Reference 6). The incremental evolution of 8VSB-based receivers has minimized multipath-interference effects in stationary configurations, but mobile applications remain prone to egregious reception setbacks. Two backward-compatible enhancements aspire to improve ATSC’s mobility capabilities. They also aim to garner patent-royalty-rich incorporation within the ATSC-M/H (ATSC-Mobile/Handset) specification, which is currently under development. Samsung and Rohde & Schwarz champion one of these enhancements, A-VSB (advanced VSB), and Harris and LG Electronics back the other, MPH (mobile pedestrian handheld).
All of the previously mentioned services rely on terrestrial-broadcast antennas and, as a result, operate only as far as those antennas’ signals reach. Satellite broadcast is the preferred approach for pervasive coverage across an entire country or geography—that is, as long as nothing is obscuring the overhead view. Sirius is the latest player in this market, supplementing its satellite-radio service with the child-tailored content of SiriusConnect Backseat TV, which Sirius launched in partnership with silicon supplier STMicroelectronics as an option on several Chrysler vehicles for the 2008 model year. The DVB-standards body also has its eye on satellite, having approved in February 2007 the DVB-SH (DVB-satellite-and-handheld) standard, which blends support for low-power terrestrial reception when available with more encompassing—albeit quicker to drain batteries—satellite reception. DMB backers similarly support the S-DMB (satellite-DMB) standard, which began operating in South Korea in May 2005, and China’s CMMB also encompasses support for both terrestrial and satellite broadcasts.
For more information, tune in to EDN’s March 6, 2008, edition, in which a follow-on feature article will delve into each of these topics in greater detail.
Author Information You can reach Senior Technical Editor Brian Dipert at 1-916-760-0159, bdipert@edn.com, and www.bdipert.com.
AT A GLANCE · Mobile television is a seemingly perpetually "almost mainstream" application that the current analog-to-digital-broadcast transition further complicates. · The 3G cellular-data networks lack the economic chops to satisfy most potential customers' video-delivery expectations, but the picture may become clearer in the 4G iteration. · MediaFlo seems to have most of the US mobile-broadcast-television momentum behind it at the moment, but you still can't disregard the "muscle" of Europe's anointed technology, DVB-H (digital-video-broadcasting-handheld). · Outside North America, mobile-television adoption is more mature, but a diversity of standards and cultural expectations complicates system- and building-block providers' strategies. · See EDN's March 6, 2008, issue for more on ATSC (Advanced Television System Committee) add-ons, WiMax, satellite-based candidates, and CES (Consumer Electronics Show) and 700-MHz auction updates.
References · Wright, Maury, "Mobile video: Participants follow multiple paths," EDN, Feb 3, 2005, pg 27, http://www.edn.com/article/CA498772. · DiBcom 2007 press kit, http://www.dibcom.info/DiBcom.zip. · Weinstock, Neal, "The 700 Club," Television Broadcast Magazine, August 2007, pg 38, http://www.televisionbroadcast.com/. · Marek, Sue, "DVB-H may not be dead yet," FierceWireless, Oct 10, 2007, http://www.fiercewireless.com/story/dvb-h-may-not-be-dead-yet/2007-10-10. · Melander, Ingrid, "Nokia-led DVB-H to be EU mobile TV standard," Reuters, Nov 29, 2007, http://today.reuters.com/news/articlenews.aspx?type=technologyNews&storyid+2007-11-29TI21827Z_01_L29516007_RTRUKOC_O_US-EU-TELECOMS-TELEVISION.xml. · Cugnini, Aldo, "World TV standards," Oct 1, 2007, http://www.broadcastengineering.com/hdtv/world_tv_standards.
Surf for CES and 700-MHz updates Because of tight deadlines, I was unable to incorporate in print any relevant news coming from the early-January 2008 Consumer Electronics Show or the results of the FCC’s (Federal Communications Commission’s) Jan 16, 2008, 700-MHz-spectrum auction. But fortunately, EDN’s Web site takes up the slack. Hit the Brian’s Brain blog for timely commentary on these and other pertinent mobile-TV topics.
Figure 1 Think mobile-television reception is new news? Think again. Sony’s TV8-301 “luggable” TV dates from 1960 (a), and the company introduced the first-generation FD-210 Watchman in 1982 (b).
Figure 2 Economic factors define cellular providers’ current reliance on low-bit-rate, short unicast video clips, as well as their encouragement for customers to send plenty of SMS messages and take lots of pictures (a). Video also consumes a substantial percentage of the cellular network’s available resources (b) (courtesy Mobile DTV Alliance).
Figure 3 MediaFlo-cognizant handsets from LG Electronics (a), Motorola, and Samsung (b) were ready to go at Verizon’s VCast Mobile TV service launch last March.
Figure 4 Nintendo’s DS Lite (a) and Sony’s second-generation PlayStation Portable (b) now support Japan’s 1sec mobile-digital-video service, a subset of ISDB-T, through adapter add-ons.
