Moving Pictures

Although the promise and the reality of wireless video are light years apart, some companies are pioneering promising solutions.

Remember AT&T's Picturephone (www.att.com) from the 1960s? Weighing 26 pounds, it debuted at the 1964 World's Fair but bombed commercially. European carrier Orange (www.orange.com) recently introduced a wireless equivalent. Weighing 500 grams (hefty by today's portable standards), the Videophone can be ordered for about $1,900. One early reviewer called it “a monstrosity.”

If reaction to the Orange Videophone is taken as an example, wireless video has a long way to go before mass adoption. With the low bandwidth of today's 2G networks, emerging wireless video applications are herky-jerky imitations of what consumers have come to expect from the word “video.”

Yet to hear some streaming-media companies tell the story, commuters soon will be watching the local TV news on their cellular phones while riding the train to work. The true state of wireless video lies somewhere between this utopia and “the monstrosity.”

Evolution, Not Revolution

Don Grises, Cap Gemini Ernst & Young (www.cgey.com) director of wireless alliances and product development, cautioned carriers against falling into the WAP trap — where expectations outstrip performance — with streaming video.

“We have an evolution happening in the industry,” Grises said. “The steps the industry can take are better education and management of the public's expectations — both consumer and enterprise. These (video-enabled) devices can be very valuable in the right circumstances.”

But there are many technological issues to be solved before streaming video can be an effective service for the majority of subscribers. The issues include handset limitations, bandwidth constraints, network reliability and management, available content and standards compliance. And that's just for 1-way streaming, or retrieving a video clip on demand. The challenges multiply — especially for the handset — with 2-way videoconferencing.

“Visual media add a lot to the wireless experience, and it doesn't have to start and end at video streaming,” said Azita Arvani, ActiveSky (www.activesky.com) vice president of business development and strategy. “At the start, we will have lower bandwidth, and there we can enhance the experience by adding images and short-clip animation. As the networks become better, and the telcos understand better how much of that they can allocate to multimedia and video, we can move into full-motion video.”

Although streaming video is possible today using some PDAs, the penetration of these devices is low compared to cellular phones. So carriers and handset vendors are looking to improve wireless phones by increasing the screen size, adding color to the display and adding memory, processor power and battery strength.

There are a few new handsets capable of receiving streaming-video content today. Kyocera makes a video phone (www.kyocera.com) for the Japanese market, and Samsung (www.samsung.com) has a color-video model currently undergoing trials in Korea.

The MPEG-4-capable Samsung SPH-X2000 incorporates the A2 video ASIC chip, designed by Emblaze Systems (www.emblaze.com). Although Korean trials have been ongoing for five months, commercial deliveries of the SPH-X2000 have not yet been announced for Europe or the United States.

Sasson Darwish, Emblaze president, said the SPH-X2000 can play video over many current IS-95B and IS-95C networks. The video phone currently is compatible with existing CDMA and 1X networks at speeds from 9.6kb/s to 144kb/s.

The applications Emblaze is testing in Korea right now include adult entertainment, location-based information such as weather and traffic, and 1-way video messages, Darwish said.

Future Samsung video-enabled phones will use Emblaze's next-generation A3 chip to allow for recording messages or scenery with an embedded digital camera. Darwish said Emblaze soon would announce a camera-manufacturer partner for the A3 project.

Preparing for the Future

Although video can be streamed over 2G networks today, the maximum number of frames per second is still in the low single digits. For mass adoption by users accustomed to television and film, faster networks will be required.

“Currently, most systems max out at 9.6kb/s, 14.4kb/s or 19.2kb/s over the air interface,” Grises said. “Even with 19.2, it takes forever to download a clip. All the major carriers in the United States have announced some form of the next step toward 3G this year, whether it's GPRS or 1XRTT.” But carriers will have to allocate higher bandwidths effectively to subscribers with advanced handsets, Grises said.

“Once you go to a packet interface, you're able to vary the amount of bandwidth you can give an individual customer,” he said. “It's not as flexible and robust as when you get to 3G. But it's definitely an improvement.”

In addition to implementing 1XRTT or GPRS, carriers have to deploy a system to manage video services and convince subscribers to pay for them.

“Just because you turn up GPRS doesn't mean you can get video all of the time, everywhere,” he said. “You just have the capability.”

The carrier's system has to make a decision on how to manage that traffic — this means balancing optimization against whatever service-level agreements you have in place with enterprises.

Wireless video platforms capable of this type of dynamic-bandwidth management are available from a number of vendors today. At Sprint PCS (www.sprintpcs.com), discussions are under way with a number of vendors to trial video-streaming solutions.

“One of the most important things to be able to support video is to increase the throughput or speed of our network,” said Ryan Slack, Sprint PCS senior director of market strategies. “And we have aggressive plans to deploy 3G technology in our network later this year and into early next year. Right now, speed before compression is 14.4kb/s. After our upgrade, the core speeds before compression will be up to 144kb/s.”

Slack said 144kb/s is fast enough for some streaming-video applications.

“There are certain apps that make sense in three to four frames per second,” he said. “And there are some apps where you need a TV-like signal.”

Although entertainment-oriented applications have proved popular in Japan, early wireless video applications in the United States most likely will focus on businesses, Grises said.

“It all depends on what the content owner is trying to convey with that piece of content and what value the consumer has ascribed to that format,” said Dale Knoop, Sprint PCS manager of new business development for wireless multimedia services. “If they've ascribed no value to a still frame with audio under it, say a picture of Peter Jennings with an audio track, then you would move upstream with your format and cause your bandwidth needs to rise as well. There will be a lot of experimentation around content formats and what people value.”

Beyond more speed and access to the right content, carriers must implement a viable platform to support wireless video.

Sprint PCS already has anounced partnerships with PacketVideo (www.packetvideo.com) and SolidStreaming (www.solidstreaming.com). However, Knoop said those partnerships have not yet reached the field-trial stage.

PacketVideo unveiled its PVPlatform 2.0, and the company has announced a field trial with Swisscom Mobile (www.swisscom.com).

The platform consists of three components: PVAuthor for encoding MPEG-4 content; PVServer for integrated billing, provisioning and authorization on Solaris, HP-UX and Linux platforms; and PVPlayer for decoding content on the handset at speeds up to 384kb/s.

The SolidStreaming System also encompasses the encode, server and decode functions needed to provide a full-scale wireless video solution. It includes the VGenerator, which converts multimedia content in formats supported by the VServer and VPlayer from input devices such as Web cams and video-capture boards.

ActiveSky approaches streaming-media content and delivery from a slightly different angle, Arvani explained. Although the company's solution does include encoding, a server and decoding, the decoding function is not chip-based. Users go to the ActiveSky Web site to download ActiveSky Media Player.

The company's philosophy is that streaming media will not begin with video but, rather, will progress toward video along a path beginning with simpler graphics.

“This is going to be an evolution, not a revolution,” Arvani said. “We're not going to go from having no multimedia to all of a sudden streaming video on these phones.”

The Birth of Wireless Photography

Although Robin Nijor thinks there will be a demand for wireless video, he thinks there's a more pressing demand for wireless still photography. Nijor, who is vice president of marketing for LightSurf (www.lightsurf.com), said the company has developed an embeddable digital eye module that can capture and transmit a photograph over wireless networks.

LightSurf was founded by Philippe Khan, who was awaiting the birth of his child when he hatched the idea for the company. Wanting to share the joyous moment, he realized that by linking his cellular phone and digital camera, he could wirelessly transmit photos to family and friends.

Although a few handsets and PDAs allow users to capture still photos, viewing is limited to that device. LightSurf has developed a solution for sharing those photos over wireless links.

To date, the company has announced a partnership with Motorola (www.motorola.com) to develop clip-on digital camera modules for wireless handsets. No deals, however, have been announced for the embedded digital eye module.

“There are a handful of leaders we're working with to take our embedded software for instant visual communications and embed that onto a chip,” Nijor explained. “So you would have a camera on a chip, powered by LightSurf technology, as well as a CMOS imaging sensor and a lens all packaged up into the size of a sugar cube.”

The digital eye technology has applications in the consumer market, as well as the business world.

To support wireless digital photography, carriers would need to integrate their wireless Internet gateways with LightSurf's wireless media gateway, Nijor said. LightSurf can remotely host and manage the back-end of the photo solution.

After the photo is captured, it is wirelessly transmitted over the network — whether it be circuit- or packet-switched — to LightSurf's back-end servers, which process, store and forward the image for sharing. LightSurf also can send the image to a Kodak (www.kodak.com) print lab to produce a hard copy for the user.

Nijor said LightSurf is studying the issue of wireless video in the context of peer-to-peer sharing of content, rather than the push model of streaming news, ads and sports that many companies are targeting.

With the bandwidth gap and the lack of affordable devices that can support wireless streaming-video applications, LightSurf has chosen to focus its energies on a more proven market, Nijor said.

“Our focus is on delivering on the promise of instant visual communication today,” he said. “And today there are over 60 billion photos taken every year. Photos have been around for well over 100 years, and they are going to be around for the next few hundred years. We don't see consumer behavior changing so dramatically that all of a sudden we want to see talking heads and don't want to share photos.”