BROWSE BY TECHNOLOGY



RTC SUPPLEMENTS


EDITOR'S REPORT

Wireless Video Networking

Wireless HD Video Streaming: Struggling toward a Standard Solution

Wireless video streaming is nothing new and various solutions for streaming HD video have been out in niche markets for a while. But getting wireless HD into the mainstream of wireless networking—and that means the world of Wi-Fi—is yet to be but coming on strong.

TOM WILLIAMS, EDITOR-IN-CHIEF

  • Page 1 of 1
    Bookmark and Share

Article Media

The world of high definition video took its time getting firmly established in the consumer space in the form of HDTV, which has since become practically so ubiquitous that we lose sight of how long it actually took to get established in the market. Now that it is established, it has raised expectations about all of our interactions with video and graphics—from smartphones, to tablets to kiosks to industrial HMI displays—everything. The embedded space is no exception and processors have been turning out with ever more powerful on-chip graphics processing units to meet those expectations.

All the latest Atom processors from Intel have on-chip graphics engines. The Fusion G series from AMD uses general-purpose graphic processors (GPGUs) based on the graphics technology developed by ATI, which was later acquired by AMD. Graphics specialist Nvidia has integrated its GPU technology with multicore ARM processor architectures in its Tegra series, which is being widely adapted into mobile tablets. The list goes on.

At the same time, display technology seems to be racing forward challenging the processing capability to keep up. In the consumer realm, TV displays are getting bigger—up to 70 and 84 inches, and new display technologies are coming online such as organic LEDs (OLED), which are still climbing the cost curve but which promise to deliver more brilliant displays with much thinner panels. OLED displays can even be made flexible. Then, of course, there are the smart displays that support touch interfaces for touch, motion sensing and gesture recognition. Tablet users can be seen sitting at their laptops (unless they’re brand new) stroking at the screen and wondering why there is no response. The new Apple iPad, for instance, has a 9.7-inch display with a resolution of 2048 x 1539.

But speaking of expectations, everyone now expects wireless connectivity and they are going to expect it of HD displays along with everything else. The industry has been responding with a number of solutions, but it is going to take a few more years to arrive at a unified solution. What will it take to bring streaming wireless HD video into a mainstream technology for mobile wireless networks?

Not surprisingly, there are efforts underway. The question is what do they actually solve and which will ultimately be successful? There are a number of technologies that have homed in on the 60 GHz range for wireless video streaming and have attracted the attention of a number of large players. In addition, there are some other players that are still successfully holding shares of the market. One such player is the Wireless Home Digital Interface (WHDI), originated by the Israeli firm Amimon and supported to some extent by Hitachi, LG, Samsung and others. WHDI can stream uncompressed 1080p video over a wireless channel to any display device equipped with a receiver. It supports data rates up to about 3 Gbit/s in a 40 MHz channel. It is basically a point-to-point, not a networking technology.

Another variant is Wireless HD, which can be thought of as a wireless HDMI. It is based on a 7 GHz channel also in the 60 GHz range and allows for either compressed or uncompressed HD video. Developed by start-up SiBeam (which was bought by Silicon Image), it has attracted the interest of companies like Samsung, Sony and Toshiba. While the initial versions had a data rate of about 4 Gbit/s, the underlying technology is thought to be scalable to about 25 Gbit/s for possible additional resolution and color depth. One disadvantage, according to Brian O’Rourke of analysis firm In-Stat, is that its power consumption limits it to line-powered devices, so it is currently not a candidate for mobile applications. It does, however, include a beam-forming technology to avoid interference and to extend the range beyond about 10 meters.

One more technology that appears to have had the opportunity to be adapted and further developed, but which is still on the market on its own merits, is Ultra Wide Band (UWB). UWB came out of the effort to develop a wireless USB—an effort that was eventually abandoned. While there is no wireless USB, there are a number of UWB devices based on chips by companies like Alereon, which go into products such as wireless docking stations and video links that can stream at 720p from a dongle on a laptop to a receiver connected to a set via an HDMI cable.

The push for UWB, however, appears to have been subsumed by efforts of the Wi-Fi Alliance to develop a multi-gigabit wireless technology that ended up known as WiGig or 802.11ad. It also operates in the 60 GHz range and will eventually integrate into existing 2.4 GHz and 5 GHz Wi-Fi networks via new devices equipped with tri-band radios (Figure 1). Since it is based on 802.11 and has a way of integrating into existing networks, it has attracted the support of several semiconductor companies that are expected to announce silicon in the near future.

Figure 1
The WiGig architecture enables tri-band communications.

WiGig is, of course, not in and of itself a video streaming technology, but is the underlying network that can carry the video data. For that, additional software and hardware is required—such as having HD codecs on either end. Intel independently implemented a set of software on top of its silicon to allow video streaming from laptop to HDTV. It called the technology “My Wi-Fi.” Subsequently, it was renamed WiDi and built into a number of laptops so the basic video streaming could be done without another dongle sticking out of the computer.

Later, Intel handed the WiDi technology over to the Wi-Fi Alliance and it is now about to release a Wi-Fi Alliance specification that will be known as Wi-Fi Display. This will truly allow HD video streaming between 802.11ab networked devices that are connected using Wi-Fi Direct. If this seems a bit involved, it is actually a simplification. Where previous HD streaming technologies involved a single peer-to-peer wireless connection (often with cables connecting receivers to displays), Wireless Direct allows connecting an enabled device in a one-to-one, a one-to-many or a concurrent Wi-Fi AP and peer-to-peer connection without necessarily having to be at a Wi-Fi hot spot (Figure 2).

Figure 2
The three configuration modes made possible by Wi-Fi Certified Direct Connect.

Wi-Fi Display runs as an application on top of Wi-Fi Direct and can then stream video to or from other devices in the configuration equipped for WiGig plus Wi-Fi Display. It will run on 802.11ad solutions but not on the lower frequency connections. In other words, devices will need the 60 GHz radio and be able to support the 5 GHz channels for a 720p or 1080p video stream. Wi-Fi Display does appear to be the wave of the future for wireless HD video streaming due to the broad and broadening industry support for Wi-Fi in the first place, and increasingly for WiGig. Adding Wi-Fi Display will take more silicon, and WiGig is also today available at a premium price.

In-Stat’s Brian O’Rourke notes that although the WiGig 802.11ad silicon is just beginning to appear and at a significantly higher price than chips for the earlier technologies, it is expected to come down in price to approximate equal cost by about 2015.

The question is, will that price drop be driven by consumer acceptance or by the needs of industrial and commercial systems? Time will tell, but there is a good possibility that the perceived value of high definition, interactive wireless streaming video may have more immediate impact from the industrial side where devices and systems could gain immediate improved utility by incorporating it. Of course, given this blurring between what was once clearly consumer and what was clearly industrial, things might happen even quicker. Managers who want to interact with their factory data from their tablets might drive the acceptance in consumer devices faster than expected. One thing that seems pretty certain: now that it’s possible, it will happen.  

Advanced Micro Devices
Sunnyvale, CA.
(408) 749-4000
www.amd.com

Alereon
Austin, TX.
(512) 345-4200
www.alereon.com

In-Stat
Port Washington, NY.
(858) 652-4012
www.in-stat.com

Intel
Santa Clara, CA.
(408) 765-8080
www.intel.com

Amimon
Santa Clara, CA.
(650) 641-3191
www.amimon.com

Silicon Image
Sunnyvale, CA.
(408) 616-4000
www.siliconimage.com