The quest for robust wireless high-def video connections
Streaming multimedia information without wires is at best marginally feasible with today's Wi-Fi. Evolutionary and revolutionary successor technologies strive to improve the situation, but do consumers even care?
Brian Dipert, Senior Technical Editor -- EDN, September 23, 2010
At A Glance
|
| View as PDF |
The market for merged audio-and-video
transport over physical wiring remains
to some degree fragmented. The dueling
DisplayPort and HDMI (high-definition-multimedia-
interface) camps, for example, continue
to make their cases to their system-design
customers and end users (Reference 1). And both
legacy and upstart alternative-cabling approaches
remain relevant to a debatable degree (see sidebar
“HDBaseT strives for ascendancy”). The technology
treadmill rolls on, however. An increasing
amount of industry attention now focuses on cutting
the cord by employing wireless transport from
source to destination. The approach has notable
merits, including enabling the portability of both
the multimedia source and the destination locations
and dispensing with unsightly and cumbersome
wires. The display’s power cable must remain
for now, however (see sidebar “Wireless power:
the hype of the hour”). The wireless approach also
potentially extends the source-to-destination span
beyond wires’ attenuation-defined limitations, including,
in some cases, routing signals not only
intraroom but also inter-room, again without the
need for unaesthetic, difficult-to-install, and costly
wall-spanning cable-routing topologies.This article does not focus on so-called smart media adapters that search for and pull file-based information from computers and NAS (network-attached-storage) devices (Reference 2). Such products undoubtedly have their place, but the substantial required processing intelligence negatively affects their cost, and they must support a vast number of file-system and network ports and protocols to create a robust design, representing an implementation and maintenance nightmare. This write-up instead focuses on implementation scenarios in which the source device not only provides temporary or permanent housing of the content but also “broadcasts” it over the network for one or multiple comparatively “dumb” playback destination devices to tune in. The promoters of IEEE 802.11n Wi-Fi have long positioned it as the Holy Grail of multimedia transport, even in high-definition-video scenarios. And it can act in that role, but only in certain situations. More generally, as testing last year made abundantly clear, conventional one- and two-stream 802.11n configurations cannot reliably handle large video payloads, therefore rendering the current technology inappropriate for widespread consumer advocacy and adoption (Reference 3).
Technologists are tackling the current generation’s problem, however, initially in a more or less proprietary fashion, but with inevitable standards-based interoperability to follow (see sidebar “Making the best of a technology that’s hard-pressed”). They’re also hard at work on the next-generation standards documentation and subsequent implementations, which will make even meatier improvements. And at least one manufacturer has transformed the 5-GHz 802.11n variant into a video-tailored point-to-point transport approach, which is incompatible with the prevailing industry standards, however. Other developers feel that a more substantial frequency migration away from the 2.4- and 5-GHz ISM (industrial/scientific/medical) bands is necessary to finally transform wireless networking’s promise into reality. UWB (ultrawideband) advocates include Wi-Media Forum participants, along with at least one proprietary approach. The WirelessHD Consortium has focused on another unlicensed spectrum swath, 60 GHz, and a suite of semiconductor heavy hitters, the WiGig (Wireless Gigabit) Alliance, recently also directed its attention at this high-frequency spectral region.
Which of these contenders for the throne will eventually seize the crown is as yet unknown and will likely remain unclear for some time. Equally unclear, however, is how big the wireless-video-market prize will ever be and how long it will take to get to that size. This article includes the observations of a knowledgeable wireless-video-industry participant who requested anonymity. This person undoubtedly has at least a slight bias as to the preferred outcome, but, then again, doesn’t everyone with some skin in the game? Nevertheless, I hope you’ll still find the insider’s comments informative.
Evolving augmentation
I initially struggled last year with streaming wireless video partly because of the limitations of the gear I was using. The number of discrete streams that a piece of 802.11n equipment can handle depends on both its antenna-array configuration and its radio implementation. As the relevant Wikipedia entry concisely states, the number of antennas in use on both sides of the link limits the number of simultaneous data streams (Reference 4). “However, the individual radios often further limit the number of spatial streams that may carry unique data,” says the Wikipedia entry. The A×B:C notation helps identify what a radio can do. The first letter, A, represents the maximum number of transmitting antennas or RF chains that the radio can use. The second letter, B, represents the maximum number of receiver antennas or RF chains that the radio can use. The third letter, C, represents the maximum number of data spatial streams that the radio can use. For example, a radio that can transmit on two antennas and receive on three but can send or receive only two data streams would be 2×3:2. The 802.11n draft allows configurations as large as 4×4:4. Common configurations of 11n devices are 2×2:2, 2×3:2, and 3×3:2. All three configurations have the same maximum throughputs and features and differ only in the amount of diversity the antenna systems provide. A fourth configuration, 3×3:3, is also becoming common, according to Wikipedia. It has a higher throughput due to the additional data stream.
I think that the authors of the Wikipedia
entry are, however, too enthusiastic
about the extent of near-term 3×3:3
adoption. Most 802.11n-cognizant gear
now on the market is capable of two-stream—mainstream and high-end—performance or one-stream—entry-level
and portable—performance. However,
three-stream equipment, employing
silicon chip sets from such companies
as Atheros, Intel, and Marvell, is beginning
to appear. Such multistream
capabilities are desirable when, for example,
a computer is simultaneously receiving
information over the Internet
and transmitting data to a LAN (local-area-network) client peer. Apple built
triple-stream capabilities into the latest
iterations of its Airport Express router
and Time Capsule router-plus-NAS
product, although the company didn’t
explicitly promote these enhancements
(Figure 1 and Reference 5). And triple-stream support is also reportedly
at the heart of Intel’s WiDi (Wireless
Display) initiative, which the company
rolled out at the January 2010 CES
(Consumer Electronics Show) in Las
Vegas (Figure 2).
At January’s announcement, WiDi
supported only a short list of relatively
robust Intel CPUs, reflecting the
fact that the video is lossy-encoded
in the PC before transmission; core-logic
chip sets; and Wi-Fi radios (Table
1). On the other end of the wireless
chain is a receiver, such as Netgear’s
PTV1000 Push2TV adapter, with
HDMI and component-video outputs
to connect to a tethered display. Inside
the PTV1000 is a Sigma Designs
media processor, along with a singlestream
Ralink 802.11n transceiver. The
audio and video codecs WiDi uses are
unknown, but the results at crowded-spectrum
CES were impressive, with
adequate frame rates and few to no visible
image artifacts (Reference 6).The approximately 2-second latency from the WiDi transmitter to the receiver is problematic only if, for example, the source is outputting the soundtrack while the destination is displaying images. Audio and video coming from the common wireless-link endpoint will preserve lip sync. Current-generation WiDi supports the 720p image resolution and dynamically upscales all content to that resolution before transmission; 1080p resolution capability is on the WiDi road map. Intel also plans HDCP (high-bandwidth digital-content-protection) support, which will enable DVD (digital-videodisc) and Blu-ray-disc playback (Reference 7). At CES, only display mirroring was possible. Intel subsequently added the ability to instead horizontally expand the desktop onto the remote display so that, for example, a PowerPoint presentation can run in full-screen mode at a destination projector while a source laptop computer displays a speaker’s notes.
Also at CES, Netgear demonstrated
the WNHD3004, a prototype 4×4
MIMO (multiple-input/multiple-output)
802.11n wireless video-bridge-device
pair that the company based on
Quantenna’s transceiver technology,
and a fuller-featured follow-on to the
WNHDE111 that I tested in last year’s
hands-on project. The WNHD3004
also comes in a WNHDB3004 version
in a two-adapter bundle. The product
was scheduled to enter the retail sales
channel this month. It signifies a notable
business expansion for Quantenna,
which to date has focused on carrier-grade
and enterprise-infrastructure deployments.
The IEEE and Wi-Fi Alliance
have historically devoted most of
their not-unlimited attention to one- and
two-stream chip sets and systems employing them, but the organizations’
standards-development and interoperability-confirmation focus are now broadening
to three- and four-stream products.
For example, Quantenna reported
in June that its QHS600 802.11n wireless-access-point chip set had received
Wi-Fi Alliance certification, including
WMM (Wi-Fi multimedia) and WPA2
(Wi-Fi Protected Access 2) enhancements
to the base designation.The IEEE 802.11ac committee is looking at support for more than two streams as a key capability. The committee’s target for backward-compatible enhancement is more than 1-Gbps peak PHY (physical-layer) wireless speeds (Reference 8). However, currently available 802.11n streams can each theoretically support only 150-Mbps maximum bandwidth even in their optional 40-MHz, wide-channel mode, so even a four-stream configuration cannot alone achieve that formidable goal. As such, the 802.11ac group is also considering increasing the per-stream channel width to 80 or even 160 MHz. The 5-GHz band therefore represents 802.11ac’s primary focus. Additional bandwidth improvement of an estimated 10% may come from more efficient modulation algorithms, and the committee is also considering multiuser MIMO antennas and algorithms, which claim to enable one channel to simultaneously broadcast streams to different destinations. The current working date for 802.11ac ratification is December 2011.
Transport adaptation
Another fundamental streaming-multimedia issue is the traditional router-centric, star-networking-topology model. The audio and video stream coming from the source must first go to the router before continuing to the destination; large-payload multimedia material requires dedicated spectra for each path to accomplish the desired glitch-free playback. To address this concern, the Wi-Fi Alliance is now testing and certifying Wi-Fi Direct, a peer-to-peer communication scheme employing the IEEE’s 802.11s standard, and a successor to the poorly implemented 802.11 ad hoc mode.
More generally, 802.11n, like its b, a, and g predecessors, targets use as a generic networking protocol, albeit with increasing degrees of enhancement for multimedia and other latency-critical applications in each wireless generation. As such, Amimon has dispensed with some 802.11n features in developing its multimedia-optimized WHDI (wireless-home-digital-interface) technology. For one thing, WHDI has from the beginning implemented direct source-to-destination interaction without an intermediary router or switch. For another, WHDI is 5-GHz only, trading off broadcast range versus 2.4 GHz for a spectral environment with decreased interference. The company claims, however, that the technology spans 100 feet even through walls, with less-than-1-msec latency. Each 720p or 1080i video stream relies on one 18-MHz channel; 1080p streams each use two channels.
Amimon has not disclosed other implementation details of WHDI’s protocols, along with their variations from 5-GHz-based 802.11a and n. Multiple documents on the company’s Web site, along with multiple postings to EDN’s How We See CE blog, describe WHDI as lossless. However, these claims also include qualifiers, such as “In video, different bits have different level of importance, and the effect of an error greatly depends on which bit was corrupted.” For example, the online technical summary says, a stream of 8- or 10-bit numbers, each representing the primary-color value of a given pixel, represents a typical uncompressed stream. The MSB (most-significant bit) of each of these numbers has greater visual importance than the LSB (least-significant bit). If an error occurs on the MSB, that pixel gets a different and unwanted value. However, an error in the LSB results in a minor change in the pixel’s value. According to Amimon’s documentation, WHDI breaks down the uncompressed HD (high-definition) video stream into elements of importance and then maps the various elements onto the wireless channel in a way that gives those with more visual importance a greater share of the channel resources. In contrast, WHDI allocates fewer channel resources to elements that have less visual importance and therefore transmits them less robustly (Reference 9).
Does WHDI discard undetectable low-order bits only after transmission from source to destination, or does it also as needed discard those bits at the source? If it discards them at the source, is it accurate for Amimon to label WHDI lossless? “How Amimon achieves what it claims seems to be out of sync with physics fundamentals,” says the earlier-mentioned knowledgeable—albeit anonymous—wireless-video-industry insider. “Obviously, the company has demonstrable technology, but reading the explanation of how it works raises more questions than answers. Amimon claims support for uncompressed high-definition video to 1080p.” He notes that 1920×1080=2,073,600 pixels per frame; 60 frames/sec translates into 2,073,600×60 frames/sec=124,416,000 pixels/sec. Assuming 24-bit per-pixel color results in a data rate of 2976 Mbps, or 2.976 Gbps.
“Amimon claims it can support these data rates because WHDI uses joint-source coding,” my source says. Joint-source coding is a form of unequal error protection giving higher FEC (forward-error-correction) protection to the MSB and less FEC to the LSB—a technique that the JPEG (Joint Picture Experts Group)-2000 codec first employed. “Any FEC, no matter how it is applied, requires incremental bandwidth,” says the source. “The data rate is only going to go up—not down—when you apply FEC.” Amimon claims a 40-MHz spectrum occupancy, he says, noting that a quick calculation shows that, to remain lossless, WHDI must send data at a data density exceeding 75 bits/Hz, even before applying any FEC. “Such bit density would require a QAM [quadrature-amplitude modulation] on the order of more than a trillion and a dynamic range off the charts. There is something more going on here that physics—specifically, communication and information theory—cannot explain. The raw numbers speak for themselves,” he adds.
My industry source notes that, at
the 2008 CES, an engineer from Amimon
at Gefen’s booth, which was then
demonstrating an FPGA-based WHDI
prototype, said that WHDI doesn’t use
compression but instead uses compaction,
a process involving the removal
of some of the LSB color information
before transmission and its re-creation
on the receiver end of the wireless
link. He recently ran the HDMI output
of a Blu-ray player into a two-port
HDMI splitter. The output of one HDMI
splitter port ran directly to a flat-panel
display. The source then connected
the output from the second HDMI-splitter
port to an Amimon transmitter
and broadcast the video over a WHDI
link, hooked up the receiver to an identical
flat-panel display, and positioned
the two screens side by side. He found
that the video quality for the WHDI-based
system, although viewable, had
a washed-out color skew. When you
looked at it alone, it was not too noticeable;
however, once it was next to the
original content on the other display,
the alteration was obvious. “It appears
that WHDI is manipulating the color-space
conversion by dropping some of
the pixels’ LSBs and maybe even sending
some pixels as monochrome interspersed
with color pixels that change
from frame to frame,” he says.Note that the degree of calibration
between the two displays used in the
test setup is unknown. Pixel-bit discarding
and distortions between the WHDI
transmitter and the receiver versus the
wired HDMI-only alternative are therefore
not the only possible reasons for
the perceived disparity between them.
Potential gaps between marketing hype
and reality aside, the market success of
Amimon’s technology is notable. At
the 2010 CES, for example, the company
announced that LG Electronics and
other customers had adopted WHDI
(Figure 3a). By May, the company
claimed that it had surpassed 500,000
units’ worth of chip-set sales and orders.
And Amimon in June unveiled preliminary
details of WHDI Version 2.0, with a specification to follow next year. This
version will include 3-D-video transport,
support for the 4000×2000-pixel
ultra-HD format, Wi-Fi integration and
channel coexistence, and mobile-friendly
reductions in power consumption and
silicon footprint (Reference 10).
Ultrawide relocation
Although officials at Amimon feel that the company can accomplish its objectives within the 5-GHz ISM band, other manufacturers believe that other frequencies will better suit multimedia’s needs. As such, the WiMedia Alliance has harnessed a UWB approach. UWB occupies a swath of spectrum spanning 3.1 to 10.6 GHz, which depends to some degree on regional regulatory policies. Its policy aspires to be friendly to other concurrent frequency-spectrum inhabitants, although the additive background broadband noise that additional UWB transmitters create may ultimately interfere with traditional narrowband and carrier-wave systems. Its backers also regularly tout peak transfer rates of 480 Mbps at distances of as much as 3m, or approximately 10 feet, and 110 Mbps at up to 10m, more than 30 feet.
People often incorrectly use the terms “WiMedia” and “Wireless USB” interchangeably. “WiMedia defined a standardized UWB-radio technology that is protocol-independent,” says Mike Krell, Alereon’s senior director of communications and business development. “Wireless USB implements the USB standard on this radio. It is also possible to run any other protocol on that same radio—proprietary, for example, or TCP/IP [Transmission Control Protocol/Internet Protocol] or Bluetooth.” WiMedia was once the planned foundation for high-speed Bluetooth, running at frequencies higher than 6 GHz in this case to avoid European-spectrum regulatory issues. However, the initial strategy to move WiMedia development to the Bluetooth SIG (special-interest group) and then to wind down the WiMedia Alliance hasn’t happened as planned. Bluetooth’s WiMedia aspirations are unclear, as is Bluetooth’s broader vision for high speed, as the organization focuses most of its attention in the low-power realm.
Many observers have over the years
repeatedly tested WiMedia gear and
have regularly concluded that real-life
speeds are fractions of the marketing
claims. The long-standing standardization
squabble between promoters of implementation
alternatives, which never
reached resolution to anyone’s satisfaction
and led to the shuttering of several
start-ups, didn’t help the market to
embrace the UWB technology, either.
As such, today’s dominant UWB applications are in Bluetooth- and other RF- and infrared-competing, low-bit-rate wireless-USB usage scenarios that are largely insensitive to speed, such as computer keyboards, mice, low-resolution webcams, and still-image-camera transfer setups. Nonetheless, WiMedia technology backers remain undeterred; several multimedia streaming setups exist, based on chip sets from companies such as Alereon, Realtek, and Wisair (Figure 2c).
WiMedia derives from one of the two PHY contenders for the IEEE 802.153a higher-speed variant of the specification. It leverages MB-OFDM (multiband-orthogonal-frequency-division-multiplexing) technology and either QPSK (quadrature phase-shift keying) or QAM-16. An alternative approach from Pulse-Link, CWave, operates over wired coaxial cable and wireless connections. It instead harnesses BPSK (binary phase-shift keying)- and QPSK-modulation techniques and is based on the other historical IEEE 802.15.3a contender, DS (direct-sequence)-UWB. Its backers claim that it has a longer broadcast range for a given bit rate and a less costly implementation potential than WiMedia. Nonetheless, they admit that neither CWave nor any of its competitors have yet hit the price points necessary for broad market adoption. This situation may make you wonder, after years’ worth of various companies’ and groups’ public advocacy of the wireless-video concept, why early adopters aren’t creating the demand necessary to drive down costs.
Uncompressed video transmission is a desirable attribute for a number of reasons. It reduces the cost of the system implementation because it requires neither compression horsepower at the transmitter nor a decompression engine at the receiver. A compression-free approach can also minimize the overall latency of the transmission system. And video content is likely already lossy-compressed when it gets to the consumer through a codec from Microsoft, MPEG (Motion Pictures Experts Group), On2 (now Google), Sorenson, or another developer (references 11 and 12). Additional lossy compression that occurs before display results in incremental image degradation, especially egregious if the artifacts don’t synergize with those in the source material.
As such, SiBeam has decided that a more radical spectral relocation, to the 60-GHz, millimeter-wave unlicensed band, is necessary. The company’s WirelessHD technology employs a 7-GHz-wide channel, currently delivering a 4-Gbps data rate. The company claims, however, that bit rates as high as 25 Gbps are possible. WirelessHD supports DTCP (digital-transmission-content-protection) encryption for content-access control. Although line-of-sight transmitter-to-receiver linkage is normally necessary at this frequency threshold, WirelessHD uses beam-forming-MIMO-antenna techniques to create alternative signal paths that, for example, reflect off room walls. Still, WirelessHD remains an intraroom approach. Oxygen molecules create atmospheric absorption and, therefore, attenuation limits spans to 10m, or approximately 30 feet.
WirelessHD uses the 60-GHz IEEE 802.15.3c PAN (personal-area-network) specification, according to the anonymous industry insider. Current-generation WirelessHD uses 1.76-GHz-bandwidth, OFDM, QPSK, and QAM-16. Its maximum RF-power level is slightly less than 10W. Due to RF’s directivity at 60 GHz, WirelessHD requires a steerable antenna array. SiBeam has demonstrated an array of 6×6, or 36, antenna elements, translating to 36 transmitting chains and 36 receiving chains. Thus, 36 low-noise amplifiers must couple to 36 VGAs (variable-gain amplifiers). The anonymous source considers 36 elements to be overkill, saying that a 4×4, 16-element array would work almost as well with less than half the complexity. For the RF transmitter, OFDM requires two DACs at 4G samples/sec with 6 bits of resolution. One DAC generates the I (in-phase) component of the signal, and the other focuses on the Q (quadrature) component. The 4G-sample/sec requirement for the DACs achieves a Nyquist sampling rate for 1.76-GHz bandwidth, including minor oversampling. Each RF-receiver/antenna chain, comprising an antenna, a low-noise amp, and a VGA, requires independent processing until it reaches an analog correlator with 36 inputs—one for each receiver chain, the source says. He adds that the correlator coherently sums the energies, parsing time to picosecond-accuracy levels, almost at the level of an atomic clock. The output of the correlator feeds two 4G-sample/sec, 6-bit-resolution ADCs for the I and Q components.
“Next, let’s look at digital baseband processing, starting with the transmitter side,” says the source. Each DAC, he explains, requires 24 Gbps of baseband-sourced data. Two DACs translate into 48 Gbps of digital data to drive them. This requirement is not only performance-intensive but also, even at 65- or 45-nm-process technologies, highly power consumptive. The broadcast destination, receiving QAM-16-encoded OFDM, requires recovery of its I and Q components, translating to 48G samples/sec of total data. The two ADCs generate this data and subsequently feed it into a digital baseband subsystem. “Think of a 60-GHz RF-transmitter front end with a 10W output,” the source says. He then asks how many watts of dc power you must put into a CMOS RF-power amplifier to yield 10W of power at the antenna. Combining the power consumption of the ADCs, DACs, baseband circuitry, and MAC (media-access controller) yields a 60-GHz system that could easily consume more than 30W. “People don’t particularly like the idea of being exposed to a mobile phone’s approximately 500-mW RF output,” he adds. “What will they think when they find out that the entire time they’re … watching a movie, they’re being exposed to almost 10W of RF energy at 60 GHz? How much energy does a small microwave oven use in comparison: 100W at 2.4 GHz?”
In some ways, WirelessHD revisits WiMedia, albeit with a steerable antenna array, says the source. “I sing the praises of OFDM for narrowband applications, such as Wi-Fi, Homeplug, and MOCA (multimedia over coaxial cable). OFDM works great for these applications because the effective RF bandwidths in use are tens of megahertz. As a result, the required ADCs and DACs can be more than 10 bits because of the lower required sampling rates.” Each ADC or DAC bit is roughly equivalent to 6 dB of dynamic range; hence, 10 bits equals 60 dB. In contrast, both WiMedia and WirelessHD operate over hundreds of megahertz of bandwidth, limiting the ENOB (effective number of bits). Thus, their ADCs and DACs can operate at no more than 6 bits, or 36 dB of dynamic range. The use of QAM, which requires a greater-than-20-dB SNR (signal-to-noise ratio) to reliably recover the signal at the receiver, leaves little link margin for propagating the signal, translating to a fragile link. This issue hurts WiMedia’s range and performance due to the technology’s limited transmitting power; it is also the reason that WirelessHD needs 10W of RF power to compensate for both low dynamic range and high attenuation at 60 GHz.
60-GHz standardization
Observer skepticism aside, SiBeam strives onward. At this year’s CES, the company announced a second-generation chip set, which is reportedly now in production. The SB9220 network processor and SB9210 RF transmitter target use in multimedia sources, and the SB9221 network processor and SB9211 RF receiver target use in displays and other destination devices. At CES 2010, SiBeam also announced partnerships with Vizio, the latest in a list of notable OEM adopters, and retailer Best Buy, which made an equity investment (Figure 3b). In May, the company unveiled the WirelessHD Version 1.1 specification. Reminiscent of Amimon’s WHDI Version 2 specification, WirelessHD Version 1.1 ups the data rate to 10 to 28 Gbps, making the technology capable of handling so-called 4K (4096×3072)-pixel-resolution, 3-D, and other large-payload video streams. It also broadens encryption beyond DTCP to include HDCP Version 2. Networking support encompasses portable-device synchronization and IP (Internet Protocol) encapsulation, and a power-consumption reduction is equally amenable to mobile-electronics applications.
In May, SiBeam also announced dual-
mode support for its 60-GHz rival,
the WiGig Alliance. The support begins
with the SB8110 RF transceiver
and its associated SK8100 development
kit, which are available now. WiGig first
publicly unveiled its plans in May 2009;
a press release touting the completion of
its Version 1 specification followed in
December (Figure 4). Its current board
of directors includes Atheros Communications,
Broadcom, Cisco Systems, Dell, Intel, Marvell International, MediaTek,
Microsoft, NEC, Nokia, Nvidia,
Panasonic, Samsung Electronics, Toshiba,
and Wilocity. The WiGig contributor
list is similarly flush with notable silicon,
software, and systems developers.Single-chip-set compatibility with both 2.4- and 5-GHz 802.11 and with 60-GHz networks has been a WiGig Alliance objective from the group’s founding. The alliance formalized this intention in May, when it and the Wi-Fi Alliance announced a cooperative arrangement to share technology specifications, with the goal of creating a next-generation certification program that also supported networking in the 60-GHz frequency band. The organizations intend for 60-GHz-cognizant devices to automatically down-shift to the 2.4- or 5-GHz band beyond WiGig’s ultrahigh-frequency-broadcast reach, which the alliance hopes to extend beyond WirelessHD’s 10m through advanced adaptive-beam-forming and other techniques.
WiGig Alliance literature also clearly documents variable-bandwidth performance that depends on a device’s target power consumption. Some WiGig Version 1-based systems offer peak data-transmission rates as high as 7 Gbps, including EDAC (error-detection-and-correction) overhead. This rate leads to the claim that WiGig is more than 10 times faster than four-stream, 600-Mbps 802.11n. However, all devices, including battery-operated devices, that meet Wi-Gig specifications can achieve 1-Gbps peak data-transfer rates. This bandwidth discrepancy is partially due to differences in the leveraged modulation and coding schemes. According to the WiGig Alliance Web site, OFDM supports communication over longer distances with greater delay spreads, providing more flexibility in handling obstacles and reflected signals. OFDM allows transmission speeds as high as 7 Gbps. Conversely, single-carrier encoding typically results in lower power consumption, so it is often a better fit for small, low-power, handheld devices. Single-carrier technology supports transmission speeds as high as 4.6 Gbps. (Reference 13).
This situation is analogous to that of today’s 802.11n products, in which cellular handsets and other small-form-factor mobile-electronics gear might incorporate only a single-stream Wi-Fi transceiver rather than beefier radios and the associated antenna arrays in ac-powered, larger products. According to Wi-Gig Alliance literature, modulation and coding schemes share elements, such as preamble and channel coding, simplifying implementation for manufacturers of WiGig devices. It’s unclear from published documentation whether WiGig will extend the 802.11 MAC to 60 GHz or craft a dual-MAC approach that leverages 802.15.3 or another approach at 60 GHz. At least some WiGig participants will accomplish the merged-technology objective through partnerships. Wilocity, for example, announced in July that it was working with Wi-Fi veteran Atheros. Further extending the relationship between WiGig and the IEEE, the alliance also touts that its technology is the foundation of the 802.11ad specification for very-high-throughput 60-GHz networking.
You can reach Senior Technical Editor Brian Dipert at 1-916-548-1225, brian.dipert@cancom.com, and www.bdipert.com.
Talkback
-
So......I still have to live with all my cables hanging on the walls every time that I move my HDTV???? Or do I still have other options that I don't know about????
Celso de Souza - 2010-19-10 10:24:42 PDT -
Brian:
I bristle at your assertion that wires are unsightly. I consider wires and antennas things of beauty. Of course, my wife does not, so guess who wins that battle:-)
It seems to me that we have missed the forest for the trees. Why on earth are we sending uncompressed data around the house? Seems a pretty stupid concept to me.
H.264 (the de facto standard for all HD content)is how *every* consumer receives his HD content. When will consumer electronic folks wise up and put a H.264 decoder in every device and keep the bandwidth down until the data reaches its ultimate destination? The answer, of course, is that H.264 decoding requires paying for IP and adding additional silicon to the display devices. It would certainly be a boon for the likes of TI, ADI, and Trident, though.
Ray Mack - 2010-4-10 11:30:29 PDT -
Wireless high-definition video by almost any means will certainly be quite an accomplishment if it can be made to work well most of the time. Unfortunately it is a waste of spectrum and hardware, something that we really don't need. OF course if you make it some will buy it, but that certainly does not mean that it is a good idea. Quadraphonic sound was a similar fad, a while back.
Consider the probability of interference, if it is in a band used for other functions as well, and consider the lack of privacy, unless expensive encryption is used. By the way, high speed encryption systems add cost and use power.
So a responsible engineering and marketing community will instead figure out how to make the wires less visible, and possibly how to send the signal over a single fiber optic strand. except that it seems like that technology may already be available.
William Ketel - 2010-23-9 18:10:43 PDT
