[1st-mile-nm] 802.11-VHT: Gigabit Wireless Future?

[Richard Lowenberg]

Gigabit Wi-Fi Could Follow 802.11n

http://news.yahoo.com/s/zd/20080902/tc_zd/231522

Mark Hachman - ExtremeTechTue Sept. 2.

Although the IEEE 802.11n specification is still a year or more away from
release, a small group of engineers is already moving ahead to the next
generation of wireless networking. The goal? Gigabit Wi-Fi, to match the wired
gigabit Ethernet links of today's PCs.

A working group is preparing to propose what may eventually be known as IEEE
802.11 VHT (Very High Throughput), in what some call the successor to 802.11n.
Two working groups each are suggesting proposals to push throughput in excess
of one gigabit per second, roughly ten times that of the 802.11n specification.

At this point, the IEEE has yet to formally approve what's known as a PAR, or a
Project Approval Request, the first step on the road to an IEEE standard.
However, that approval is expected. The proposed technology has also not yet
been blessed by the Wi-Fi Alliance, which governs the technology.

On the other hand, much of the group's work has been in cooperation with the
Alliance, including plans to use the technology in wireless display
technologies for HDTV, fast file transfer, and campus deployments, among
others.

If approved, the increase in data rates would be dramatic, at least by today's
standards. The 802.11n standard calls for bandwidth on the order of 600
Mbits/s; today, so-called "pre-n" devices offer roughly 300 Mbits/s. But actual
throughput can be much less, or only about 100 Mbits/s, after overhead and other
traffic. The 802.11 VHT proposals call for throughput of at least a gigabit per
second, which could place actual data rates many times higher. Interestingly,
none of the PAR documents mention the estimated range for the wireless link.

When could such a standard be ratified, and IEEE 802.11 VHT products hit the
market? Possibly around 2011 or 2012, according to James Gilb, the technical
editor for the WirelessHD consortium and the maintainer of the 802.11 VHT PAR
page.

IEEE specifications typically take about four to five years from proposal to
product; the confusion surrounding the 802.11n standard will probably push
formal 802.11n products out until 2010, Gilb said. In July, the IEEE P802.11n
task group pushed the timeline for the official publication of 802.11n out by
six months, to November 2009.

Gigabit Wi-Fi has been proposed before, although not in specification form. In
2007, a research group from the Georgia Electronic Design Center (GEDC) at
Georgia Tech developed a 15-Gbit link at 1 meter.

The details

Two IEEE 802.11 VHT PARs have been submitted, according to documents the group
has posted: one dealing with sub- 6 GHz communications, and one using
frequencies of 60 GHz and above. While a few key details have been released,
more updates are expected in the second quarter of 2009, Bruce Kraemer, the
chairman of the IEEE Working Group and a senior manager of strategic marketing
for Marvell Semiconductor, said in an email.

The author of both the sub-6-GHz and 60-GHz PAR is Eldad Perahia, a senior
wireless systems engineer at Intel, who did not reply to requests for comment.

The 802.11 VHT sub-6-GHz spec has two overarching goals. "It would be fair to
say a goal is to find ways to either serve more users per unit area or to
increase the bits/second delivered to users in a given area and backward
compatibility is the norm," Kraemer said in an email, when asked about the goal
behind 802.11 VHT.

"There are no technical agreements on the project plan since it has been
concentrating on a basic objective statement for the PAR," Kraemer added.

At this point, it's difficult to say which of the two PARs might have more of an
effect on PC users and consumer-electronics aficionados. According to Gilb, the
sub-6-GHz PAR looks more likely to replace traditional Wi-Fi, if only because
the proposal looks more like a traditional wireless network. The goal is for
the sub-6-GHz spec to be backwards-compatible with traditional Wi-Fi, but to
avoid using and interfering with the 2.4-GHz frequency band used by today's
802.11b and 802.11g.

The sub-6-GHz PAR proposes a single-link throughput of only 500 Mbits/s, but
contains an interesting provision: "a maximum multi-STA throughput (measured at
the MAC data service access point), of at least 1 Gbps". Put simply, it means
that the proposal suggests that bandwidth be aggregated among STAs (stations),
which include the Wi-Fi client cards and chips found within PCs.

How? A July 2007 proposal by Motorola may have the answer. It suggested using
Orthogonal Frequency-Division Multiple Access (OFDMA) bandwidth allocation,
which would assign a user "chunks" of bandwidth measured in time and frequency.
OFDMA is an advanced form of CDMA, a channel access method employed by
CDMA-based cellular phones, among other devices.

According to the document, 100 MHz of bandwidth would represent 320 subcarriers,
with 8 adjacent subcarriers grouped into one chunk. The forty chunks would then
be divided by the number of users. Using a technique called optimized chunk
allocation, throughput increased as the number of users also increased,
Motorola found: 33 percent for two users, and 66 percent for six users. The
performance gains trailed off sharply after that.

The authors of the Motorola report suggested that the 802.11 VHT PAR "adopt a
requirement on the aggregated throughput and not the peak throughput to
introduce multi-user component into the standard" ? which, apparently, the
sub-6-GHz PAR has done: "The project may include the capability to handle
multiple simultaneous communications," according to the PAR. "The 1Gbps maximum
multi-STA [client] throughput may be achieved when considering multiple
simultaneously actively-communicating STAs, e.g., a BSS [base station] with 1
AP [access point] and at least 3 STAs."

Modifications to both the physical access layer (PHY) and media access control
(MAC) would be required, the PAR says. However, the working group also hopes to
design in power-saving technologies above and beyond what is offered in Wi-Fi.

The 60-GHz PAR

The 60-GHz proposal, however, aims for pure, blazing speed. Like the sub-6-GHz
PAR, the 60-GHz flavor of 802.11 VHT also proposes altering the MAC and PHY. In
this case, however, each single link would be capable of a gigabit per second
throughput, primarily through the use of a wider swath of spectrum, or 57 to 66
GHz.

If the sub-6-GHz proposal seeks to avoid interference with traditional Wi-Fi,
the 60-GHz version is hoping to design around the wireless multimedia
specifications that operate in the 60 GHz range. The problem here is twofold:
first, there is an existing specification that offers similar capabilities:
IEEE 802.15.3, designed for "personal area networks" running at speeds up to 55
Mbits/s or so. Second, specifications like the WirelessHD group also operate in
the 60-GHz band.

According to the 60-GHz working group, the proposal will maintain the "802.11
user experience" ? basically, a familiar universe of base stations, access
points, and clients, together with the underlying infrastructure. The group
anticipates that future Wi-Fi radio will access the familiar 2.4-GHz/5-GHz
Wi-Fi networks, but also 60-GHz networks as well. And if a 60-GHz network isn't
available, the radio would search out a lower-bandwidth Wi-Fi connection.

Co-existence with other 60-GHz systems is so important that it is specifically
called out in the PAR as a requirement.

What appears to be happening, however, is that 802.15.3 and the new 802.11 VHT
technology "will go out as separate standards, and we'll let the market
decide," Glib said. While the physical chips are relatively similar, the fact
that both use the same frequency band means that they'll both share the same
airwaves. The question will be whether the two standards can truly cooperate,
or whether manufacturers will have to choose one over the other.


---------------------------------
Richard Lowenberg
1st-Mile Institute
P.O. Box 8001, Santa Fe, NM 87504
505-989-9110;   505-603-5200 cell
rl at 1st-mile.com  www.1st-mile.com


----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.