802.11 Planet Spring 2003
by John Latta
Wave Issue 0328 8/15/03
June 25-27, 2003
Boston, MA
At this, the third 802.11 Planet we have attended, there are major changes taking place. The number of exhibitors is up to nearly 100. There is considerable diversity of product on the show floor. In the past the demand in the market has been largely from consumers and road warriors but now the enterprise is getting serious. Yet, the hot spot market remains chaos with no clear business models. As Jupiter Research said on the first day: The Facilities Providers drive the market because they control access, be it in the hotel or airport. Further, it is clear that the current specifications need additional work to reflect the realities of the technology and market. For example, it was stated that a new IEEE working group under 802.11 was formed to support hand-off. In addition, high speed hand-offs are also being supported in the security working group of 802.11i. Other signs of market development are that $3.7b was spent in equipment last year. Intel has created within Intel Capital an investment fund of $150m for wireless.
Microsoft Keynote
Jawad Khaki, Corporate Vice President, Windows Networking and Communications Technologies gave the first keynote. This was significant in what was and was not said. Microsoft used the occasion to disclose a number of items and its view of the role of networking. Some impressive demonstrations were given, especially in the transport of 2 streams of HD over an 802.11a network. Given that this was seen as a corporate position statement the presentation merits some discussion.
Jawad described the Dream Network as
Easy-to-Use, Always Available, Secure, Scalable and Manageable.
As a network that is more than just a data conduit. It supports integrated data, voice, and video, which leads to new digital experiences.
Ubiquitous peer-to-peer connectivity.
It was especially important that Jawad stated that the
Take up rate for broadband does not follow the availability.
This was taken as an implicit recognition that lacking are applications that drive consumers to broadband. Along these lines, it was cited that internal data at Microsoft showed that its internal wireless network saved ½ to 1 ½ hours per day. Jawad joked that he hoped that this time was used as a gain for Microsoft and not to improve life.
One of the most important parts of the presentation was the Wireless Technology Roadmap, which was broken into Today, Next and Future. The next chart was a response to the question “What is Microsoft doing about this?” From this the following disclosures were made:
Microsoft is developing a Self-Organizing Wireless Broadband Mesh network;
Microsoft is developing a means to accomplish fast seamless wireless roaming across media;
Microsoft is developing smarter applications that adapt to latency, bandwidth and location;
Windows will enable secure hot spot sign-in, unified provisioning and discoverabilty;
There will be PAN support in Windows; and
Windows will support Wi-Fi and WAN roaming.
During his talk, Jawad Khaki stressed the need to confirm and support standards. Yet, the presentation was silent in this area, other than Wi-Fi. Curiously absent was any mention of:
802.16 & WiMAX;
802.20; and
SDR Forum
where Microsoft is not a player. In contrast, Intel is active in two, but apparently not 802.20. Yet, Microsoft has typically been late to the table in wireless standards efforts such as its membership in the Bluetooth SIG. Jawad is correct, Microsoft should play a strong role in standards.
Nomadix Keynote
Nomadix Chief Technology Officer and SVP, Joel Short gave the keynote on the second day. He used it to characterize the state of the industry in nomadic computing. Some of the points raised included:
The vision of nomadic user is one where there is ubiquitous secure Internet access from any place at any time.
The objective of transparent ubiquitous roaming is a convergence, of seamless interoperability and simplicity for the nomadic user and multiple wireless systems.
There are 50m mobile enterprise users today, 35m broadband ports by 2005 and an expected 500,000 hot spots by the same year.
The current ecosystem is made up of hot spots, clients and providers. The provider space is one of the most complex with:
Mobil carriers
Traditional ISPs
WISPs
Wholesale Providers and
Venue Owners.The public access space evolution, to the eventual goal of transparent ubiquitous roam, can be seen in terms of a pyramid. At the bottom is the first generation, which is based on link awareness and single point purchase of access, i.e., ad hoc, purchase of access. The second generation of the market comes with one bill roaming between WISPs. The market has just begun at the early stages of this. The last phase enables for seamless handoff between systems and mobile IP-like sessions.
Part of the 2nd generation are the hot spot finders that include HotSpotList.com; EZGoalHotspots, and ZoneFinder.org. Branding includes the war roaming symbols, Wayport, McDonalds, Intel Centrino, T Mobile HotSpot, Wi-Fi Zone and PassOne.
Currently the payment and authentication methods are all over the map. These include credit card, Virtual WISPs, Corporation, cell phone bill, hotel bill and home ISP. This latter category is one where the home broadband provider is also a hot spot provider. Currently Verizon is leading in the US with free hot spot access if the home broadband user is also a Verizon user.
In answer to the continual question – where is the business model in the morass, Joel states, there is not one but many.
Joel ended this presentation with the call for all to participate in industry alliances, to implement best practices and deployment guidelines, and advance the technology with:
Smarter clients;
More intelligent hot spots; and
Interoperable and converged networks that include 3G, WLAN wholesalers and ISP retailers.
Jupiter Research on the Home Market for Wireless
Joseph Laszio, Senior Analyst, Jupiter Research, spoke for a whole session on Extending Wi-Fi’s role in the Networked Home. Most simply correlated to what we have heard many times before. Some of the market statistics were interesting and these will be included in the presentation. Take aways include:
There is a high correlation between broadband to the home and the need for home networking. 38% have home networking, or a need for it, in those with broadband, and only 11% with dialup.
Only linear growth is forecast for broadband penetration from now to 2007;
The same asymmetries on the value of broadband to the home exist that have been seen many times before. Always-on and fast dominate the prioritization of importance of broadband in the range of 59% – 53%, while sharing between PCs falls to 25% and other applications such as listening to music drop to 20%.
The highest specific application identified was listening to PC music files on a stereo at 34% of home network users. Yet, 31% of those same uses said that they were not interested in any of the applications suggested – which included sharing and home security. This says much about the lack of applications.
Jupiter predicts that broadband households with PC networks will rise to 17.8m by 2005.
Jupiter went way out on a limb to claim that there would be two wireless networks in the home, one for TV/CE/Media use, and the other for PC/Utility/Communications.
In a recognition of the shift from retail, it was stated that the service providers will play a much larger role in distribution. Cited were SBC, BellSouth and Earthlink.
Proxim -- Intelligence Everwhere
Proxim argues that as wireless becomes pervasive in the enterprise that there should only be one wireless network. “Wire only once for wireless.” The wireless network will be for both data and voice, thus, dual mode phones will exist for both VoWLAN and cellular. This then makes seamless wireless coverage in the enterprise essential.
Proxim sees the future wireless LAN as being a very dynamic environment where the WLAN is actively managed in terms of load balancing, priorities, handoffs and AP linkups. Further, in these high demand environments the use of all available spectrum is mandatory – a + b + g + new spectrum. This is what they call intelligent mobility. The bottom line is that these WLAN networks will be intelligent, actively managed and that the handoffs should be intelligent and seamless.
A critical element in the requirement for this level of management, that does not exist in switched Ethernet, is that wireless is a shared medium whose precious commodity is spectrum. As the demand for bandwidth and mobility increase there is a corresponding requirement for intelligence to lessen the impacts of the resource scarcity.
Left unsaid is the role of the radios in this environment and it appears only a matter of time when the antennas and radios are managed elements in the network and will share the intelligence requirements.
Focus on Antennas
The IEEE has recently begun a group on high data rate wireless, 802.11n - High Throughput Study Group. During the panel entitled "Better WLANs through Better Antennas," Jack Winters of Motia stated that 216Mb/s would be possible with a 4 antenna MIMO configuration. As was learned at this panel discussion, the antenna has become the focal point for improved performance in 802.11. Zeewaves is proposing that the antenna contain the network security with a chip that resides there. ethertronics, the antenna design company that did the reference design for the Intel centrino implementation, stated that this design for notebooks was one of the most difficult they have done, especially given that the target price is 50¢.
Motia
In the antennas panel, Jack Winters described Motia's work on Smart Antennas. There are two ways in which these antennas can be implemented. One is a switched Multibeam Antenna, and the other is an Adaptive Antenna Array. Both track the wireless environment to maximize the effective gain and decrease the impact of interfering signals. In multipath environments, there is a diversity gain of these antennas, which raises the reliability of the network. The smart antennas can be applied to the access point as a low cost implementation, but the larger gain comes when a MIMO configuration is created on both the AP and clients.
It is Motia’s position that two antennas can be added to either or an AP and client and still be standards compliant, for a slight cost increase. When added to the AP it provides greater range and higher capacity. When added to the client it provides greater range and interference immunity. When both are used, i.e., it is a MIMO configuration, the gains are additive. However, MIMO is not a part of the 802.11 standard and currently not specifically included in the charter for 802.11n--but may be implied within the channel definition. It is Motia’s position that they will only provide antennas to their OEM partners which are standards compliant. The smart antennas for today’s 802.11 compliant systems will go be available for sale from Motia in 6 – 12 months.
DFS Defined
With the WRC approval today of the additional 250MHz of spectrum from 5.470 – 5.725GHz and 5.250 – 5.350Hz it is expected that the FCC will follow its already released NPRM to release this spectrum for WiFi usage. There is one problem – since this spectrum is already in use by the military it will likely be mandated for use only on a non-interference basis. Thus, radios will have to first determine if the band is in use before transmitting using a technique called dynamic frequency selection (DFS).
Here are more details on DFS as announced by NTIA:
Recent technological developments made successful sharing possible between the existing allocated services and WAS in a large portion of the 5250-5825 MHz band. Spectrum sharing between the radars and WAS at 5 GHz can be accomplished by employing a new breakthrough technology known as dynamic frequency selection (DFS). DFS uses the same principle as listen-beforetransmit communications systems, but operates automatically and has a much faster response time. The NTIA was instrumental in developing and validating the DFS technique as a method to allow sharing. This was accomplished through analysis employing rigorous simulation over a two-month period. Without proper specific characteristics for DFS, however, sharing will still not be feasible.
The following values are required for DFS to allow successful spectrum sharing at 5 GHz:
- A DFS detection threshold of -64 dBm for WAS devices operating at a total effective isotropic radiated power (e.i.r.p.) of 200 milliwatts to 1 watt, and -62 dBm for devices operating at an e.i.r.p. of less than 200 milliwatts, measured over a period not to exceed 1 microsecond, as normalized to a 0 dBi gain antenna. These measurements must be accomplished during quiet periods between or within each WAS frame or packet;
- A channel non-occupancy period of 30 minutes to ensure that fixed radars will be protected for any channel in which the DFS detection threshold has been exceeded;
- A channel availability check time of 60 seconds upon initial startup or monitoring of WAS on any particular channel to ensure all radars present around a WAS are detected prior to it utilizing a channel; and
- A DFS Channel Move Time of no more than 10 seconds. DFS Channel Move Time is the period that WAS systems will have to move off of the channel once the DFS detection threshold has been exceeded. It takes an average of 200 milliseconds for all normal traffic to suspend and then intermittent control signals can continue for up to 10 seconds.
WAVE Comments
802.11 Planet was another excellent venue to step back and see the big picture. This is our 3rd Planet event and we can now see important long term trends. Many of the sessions and speakers developed their vision.
Wi-Fi and wireless networking is getting complicated, in the enterprise, and in the home. We have believed that there is likely to be a divergence between the WLAN requirements for the home and the enterprise, with the home WLAN being much easier. The Proxim presentation left us wondering. Consider the following:
As the number of home network nodes rises it is likely that the number of access points will rise.
These nodes, based on what is done with them, could well tax the shared access wireless network and, thus, mandate a greater level of management;
Frequency interference, bandwidth needs and congestion are likely to drive usage to 5GHz;
The CE companies are solidly behind 802.11a not 802.11b, yet, in these data-centric conferences this is largely ignored; and
Mobility of wireless devices in the home will further put pressure on the intelligent wireless networks Proxim describes.
Thus, we are left wondering is the home all that simple?