Since the first short-range wireless networking technologies arrived in the late 1990s, one standard has climbed to the top. Wi-Fi, also known by the IEEE standard designation 802.11, maintains its domination over the world of computing and electronics in general.
Yes, Bluetooth has it beat in terms of total deployment and chips sold, thanks to its incorporation into a few jillion cell phones. But with over 120 million chip sets shipped in 2005 and over 90% of all laptops enabled with the technology, Wi-Fi has changed the way we network and use computers. Seemingly everyone's laptop is wirelessly linked to company networks, home networks, and public access points.
The cool thing about Wi-Fi is how it has developed over the years. Each new version of the standard boosts its operating range, data rate, reliability, and security. But that's not all. Wi-Fi continues to progress on numerous fronts, including work in some unexpected areas. In fact, there are five keys to its success.
1: CONTINUOUS STANDARD DEVELOPMENT
A quick history lesson of 802.11 reveals a progression of speed increases and other improvements over the past 10 years (Table 1). Many older, original 802.11b products are still in play. However, most wireless connectivity today is achieved through the latest standard, 802.11g, which provides up to 54 Mbits/s in the 2.4-GHz band.
Maximum range is about 100 m, if the path is unobstructed. Most connectivity is in an obstructed multipath environment, though, so the rate backs off automatically to a much lower number (about 20 Mbits/s) with a range far less than 100 m.
The 802.11n standard promises data rates in excess of 100 Mbits/s using multiple-input/multiple-output (MIMO) technology (see "How MIMO Works," ED Online 12998, at www.electronicdesign.com). The standard remains in draft form, but IEEE Task Group n is getting close. Companies already are selling "draft-n" chip sets and routers/gateways. These products mostly comply with the standard, but they may not be fully compliant in the end. Some companies say they can bring the product into full compliance later with a software upgrade, yet that remains to be seen.
These "draft-n" offerings are the result of the pent-up frustration with the standardization process. It takes time to reach a consensus, and in some cases, it doesn't even result in a final standard. The recent Ultra-Wideband (UWB) standard process didn't yield a final design, so the IEEE task group was abandoned. Most designers hope that won't happen with 11n, but progress has been rocky.
Back in 2005, the two primary factions for the 11n technology essentially agreed to disagree, and work on 11n slowed. Some companies like Airgo Networks decided to go ahead with their own proprietary versions and started selling MIMO chips. Most of these products are quite successful, providing user speeds in excess of 100 Mbits/s over a greater range than traditional Wi-Fi.
More recently, a group of manufacturers came together to form the Enhanced Wireless Consortium (EWC) to create a compromise version of the standard. The EWC submitted its draft to the 11n Task Group, which accepted it with an 87% vote. That draft is now working its way through the standards process, with a target date of September 2007 for ratification. With both hope and an agenda, the 11n group is confident of reaching a final version.
Meanwhile, companies like Atheros, Broadcom, and Marvell are pushing their draft-n designs. Atheros and Broadcom also recently demonstrated the interoperability of their chips as a way to kickstart the market for this standard. Airgo offers its 802.11a/b/g-compatible chips with MIMO but doesn't characterize them as draft-n. Intel, Texas Instruments, and some of the other chip companies remain rather silent on the subject.
While non-standard products can be successful, the draftnproducts won't meet the desirable goal of full interoperability among similar products. That may be okay for home networking or small business products. But widespread use in enterprise local-area networks (LANs), public hot spots, and laptop computers requires fully interoperable products.
The testing and certification process provided by the Wi-Fi Alliance now ensures that interoperability. You can't put the Wi-Fi label on your product until you pass the test, and the test is based on the standard. As soon as the standard is final, the Wi-Fi Alliance will begin its 11n program. Caveat emptor (let the buyer beware) as far as draft-n products are concerned.
Big standards efforts like 11n aren't the only reasons why Wi-Fi stays strong. Consider all of the smaller standards efforts that have been initiated to expand, improve, and fine-tune 802.11 (Table 2). Security is a good example of how the 802.11 standard continues to evolve. The earlier security part of 802.11b known as Wired Equivalent Privacy (WEP) worked well, but few users sought to activate it in their equipment. Also, WEP using the RC4 encryption standard could be defeated, though it takes lots of effort.
This led to vendor-developed security measures such as the Temporal Key Integrity Protocol (TKIP). The Wi-Fi Alliance adopted TKIP as Wireless Protected Access (WPA), which uses RC4 but encrypts every packet with its own key. Now, WPA has evolved into WPA2. WPA2, using NIST's Advanced Encryption Standard (AES) with a 128-bit key, has become the 802.11i standard.
2: MESH NETWORKS
In mesh networks, any wireless node can talk to any other nearby adjacent node to exchange information. This concept lets individual nodes extend their range by sending data to a remote node through intermediate nodes that act as repeaters or routers.
With such an arrangement, a few nodes can provide blanket coverage over a very wide area as signals hop from one node to another. The mesh scheme also makes the whole network more reliable by providing alternate paths from one node to another, even if one or more nodes are disabled.
The mesh scheme is inherent in short-range, low-power personal-area networks (PANs) such as ZigBee (IEEE 802.15.4). Proprietary mesh networks also can provide such coverage. This permits the implementation of wireless sensor networks and networks of wireless monitoring and control in large buildings and homes. Mesh networks are inherently self-forming and self-healing.
Wi-Fi mesh networking was first used in small community wireless networks in towns that lacked any kind of formal broadband service. Now, such mesh techniques have been adopted to implement municipal (muni) networks. Cities around the country are establishing these mesh networks for public safety (fire, police, EMS) as well as public works (street maintenance, water department) communications services. They also supply broadband access to local citizens. By using many access points (APs) with a mesh software overlay, any laptop, or other 802.11 transceiver, can communicate with the system over several square miles.
The main goal of most muni meshes is to improve city government services. They give city workers access to e-mail, the Internet, and any city databases or other resources from a laptop in a truck or from virtually any location. It saves time and, it's particularly helpful to police and fire services.
Another goal is to help cities bridge the "digital divide." Most large and midsize cities offer broadband services, but they're used primarily by their more affluent citizens. Whole areas of some cities are totally without broadband access, putting the citizens in these areas at a disadvantage. With a Wi-Fi mesh, any citizen can have access.
Finally, muni meshes are intended to boost the competitive nature of the local broadband markets. Cable TV and DSL companies initially fought the muni Wi-Fi movement. Now, some are joining in to get a piece of that action.
While muni mesh networks are a recent phenomenon, there's been considerable activity around the country. Some smaller cities are already in operation, such as Alexandria, Va., and Tempe, Ariz. Larger cities have systems under construction, such as Philadelphia, San Francisco, and Phoenix. Pasadena and Ripon, Calif., Buffalo, Minn., and Oklahoma City have systems in operation or on the way. Even a huge city like Houston, Texas, has a system in the request for proposal (RFP) stage.
Wi-Fi was chosen simply because it's low in cost and easy to expand. It also works in portable (not mobile) situations. The availability of mesh systems from companies like Cisco, Motorola, and Tropos makes installation fast and easy, typically as easy as bolting the AP to a light pole.
More elaborate systems like Motorola's Motomesh combine the regular 2.4-GHz unlicensed APs with special APs that operate in the relatively new 4.9-GHz public safety band (see the figure). This band is intended for use by police, fire, and other emergency services during disasters and major security events.
While most of the current crop of muni Wi-Fi meshes use proprietary technology, the IEEE is working on a mesh standard in its 802.11s task group. The standard isn't final, but it's expected to be ratified next year.
The jury continues to deliberate on how well these muni meshes perform. With enough APs, Wi-Fi can perform well even in environments with lots of buildings and other obstructions—even indoors, to some extent. But Wi-Fi definitely isn't a mobile technology. Companies like Cohda Wireless are developing complementary technologies that should significantly improve Wi-Fi's performance in a hostile mobile environment.
Because of the muni meshes, Wi-Fi may be a real competitor for the latest metropolitan broadband wireless system, WiMAX. WiMAX was designed to provide metro-wide wireless broadband access to cities using multiple cell sites similar to the cell-phone system. Such systems would compete with established cable TV and DSL broadband servcies.
Able to supply from 1.5- to 20-Mbit/s data services over a range of several miles, WiMAX is expected to be the answer for customers in rural areas that lack broadband service.
How it will compete with cable TV and DSL remains to be seen. Surely, there's an overlap with the muni Wi-Fi mesh phenomenon.
The new mobile WiMAX standard (802.16e) will outshine Wi-Fi when it comes to mobile access. But so will already available 3G cell-phone data services. Using GSM/EDGE, WCDMA, or cdma2000 with EV-DO available in many cities, a laptop can easily access the Internet, e-mail, and other broadband services using a plug-in card. Even the elusive 802.20 mobile wireless standard may get involved—if it ever gets off the ground.
3. HOT SPOTS EVERYWHERE
The Wi-Fi Alliance reports that over 100,000 hot spots are now available internationally. These are in addition to the regular access points in enterprise wireless LANs (WLANs) and home networks. And, the number of places where you can tap into a Wi-Fi network continues to grow. The muni Wi-Fi networks further extend that capability.
Airline Wi-Fi is now in the works. Laptop owners traveling by plane can access their e-mail and the Internet via an on-board AP. Boeing's Connexion system already is in effect on some Asian and European routes, and it soon will be available on longer flights in the U.S. Jet Blue and AirCell recently purchased the FCC's air-to-ground spectrum, with the potential for setting up future in-plane hot spots (see "FCC Awards Air-Ground Radiotelephone Licenses," ED Online 12828, at www.electronicdesign.com). Most travelers will welcome e-mail and Internet access on their laptops in a plane, but they may give a resounding "no" to Voice over Internet Protocol (VoIP) phone calls, which are becoming increasingly popular.
4: VOICE OVER WI-FI
With VoIP rolling out nationwide, it's no wonder that someone quickly thought of making phone calls using the technology attached to Wi-Fi. "VoWi-Fi" phones are already available for making calls from any AP or hot spot. Such phones could become standard for internal enterprise use. Several vendors already add Wi-Fi to standard cell phones, making it possible to call from a hot spot if cell site coverage isn't good—or vice versa.
The converged cellular/Wi-Fi phone has lead to the creation of systems such as unlicensed mobile access (UMA). Also known as a generic access network (GAN), UMA lets users seamlessly roam by regular cell towers or via some wireless technology in unlicensed spectrum, such as Bluetooth or Wi-Fi. Wi-Fi hot spots are the most likely links.
In GAN, if a subscriber's handset is within range of a hot spot, it automatically will connect with an IP connection through a gateway to a GAN server that in turn is linked to the regular core mobile network. This happens without subscribers knowing which system they're using. UMA/GAM is defined by the Third Generation Partnership Project (3GPP) TS 43.318 specification. So far, its implementation is limited.
A similar parallel effort, called IP Multimedia Subsystem (IMS), comes via the 3GPP as well as the 3GPP2 organization. Based on Internet Engineering Task Force (IETF) protocols, this new standard provides IP sessions over Wi-Fi and other packet-based systems like GSM/GPRS/EDGE and cdma2000 cellphone systems. This architecture uses standard VoIP voice technology with the popular session initiation protocol (SIP).
The goal of IMS is fixed-mobile convergence (FMC), which permits any cell phone to connect to any service via any available network (cell, Wi-Fi, WiMAX, whatever). It eventually may let carriers offer almost any service via any communications medium and charge appropriately. We shall see.
5: CONSUMER ELECTRONICS |APPLICATIONS
Wi-Fi is already well entrenched in the consumer space—it's by far the most popular home networking technology. Thousands of home gateways/routers are attached to cable TV or DSL lines for broadband access via multiple PCs and laptops.
Another movement, though, is the inclusion of Wi-Fi into dozens of other consumer items. For example, Nokia's digital camera interfaces flawlessly with any nearby AP to transfer photo files. Camcorders and personal media players (PMPs) also can connect wirelessly with a Wi-Fi transceiver. Wi-Fi-enabled MP3 players are already available.
What's the greatest coming application for Wi-Fi in the home? No doubt, it's wireless video technology. Consumer electronics companies are anxious to link PCs, HDTV sets, DVD players, personal video recorders, and other video devices together wirelessly. While the UWB companies claim this space for themselves, it isn't clear if they will totally dominate.
With the new 11n standard, Wi-Fi should be more than competitive, since 11n is expected to be far more robust at longer ranges than UWB. Wi-Fi (11n) can't top UWB's 480-Mbit/s speed. But its more than 100 Mbits/s is more than adequate to deal with video. And, its longer range and super reliability are sure to ensure it a piece of the consumer video action.
The big issue with Wi-Fi and video is quality of service (QoS). The 802.11 standard wasn't designed for continuous streaming of data like video or audio. It can result in disruptions of data that make for unacceptable TV viewing. This has led to the 802.11e QoS standard as well as the Wi-Fi Alliance's Wi-Fi MultiMedia (WMM) certification. Based on the 802.11e standard, this new certification ensures that certified devices deliver acceptable performance.
Companies also are developing technology that will help deliver video by Wi-Fi. Kiyon Autonomic Networks has developed software that runs on a wireless router or AP and converts into a multichannel system using TDMA instead of the usual CSMA/CA MAC technology. This software supports two HDTV channels and up to 50 VoIP calls over a fivehop mesh Wi-Fi network.
Another company, Metalink, recently demonstrated its WLANPlus chip set based on the 802.11n standard and MIMO, which can easily stream highdefinition video throughout the home. It's designed for set-top boxes, broadband gateways, HDTV sets, DVD players, and other consumer devices.
Metalink has an agreement with Royal Philips Electronics to use its chips in forthcoming HDTV products.
WI-FI PRODUCTS TO WATCH OUT FOR
The usual suspects—Airgo, Atheros, Broadcom, Intel, and Marvell—will field 11ncompatible chips when the standard is finally ratified. In the meantime, other Wi-Fi chip vendors are rolling out lowpower Wi-Fi chips for portable, batteryoperated consumer products and the cell-phone markets.
Conexant Systems' CX53121 is an 802.11b/g device targeted at any portable or mobile device wishing to incorporate Wi-Fi. STMicroelectronics also delivers similar chips, including its STLC4420, which covers 802.11a/b/g Wi-Fi. A second device, the STLC4550, has only 802.11b/g coverage. Both offer very low power consumption for cell phones, MP3 players, and other portable and consumer items.
In addition, Texas Instruments offers the Consumer Electronics WLAN Developer Kit (CE WLAN DK 2.0). It gives designers of portable and mobile devices everything they need to build a successful product. The kit includes the WLAN chip set comprising the TNET1351 and TNET 3526/5100. The 1351 is the single-chip MAC baseband processor and radio, while the 3526 is an RF power amplifier. The 5100 is a power-management chip. A reference design and software are provided to support both WPA2 security and the WMM/802.11e QoS efforts.