Home > Cover Story> Full Story

Wireless LAN - it's happening at last
Uday Birje

P R O F I L E

Uday Birje is Country Manager (India & SAARC), Enterasys Networks. He has been with Cabletron for the past three years and and has been the key figure behind successfully transforming Cabletron Systems’ Indian operations to Enterasys Networks after internationally, Cabletron Systems was transformed into four independent operating entities.

Frost & Sullivan's latest market research documents says that the Wireless LAN (WLAN) market place will be worth $5 billion by 2005. Currently, wireless access points (base stations) retail at prices between $200 and $1,400. Over 100 vendors sell WLAN technology, and you can purchase a NIC (Network Interface Card) for less than $100. With so many facilitating factors, the $5 billion figure and consequently a huge installed base of WLANs will certainly not be far from reality.

It's not difficult to understand what is fuelling this growth. WLANs are a convenient plug and play technology, easily deployable without the underlying costs associated with cabling infrastructure. In times when organizations are implementing technologies that enable mobile and ubiquitous access throughout the enterprise, WLANs provide the perfect fit. The fact that employees can access company resources without the constraints of a wired network connection, provides companies with more productive and efficient workforces.

Confusion?

Developing standards are at best a long and detailed process, and they are supposed to create a clear direction for the industry and customers to follow. Lately however, in the field of wireless technologies, the standards themselves have become quite confusing to an outside observer. The world was a much simpler place when we only had IEEE 802.11b wireless products to worry about.

So what is all the confusion about? Anyone who thought there would be a nice and smooth transition from 802.11b to something like 802.11a, will be quite surprised. Originally the plan was (for most vendors and customers) to move from an 802.11b technology to 802.11a technology. The new technology would provide higher bandwidth (from 11Mbps to 54Mbps), and at the same time allow wireless data transmission to move to a less-crowded area of the radio spectrum (from the crowded 2.4 GHz band to the less crowded 5 GHz band). If this had happened smoothly, the wireless world would not be as confusing as it is now.

Data security
A few things came up over the last couple of years, which introduced a certain level of uncertainty into the plan. We discovered that WEP (Wired Equivalent Privacy) was not as secure as everyone thought. This was not an earth shattering revelation to many people in the industry, but it did catch some people off guard. The reality is that WEP was designed only to provide the same expectation of privacy as a wired network. The fact that WEP succeeded even at this level is debatable.

A more valid concern may be to look for the number of organizations that are utilizing any sort of security measures. As part of the Wi-Fi (Wireless Fidelity) standard, wireless networks can advertise (broadcast) their network names to make them easy to find and join. This advertisement is called the SSID (Service Set Identifier). The first step in providing some form of security would be to refrain from broadcasting this name.

It is smart to choose a name that cannot easily be guessed. Many organizations that set up wireless networks do not turn on any sort of security, and they allow the name of their network to be broadcasted to anyone who cares to listen for it. To address this within the standards, the IEEE 802.11i sub-committee was formed. Their goal is to provide a standard interoperable way to secure wireless data. In the meantime, there are a number of proprietary vendor solutions that aim to solve the same problem. While many of these solutions are better than nothing, they will not be as good as the standards-based solution.

Standards soup
The next bit of uncertainty centers on what the "next" technology should be. At last count there were ten different options for wireless data transmission, other than 802.11b. Why so many? Good question. It seems that every vendor and consortium has a different idea to solve the same basic problem, that is how to transport data via radio transmission, in a way that is both secure and efficient, at a high data rate. If there was only one answer the world would be a simpler place. Unfortunately (or fortunately depending on your point of view), there are many valid ways to solve this problem. Let's look at a few of these.

The first place to look is at the technology that is closest to 802.11b. Most people would think that this is 802.11a, but that's not the case. A very contentious standard, 802.11g, is closer in technology to 802.11b. IEEE 802.11g is based on doing a higher data rate (starting at 22Mbps) in the same frequency spectrum (2.4 GHz) as 802.11b.

There are a few incompatible ways for a vendor to provide an 802.11g solution. This means that you could potentially (some day in the future, because none are shipping right now) buy a solution to provide 802.11g from two different vendors. Both sets of equipment could conform to the 802.11g standard, but the two sets of equipment would not necessarily work together at any of the data rates past 11Mbps. At 11Mbps the solutions should be compatible with existing 802.11b Wi-Fi compliant solutions. This point was one of the few things that members of the standards body agreed to. Someday there will probably be a single market winner in the 802.11g race, but today it is far from certain which technology that will be.

The next set of technologies that should be examined are the ones that are currently being promoted and are considered fringe technologies at this point. They are fringe technologies from the point-of-view of serious consideration for the enterprise data network, because they lack many of the features that an enterprise data network requires. These technologies include the likes of HyperLAN 1 and 2, Bluetooth, Ultra-Wideband, Wide Band Frequency Hopping, and HomeRF. While none of these technologies have any real presence within the mainstream, with the possible exception of Bluetooth, they are all trying to gain wireless market share. All these technologies have some technical merit, and may be a good solution in some cases. The problem is that since there are so many of them, they add to the confusion in the market. It is unlikely that any of these will ever gain significant market share, although they will probably persist for a while.

Upgradeable 802.11b
So what should someone who is looking to deploy wireless technology today do? In times of risk and uncertainty, there is generally a "flight to safety." For this uncertain period of time in wireless development the safe bet is the good old standard and interoperable 802.11b technology. The best solution is one that will allow an organization to purchase an 802.11b solution today and easily upgrade it to any future wireless technology winners. The winning technologies may be based around 802.11a, 802.11g or some other specification. The upgradeable solution should be able to support them. By 'upgradeable' I mean a solution where you can change a small component, without ripping out all the old technology to replace it entirely with new technology.

Wireless and India
Now let's turn our attention to what's happening in in the wireless space locally and understand how the near future looks. WLAN is talked about very sporadically even though there was a lot of noise on Bluetooth. As one may be aware, Bluetooth has been talked about more in the telecom sense. Actually, WLAN provides a comprehensive solution which caters to the requirements of both telecom and IT. In India, both a buyer and a seller of WLAN need to have a WPC (Wireless Planning & Coordination Wing) license. Enterprise customers are finding this a barrier to implementation. One hopes that as pressure builds up due to market demand, there will be relaxation in the licensing provision. This will give an impetus to the wireless market.

On a closing note, remember the days when we went from good old 10Mbps Ethernet to something faster? For a while, there was great confusion in the market--100VGAnylan, Full Duplex Switched Ethernet, 25 Mbps ATM, Fast Token Ring, FDDI, OC-3 ATM, and Fast Ethernet--they were all proposed as replacements for Ethernet. All of these technologies were standards. Eventually the market settled on one, Fast Ethernet, but not before a lot of incorrect technology decisions were made, and a lot of gear had to be discarded. Let's try to learn from history and let's not repeat the same mistakes.