In January of this year, the IEEE 802.11n task group announced that it had approved a draft proposal for the 802.11n standard. This indicated that disagreements between members of the task group that had been stalling development of the standard had been largely resolved and marked a large step forward in 802.11n’s development. But what does this mean to the consumer and wireless network designer? This month, we’ll give you an overview of what to expect from 802.11n.
As with other iterations of 802.11 technology, 802.11n’s major difference from its predecessors is higher data rates. Whereas 802.11b maxed out at 11 mbps and 802.11g maxed out at 54 Mbps, 802.11n uses data rates of 200 to 600 Mbps, with effective data rates of about half that. This means that 802.11n will be able to offer data rates that compete with a typical wired Ethernet network. The same technology that allows 802.11n to offer higher data rates also promises to give it greater range than 802.11g or 802.11a under similar conditions.
For very small networks consisting of only a few APs, all you’ll need to know is contained in the previous paragraph: 802.11n is a faster, longer range version of 802.11. Home and SOHO users will probably simply swap out their 802.11g access points and client cards for 802.11n ones when the budget allows it. But larger networks will need to consider how to integrate 802.11n into their existing wireless infrastructure. They probably won’t have the budget to wholly upgrade their existing 802.11g and 802.11a devices to 802.11n.
Fortunately, 802.11n is backwards compatible with all previous 802.11 standards. This means that you can add 802.11n access points to your network, and they will work seamlessly with non-802.11n clients. 802.11n clients will similarly work with non-802.11n access points. Of course, when 802.11n devices talk to non-802.11n devices, the maximum data rate will be that of the non-802.11n device. Additionally, just as the effective throughput of 802.11g devices is reduced when there are 802.11b devices in the area, 802.11n devices will only achieve maximum throughput when only 802.11n devices are present.
Because 802.11n is backwards compatible, one migration strategy is to transition high-demand access points and clients to 802.11n first. For example, a hotspot that provides Internet access via a 1.5 Mbps DSL link wouldn’t be able to take advantage of 802.11n’s higher speeds. For this type of installation, even an 802.11b access point might suffice! On the other hand, if users are accessing resources on 100/1000 Mbps Ethernet corporate LAN, they can use all of the wireless bandwidth that they can get. As 802.11n equipment becomes cheaper and more common, lower-demand clients can be replaced, until eventually, 802.11n might become the de facto standard wireless technology to buy, much as 802.11g is today.
Whereas 802.11g only uses frequencies in the 2.4 GHz ISM band and 802.11a only uses frequencies in the 5 GHz UNII bands, 802.11n is specified for both 2.4 GHz and 5 GHz. Preliminary announcements from vendors suggest that 802.11n cards will be available either in single-mode (2.4 GHz only) or multi-mode (2.4 GHz and 5 GHz) models, just as we can today buy either an 802.11g or an 802.11a/g card. Multi-mode cards will likely be more expensive than their single-mode counterparts, but will be able to take advantage of more channels, giving the network designer more options for avoiding interference. The issue here is similar to the choice between 802.11g and 802.11a that network designers make today.
Potential users of 802.11n should keep in mind a few caveats. First, 802.11n is capable of delivering usable throughput in excess of 100 Mbps, but, as with all wireless technologies, this is dependent on environmental conditions. As distance and obstructions between the client and the AP increase, the usable throughput will fall. That being said, 802.11n should offer more throughput than 802.11g or 802.11a in pretty much every situation. Second, consumers should continue to buy 802.11g and 802.11a equipment until the final 802.11n standard is ratified, which is expected to happen in late 2006 or early 2007. 802.11g and 802.11a equipment bought today will work with 802.11n equipment sold tomorrow, but there’s no guarantee that pre-standard 802.11n equipment bought today (before the standard is finalized) will work with tomorrow’s standard’s-compliant 802.11n equipment.
Connect802 will continue to bring you more analysis of 802.11n as it becomes available.
Like 802.11 standards before it, 802.11n brings a lot of new technology to the table. 802.11b introduced high-rate Direct Sequence Spread Spectrum (DSSS); 802.11g brought us Orthogonal Frequency Domain Multiplexing (OFDM); and 802.11n brings us Multiple In Multiple Out (MIMO). This month, we’ll discuss MIMO and other more advanced technologies that make up 802.11n.
The major difference between 802.11n and prior 802.11 standards is that 802.11n requires the simultaneous use of more than one antenna—MIMO. Although 802.11b radios commonly had two “diversity” antennas, only one of these antennas was ever used at a time. Normally, when two antennas transmit at the same time on the same frequency, the transmissions corrupt each other. MIMO takes advantage of the fact that the characteristics of a radio signal are affected by the relative location of the transmitting and receiving antenna. By aligning the transmitting and receiving antennas relative to each other in a specific way, the receiving radio can distinguish the transmissions of each of the transmitter’s antennas. Of course extra antennas alone are not sufficient to enable MIMO—it also requires some signal processing magic in the radio.
MIMO allows for several streams of data to be sent simultaneously. Even if this were the only change between 802.11g and 802.11n, it’s easy to see that the data rate would be increased. If MIMO allowed for 2 streams of data, data rate would be doubled; 3 streams of data would triple the data rate, and so on. Each additional stream of data requires additional components—analog-to-digital converters, modulators, and so forth—in the radio, increasing its size and expense. And not all 802.11n users require the maximum data rate that 802.11n is capable of producing. Therefore, 802.11n gives manufacturers the option to choose how many transmit and receive radios will be in a single chipset. This is commonly expressed in the format “TxR”, where “T” indicates the number of transmit radios and “R” indicates the number of receive radios. So, for example, a 2x2 chipset would have two transmit and two receive radios. A 3x3 chipset would have three of each and would offer approximately 1/3 more throughput than a 2x2 radio. 802.11n allows for configurations of 2x2 up to 4x4. Common configurations include 2x2, 2x3, 3x3, and 4x4. Of these, 2x3 is the only one that has an unequal number of transmit and receive radios. Having more receive radios than transmit radios increases the ability of the chipset to pick up transmissions without the additional cost of an additional set of radio circuitry.
802.11n systems can be either “open loop” or “closed loop”. A “closed loop” system provides a mechanism for the receiver to give the transmitter feedback about the signal-to-noise ratio of each of the transmission paths between the two devices. For example, in a 3x3 system, three independent transmission paths exist. Perhaps one of the paths happens to be experiencing interference or multipath fading. In a closed loop system, the receiver could notify the transmitter of this and the transmitter would have the option of reducing the data rate or increasing the transmit power of that path in order to increase its reliability. By comparison, in an open loop system, the transmitter uses the same data rate and transmit power on all paths and the receiver has no means of modifying the transmitter’s behavior.
Closed loop systems can be more efficient than open loop systems, but they require a certain amount of overhead for the two stations to give each other feedback on the quality of each transmission path. If the systems are stationary, then the characteristics of the transmission paths between them shouldn’t change much, and the overhead of giving feedback on those characteristics will be minimal compared to the throughput gained by the more efficient use of transmit power. On the other hand, if the stations are mobile or the environment is changing a lot, the characteristics of the transmission paths will also change a lot, and the stations will have to constantly communicate about those changes, increasing overhead and decreasing the effective throughput of the link. Therefore, vendors that plan to make stationary devices like media extenders tend to prefer closed loop designs, while vendors that plan to make mobile devices like PDAs tend to prefer open loop designs. Currently, the 802.11n standard provides the option to go either way.
Several new converged voice/Wi-Fi handsets were announced recently. The Linksys WIP300 VoIP Handset uses any 802.11b/g network, supports SIP 2 for VoIP, and includes a POP email program (allowing email to be read on the phone’s screen). The phone retails for $250. Meanwhile, in the cellular camp, Philips ahs announced that sometimes next year, a U.S. cell phone carrier will be using its Nexperia 6120 chipset to allow voice-over-Wi-Fi calls from your home network. The advantage of cellular integration is that Wi-Fi calls from the cell phone can be routed onto the cellular or wired telephone network. The disadvantage is that cell carriers are certainly going to charge for the calls, whereas calls from a pure-802.11 handset like the Linksys are free. Finally, the Nokia 6131 was announced, with the unusual ability to seamlessly roam between GSM and 802.11.
In-building Wi-Fi is quickly becoming more than just a means of browsing the Internet or accessing the corporate LAN. Wi-Fi networks are becoming a utility, like electricity, rather than a luxury. As this occurs, users will expect similar service levels and uptime from their wireless LAN. Expert wireless engineers with the skills to deliver on users’ expectations will be in high demand.
255 MHz of New 5 GHz Channels Available For 802.11a Use
Effective January 20, the FCC added a 255 MHz band of frequencies to the unlicensed 5 GHz band. This means that 802.11a devices have 11 more frequencies available for use—up from the 12 that were previously available. It remains to be seen when products supporting these additional channels will come to market or whether existing 802.11a products can be modified to support the new channels via a firmware upgrade.
802.11a provides a compelling alternative to 802.11g because the 5 GHz frequencies used by 802.11a are usually much less congested than the 2.4 GHz frequencies used by 802.11g. The major issue preventing wider use of 802.11a is the lack of 802.11a client cards. As such, 802.11a is primarily used in backhaul or point-to-point links, where client access is not an issue.
Strix Adds Voice Over WLAN to Bangladesh
A network of 90 Strix Systems OWS mesh routers will be installed over an eight square mile area to support 10,000 voice lines in Cittagong, Bangladesh. The network is planned to ultimately offer 200,000 voice lines over three years, in addition to video and data services. The success of this system will demonstrate the feasibility of using wireless mesh systems to supply voice and data connectivity to metro-scale areas. Connect802 is proud to be a Strix reseller.
At Connect802 we're your PAGE ONE resource for wireless networking!
Connect802 has the experience, expertise, and resources to help you with your wireless network system. Use your favorite search engine and see what Connect802 is doing. Each month we give you some suggested search terms for you to explore. Here's this month's list. As you look down the search engine results you'll find Connect802 either at the top, or on the first page (true for Google and Excite, unknown for the rest).