MIMO—multiple-input multiple-output—promises dramatically improved throughput and range for wireless networks. The technology is compelling for enterprises and worth keeping a close eye on, but the gains that available MIMO-based products can provide arent enough to offset the current lack of an interoperable standard or concern over MIMOs interaction with legacy networks for corporate deployments.
In the strictest sense, MIMO allows for multiple receive-and-transmit antennas in the same device. However, what data is being transmitted by these antennas is up for interpretation.
MIMO-OFDM (MIMO-Orthogonal Frequency Division Multiplexing) technology will be a part of the IEEE 802.11n standard, which is expected to be ratified late next year. The TGn Sync proposal, which gained favor in this months downselect vote, provides for multiple antennas and spatial multiplexing technology, in addition to several other components.
Whats less clear at this time is what role channel width will play in the standard. It is likely that 40MHz channels will be used (as a required or an optional part of the standard) to boost capacity. One of the 802.11n working groups mandates is to maintain backward compatibility with 802.11 a/b/g networks, so n-enabled devices will need to detect legacy traffic and step back to 20MHz to avoid interoperability and interference problems. This could make it difficult to maintain optimal performance.
Enterprise-class wireless makers are forgoing MIMO in the near term, but companies can still accommodate greater bandwidth demands with denser and more intelligent wireless network deployments. As the standard ratification nears, however, and MIMO-aided performance escalates, there will be an increasing need to support the technology.
Indeed, all indications are that MIMO performance will ramp up quickly. Airgo Networks Inc. representatives have already hinted at chip-set revisions in development that provide as much as six times more bandwidth than were currently seeing, promising fantastic performance by the time the standard is ratified.
For now, vendors are releasing products based on proprietary MIMO implementations. eWEEK Labs has seen several consumer-oriented MIMO-based products come to market recently, and their different interpretations of what MIMO really entails can lead to some confusion.
Cisco Systems Inc.s Linksys Wireless-G Broadband Router with SRX (Model WRT54GX) and Belkin Corp.s Wireless Pre-N Router are based on True MIMO technology, Airgos term for MIMO, using multiple antennas and spatial multiplexing to boost performance.
Spatial multiplexing takes advantage of multipath reflection to simultaneously transmit different data streams on the same channel with separate spatial signatures. Receiving antennas collect the streams and recombine them quickly, effectively boosting the transmission rate.
Netgear Inc.s RangeMax Wireless Router leverages another MIMO variant, Video54s BeamFlex, to intelligently transmit multiple identical data streams from some combination of RangeMaxs seven antennas to boost range and eliminate dead spots in coverage. Called beamforming, this transmission approach does not use spatial multiplexing because each data stream is identical. However, it does improve performance over long distances by improving antennae gain, allowing wireless devices to maintain higher throughput modes at greater distance.
D-Link Corp.s Super-G with MIMO Wireless Router leverages an Atheros Communications Inc. chip set and four antennas. The wireless router also uses beamforming without spatial multiplexing. The Super-G feature uses two channels to boost throughput performance. To minimize impact on radio-frequency environments, the device must step back to one channel when other devices are present, minimizing the features impact in a crowded RF area.
Page Two
To compare MIMOs gains in throughput and range with standard 802.11g implementations, eWEEK Labs recently tested the Airgo MIMO implementation in the Linksys WRT54GX and compared the findings with numbers culled from a pair of popular 802.11g access points—the business-class Cisco Aironet 1200 and the consumer-grade Linksys WRT54G. Using Ixias IxChariot 5.4 in conjunction with Ixias 1600T hardware chassis, we measured each products performance at several distances.
Using the WRT54GX with Linksys MIMO-enabled CardBus client adapter (WPC54GX) provided excellent throughput performance at short and medium distances—almost twice the bandwidth we saw from the other access points. However, eliciting this performance required tweaking of default access point settings that could hamper backward compatibility with 802.11b clients. Without these tweaks, we could not squeeze more than 17M-bps throughput from the WRT54GX at close distances, and we could not attach to the network at all at the longest distance.
With tweaks in place and MIMO at the access point and client, we also saw improved performance at the longest distance, but not as great as seen at shorter lengths. At the maximum tested distance of 140 feet, WRT54GX performance dropped to just above 10M bps, about 3M bps greater than that of the other standard 802.11g devices. (It should be noted that, at this distance, our testing grounds included several walls and an elevator shaft between the access point and client.)
Large organizations will likely want to leverage embedded wireless client devices rather than install and manage new CardBus adapters, so we also performed tests against each access point using Intel Corp.s Pro/Wireless 2200BG adapter that came with our Dell Inc. Latitude 610 laptop. With this client device, the WRT54GXs antenna design added a small performance boost at all distances. These results, however, were significantly inferior to our end-to-end MIMO results.
During tests, we monitored the RF environment with AirMagnet Inc.s Laptop Analyzer 4.0 and the AirMagnet Trio wireless adapter. We performed each set of tests on Channel 6 of the 2.4GHz band, which in our labs environment showed the least amount of ambient noise and interference during the late-evening hours when we performed our tests.
Like the Aironet 1200 and the WRT54G, the Airgo chip set in the WRT54GX uses only a 20MHz band for transmission. We did not notice any appreciable differences in the levels of interference or noise due to the MIMO implementation compared with that from the other devices.
Page Three
This month, the IEEE 802.11n working group selected the TGn Sync proposal over the competing WWise proposal with a simple majority but could not immediately attain the 75 percent approval necessary to confirm the proposal as the first draft of the standard. Compromise between the backers of TGn Sync and the extant WWise proposal is now necessary. According to Airgo Networks representatives, some of the issues of contention are:
- Whether 40MHz or 20MHz channels should be mandatory: TGn Sync favors 40MHz channels for improved capacity, but this may take longer to implement and may conflict with regulations in Japan.
- The number of mandatory data rate modes.
- The quantity and usefulness of proposed extensions to the 802.11 MAC (media access control) sublayer.
If a compromise is not attained after the next few meetings of the working group, the WWise proposal will be reintroduced. The next working group meeting is scheduled for May 13-15 in Queensland, Australia. Stay tuned.
Source: eWEEK Labs Reporting
Technical Analyst Andrew Garcia can be reached at andrew_garcia@ziffdavis.com.
Check out eWEEK.coms for the latest news, reviews and analysis on mobile and wireless computing.