Yesterday Cisco announced the Aironet 1140 series of Access Points (APs) and M-drive technology. The Aironet 1140 is Cisco’s next-generation 802.11n access point and is a significant advancement over its predecessor, the Aironet 1250. The most significant difference is the fact that the Aironet 1140 has a maximum power draw that is within Power over Ethernet (PoE) limits. The Aironet 1140 is also thinner and lighter than the Aironet 1250, and uses internal antennas (see the table below). In addition, the Aironet 1140 uses the same mechanical brackets as the Aironet 1250 in order to simplify AP replacement.
Cisco claims that they are the first vendor to have an 802.11n AP with a maximum power draw that is within PoE limits, but they are actually not the first vendor to make this claim. To be fair, not all PoE claims are created equal. Some vendors must reduce transmit power or turn off encryption in order to fulfill their PoE claim. So to be specific, the Aironet 1140 can use a single PoE connection to power both 802.11n radios, at full transmit power, running two spatial streams, with encryption turned on, using either the 2.4 GHz or 5 GHz frequency bands. (When your wireless vendor tells you that a single PoE connection can power their 802.11n access point you should ask them to be more specific.)
M-Drive
M-Drive Technology is not a single technology, per se, but rather an integrated collection of capabilities that aim to simplify the adoption of 802.11n. M-Drive is primarily a branding announcement because many of the capabilities already exist in the Aironet 1250. M-Drive provides Cisco with a name that they can use for competitive marketing, much like Aruba Networks did with Adaptive Radio Management (ARM). M-Drive includes capabilities such as:
- Automatic client load balancing
- Dynamic channel and power setting
- Extensive client compatibility testing
- Beamforming technology (that Cisco calls ClientLink)
Cisco’s competitors already support many of the M-Drive capabilities. For example, Aruba’s ARM 2.0 software supports features such as automatic client load balancing, dynamic channel/power setting, and airtime fairness. Aruba’s access points do not incorporate transmit beam forming but instead implement Cyclic Delay Diversity (CDD) to improve the downlink transmission quality for each client.
I anticipate that as this market matures, competitive differentiation will become more RF oriented and more technical. So we can expect to see Cisco enhance and market M-Drive aggressively.
ClientLink
Cisco’s ClientLink technology is interesting because it works in both the uplink (Client to AP) and downlink (AP to Client) directions to improve transmission quality for legacy (802.11b/g/a) clients. Most 802.11n access points can improve uplink transmission using an algorithm called maximal ratio combining (MRC) that combines several received signals from a client to create an single optimal signal. ClientLink remembers the MRC signal adjustments for each client, and then uses that information to form an optimized transmit beam for every client, on a packet-by-packet basis.
ClientLink can improve throughput, coverage, and reduce the lost packet rate for legacy clients. This is important because legacy clients consume an inordinate amount of airtime due to their slower transmission rate, and therefore they can significantly reduce the remaining channel capacity for 802.11n clients. (Note that ClientLink is supported on both the Aironet 1140 and 1250 and is available April 2009 via the Cisco Unified Wireless Network release 6.0).
The Bottom Line
The Aironet 1140 is a significant advancement over the Aironet 1250 and will be widely deployed in carpeted enterprise environments. The collection of capabilities called M-Drive brings Cisco into competitive alignment with similar capabilities from other vendors. Lastly, enterprises should take advantage of the ClientLink capability because it measurably improves throughput, coverage, and the lost packet rate for legacy clients.
Overall, this announcement strengthens Cisco’s competitive position in the enterprise WLAN market. It may also expand the size of the WLAN market for all vendors by helping enterprises become more comfortable with 802.11n and therefore more likely to deploy the technology.
Nice Analysis Paul, I agree with your assesment; this is a significant announcement from Cisco as it brings them more closely in line with the competition from an AP hardware perspective. Now they just need a viable solution with external antennas.
I think the interesting part of this announcement is that Cisco tries to make this a technology announcement and the only real technology is ClientLink, and even that is a bit dubious. With 2 transmit chains beamforming can only give 3db gain, this doesn’t really make up for the fact that all of Cisco’s competitors have 3 transmit chains and CDD which can provide more than 3db gain over Cisco.
As you correctly point out, it is going to be more and more difficult for the IT buyer to differentiate solutions as there is more and more innovation on the RF side. In the end I believe that this will lead to enterprises using benchmarking to compare solutions, much the way companies do when evaluating other IT solutions.
Cisco tried to sell me on the fact that in an indoor environment you could exceed 3dB gain because of all of the reflections. They even referenced a page of goo in a new book-of-RF-formulas. Then, they released a YouTube video, additional whitepapers, etc showing that ClientLink works through direct line-of-sight (which limits it to 3dB and isn’t really beamforming at all). I posted a comment on that video, and they immediately made it private so that I could no longer see or access it. Not really sure what they’re trying to hide, but they sure are hawking this “technology” to their VARs and customers.
Thanks Devin. Unfortunately, most IT managers are not equipped with the technical background you posses, nor do they have the time to thoroughly evaluate the various vendor solutions. Hence, I am working on a research report that will help the IT manager/Architect make sense of these solutions.
Antionette Whelan