Meru Networks this week announced the concept (and their implementation) of virtual ports for wireless LANs. This is a little complex, but so is virtualization, which is the art of making something appear real when it is not. Most IT professionals are very familiar with virtualization as applied to virtual machines, taking advantage of hardware features that have been in x86 microprocessors for some time to create many virtual processors that behave the same as a real one. Virtualization is saving big bucks at server farms everywhere, so the benefits are easily quantified. A virtual wireless LAN, or, as Meru calls it, a virtual port, is a little different; it virtualizes the connection between the AP and the client and allows a more granular level of control over each client station, in much the same way as switch ports on a real managed Ethernet switch can be configured to customize each client link. Meru is going after an objective that is very difficult to accomplish, because of the client-centric nature of the 802.11 protocols themselves, and the vagaries of radio propagation, interference, and other PHY-layer artifacts.
But I find this development particularly significant because Meru implements all of the required functionality on the infrastructure side - there's no special hardware, firmware, or client-side software required, although WMM (the Wi-Fi Alliance's spec for 802.11e, present on most clients shipping today) is required for best results. As you may recall, I have been in favor of standardizing client-side capabilities under the 802.11 umbrella for many years. I've made the argument for determinism in load-balancing and roaming and other components of client-side control as best I can, urging 802.11 and/or the Wi-Fi Alliance to take action here with an (albeit optional) client. But if Meru can really gain the benefits of controlling the client without modifying the client - wow, that could be significant.
I am looking forward to testing Meru's vWLAN as soon as I have the chance. And I remain impressed with the level of innovation that Meru continues to bring to the marketplace - vWLAN follows on the heels of their Virtual Reality RF analysis product and the PHY-layer security system announced earlier this year. As I have written before, there are still great opportunities to both expand and refine WLANs ahead of us, and Meru is clearly one of the industry's leaders in this pursuit.
Advertisement: |
Mathias is a principal at Farpoint Group, a wireless advisory firm in Ashland, Mass.
The Leader in Network Knowledge
Fixing the unfixable Meru's virtu
Having this and client-side capabilities under this virtual port I want to know the part that isn't real that seems like it is. Because most of this seems real to me except for the alliance analysis. I remember that I thought everyone except for familiar virtualization was roaming Phy-layer security system ahead earlier this year one leaders in pursuit. Also the Phy-layer virtual reality did the benefits clearly expand the opportunities to refine still great innovation. Inspiration is the mother of invention and the children are reality you've created wy creation. Implies the application all of Meru networks.
The virtual part is...
...the relationship between the client and the infrastructure. Meru's goal is to allow client control analogous to that in a wired netwrok, which features a one-to-one relationship between clients and ports, and thus all the control that that implies. That's the idea - now, of course, the implementation is not the same, but Meru intends to show that the benefits are.
I hope this helps, and thanks for the note.
Craig.
No answer to capacity. no consideration of neighbor networks
It's nice to have more granularity with QoS which this appears to promise.
But, one channel Wi-Fi is rediculous, if cell towers tried this no calls would ever get through.
RF fixes are nice with one channel but not at the cost of all the bandwidth you give up!
Kind of like saying you can save a lot of gas with a single cylinder engine - but who wants to drive a moped?
Most enterprise solutions are multi-channel...
... as they need to be. And I think Meru's solution will work with multiple channels...
Thx. Craig.
Anon, if your WLAN only has
Anon, if your WLAN only has 3 AP then your comment is true. Unfortunately most of enterprise WLAN have more than 3 APs, thanks god, so when you are dealing with lot access points in a Micro Cell architecture you're getting hurt by interferences and the performance of your WLAN goes down down dramatically. Not even talking about client density here... SO you're driving a 4 cylinder car but your speed is capped to 5 miles per hour and you are driving on a 3 (1, 6, 11) ways highway under heavy traffic so you get accidents, collisions, all the time. Cars are moving from one line to an other without notifying the other users... etc so in a Micro Cell architecture you have few to no control on the clients, results are unpredictable and performance is poor. Unpredictability is not good for enterprise.
Last the good thing with a single channel architecture is that you can layer, stack the 3 channels to provide capacity. You're not limited to just one ;)
Post new comment