Carrier WiFi

Futzing the Protocol

5:30 PM -- If you’ve not been following this, a recent article in Network Computing is raising a question as to whether Meru Networks Inc. is altering a field in the 802.11 header, the result giving them a performance advantage at the expense of other nearby WiFi networks. As I write this, there is no resolution to this issue. Is it really an issue? Is it just a bug? Is it a violation of 802.11 in some way? Hmmm.

While most of the people I’ve spoken with are adamant that Meru has violated both the 802.11 standard and the WiFi spec (I doubt the latter, but I’m checking), I’m going to reserve judgment for the moment, and we should most certainly give Meru the benefit of the doubt until all of this is resolved. It’s certainly possible that a given enterprise, for example, might operate a Meru-only environment, and this problem, if it is one, would never really show up. There’s certainly no violation of FCC or other regulatory rules. But there’s a more important issue afoot here.

If one really wants to boost one’s access to the airwaves, why not just use 802.11e, AKA WMM? Just set your priority to high for everything, and, voila, really great throughput until everyone else does the same thing. So, one of the key benefits of .11, playing nicely with other WLANs, is also a source of potential trouble down the road.

I personally find it hard to believe that any vendor would be so dumb as to intentionally fudge the protocol, since it’s so easy to prove this should it be the case. But stay tuned -- there will be more on this issue shortly. In the meantime, monitor those airwaves.

— Craig Mathias is Principal Analyst at the Farpoint Group , an advisory firm specializing in wireless communications and mobile computing. Special to Unstrung

<<   <   Page 2 / 5   >   >>
CleanSheet 12/5/2012 | 3:33:45 AM
re: Futzing the Protocol GTHill's suggested test misses the forest for the trees. Instead of getting into the nitty gritty, why not ask the bigger question which is:
Business sense mandates Meru must have obtained or filed for a patent for their secret sauce, if they have any, that explains their "advantage." In the absence of patents (or patent filings) it is more likely their secret sauce is cheating.

I don't see any patent filings from them that explain their secret sauce. And, if they had any patents or patent filings in the works wouldn't Juniper or Cisco have jumped at the opportunity to buy them out to obtain that intellectual property?

Those bigger questions obsolete the need for any further diligence on Meru. Companies with troubled genetic histories die early; some, in a growth market, might flail around and eventually die off. Meru may have made it this far in a thus-far growing Wi-Fi market. What follows will confirm their flawed genetic history and lead to what the market has concluded about Meru: it is a stink bomb.

What transpired at Meru for the ex-CEO and "co-founder", Ujjal Kohli, to leave abruptly and not even mention his Meru experience or affiliation at his new job as CEO of Rhythm New Media? Check his profile on www.rhythmnewmedia.com.
What explains the simultaneous departure of ex-CFO and "co-founder", Nicholas Mitsakos?
What business relationships did they have in addition to Meru?
What do executives that left Meru have to say about Ujjal Kohli or Nicholas Mitsakos?
Wht do entrepreneurs and others in companies where they invested have to say about UK and NM?
What does Ravi Sethi, a business partner of UK and NM, have to say about them?
What are the non-business relationships between the ex-CEO, Ujjal Kohli, and his VP, Kamal Anand?
What explains the departure of the most recent CFO and VP-Eng?
Which startup in Silicon Valley has a VP-International Sales and another VP for Worldwide Sales?
Which has a VP for Corporate Strategy who also performs as the above mentioned VP-International Sales?

With a genetic history as bad as Meru has, their future can't be promising. Sure, they'll get the occasional bone missed by Aruba and Cisco and others. Let's see how long that can continue.
lbknick123 12/5/2012 | 3:33:43 AM
re: Futzing the Protocol Prescient - Now you have me curious about their current management, and you seem to know quite a lot about this place. What's the story with their CEO? What's his history? Only common denominators between their old management team and new seem to be Kamal Amand and their CTO. Why are they still there if there was such a tight link with the old CEO?

This company appears to have a lot of smoke and mirrors. Looks like they cheat, industry tests show that either they cheat standards, or that their stuff is just too plain buggy to pass a test. But at the same time they brag on their website that the Gartner Group loves them.

Smoke and mirrors or the real deal?
CleanSheet 12/5/2012 | 3:33:42 AM
re: Futzing the Protocol First, about their "founding" management team, namely Ujjal Kolhi and Nicholas Mitsakos. They formed a venture partnership with Ravi Sethi and screwed him over. The LPs in the venture, including SunAmerica, got tired of being lied to and screwed around by UK and NM. Result was the two were fired from their venture responsibilities and the venture firm shut down. Ask questions around about MKS Ventures.
Entrepreneurs in whose companies UK and NM invested were lied to and later stabbed in the back by them. Ask founders at companies MKS invested in, where UK or NM served or serves as a director still.
Executives at Meru were scapegoated for business problems and fired or left on their own. Ask ex Meru executives, including the most recent CFO and VP-Eng that left several months ago.
Finally the board woke up, fired the two, and in crisis mode launched a desperate search for a CEO. Ask Meru's investors.
CEO candidates with a clue walked away, realizing a company with this genetic history has slim chances. Current CEO, formerly a VP at Proxim and SGI--hardly known for business success or ethics-- took the job as it gave him a chance to be CEO and makes his next job as CEO easier. Meru's circumstances didn't allow for a longer search and there was no assurance the candidate pool would get better. They are having a big problem today trying to fill the VP-Eng and CFO positions. Explains why those positions remain unfilled for many months.

CTO retained for technical value.
The original VP-BD is now VP-International Sales/Corporate Strategy and was retained to project an image of management continuity.
In companies Meru's size the CEO drives strategy (with, if it's a growing company, the VP-BD). Here we have a CEO who clearly acknowledges he can't do strategy and has given that responsibility to someone who is also running International Sales. All this when there's a VP-WW and a VP-BD. Says enough about the CEO's inexperience and poor judgment.

If Meru had confidence in its secret sauce it'd have gone ahead with 3rd party tests, competitive bake-offs, patents, etc. That they are paranoid about 3rd party tests and bake-offs, and have little to show by patents, confirms company's DNA is the same as the founding management team's. Hardly surprising.
lrmobile_GTHill 12/5/2012 | 3:33:41 AM
re: Futzing the Protocol I agree that my test (again, I wrote that before seeing that they performed the exact test) is very focused on the technology itself.

You seem to be very knowledgeable about Meru as a corporation and their business practices so I'm in no position to discuss those points.

However, do you really think that Meru would file a patent? I'm sure that it would be granted, but that would be publicly saying that they are not conforming to the standards. Enterprise implementers are very leery of any product that isn't standardized so they would almost guarantee their demise.

After filing a patent, Meru's only hope would be to get their technology standardized and have a significant jump on the market. Meru doesn't have the clout in the industry to get their technology past an IEEE TG.

Meru is gambling that people will be so happy with their improvements that they will look past these articles and implement their technology. They will continue to deny any variation of the standards and hope that the average decision maker either won't care or won't understand the differences in the "nitty gritty" of the technology.

I do agree that Meru's future is in jeopardy. I do know that more and more of my students are aware of Meru. But, the focus of vendors in any discussion is about Cisco, Aruba and Symbol in that order. Meru is still currently a footnote.
lbknick123 12/5/2012 | 3:33:41 AM
re: Futzing the Protocol
Unless you assume that Meru itself is going to be around for a long time (and given the rash of negative test results that have shown in in the press lately, that's a open question), a customer must consider the prospect needing to go elsewhere in the future to expand their WLAN.

Customers, particularly big ones, rarely roll out complete coverage. So let's say a customer deploys meru gear in one department or area. One year from now, if meru isn't around, they'll be forced to rip out the meru gear because it kills the performance of adjacent standard gear.

I recently read Meru's rebuttal to the Network computing article, which btw read like a very panicky, over the top response. Essentially, meru's cto said that network computing guys didn't know what they were doing, and that the only reason Cisco's AP had lousy performance next to a meru AP was due to a cisco bug. Look .. as a cisco user, I know their gear is certainly not perfect. But I also know that my site has one area (operating next to Cisco WLAN) with aruba gear. (I won't say which we prefer.. ;) What I will tell you is that there's no performance stomping going on one way or the other. Really a rather frivolous argument by Meru's technical guru.

When we started our evaluations over a year ago, we simply couldn't find any third party validation of meru's system. read a lot (a whole lot) of press about new products and partnerships and awards. What we couldn't find was any testing where meru agreed to participate. Now, all of a sudden, three show up this month, and I guess the results assure me that we didn't miss anything here.
peanutoat 12/5/2012 | 3:33:40 AM
re: Futzing the Protocol Howdy,

Looks like I erred by ever sending a message to these boards. I think most people on here must work for one of the four companies mentioned (Aruba, Cisco, Meru, and Symbol), and so it is just a crappy high school prom, with the posturing and all.

I could care less about Meru's management. The way I hear it from the sales reps, this industry has gotten down to mud fighting at this point. Everyone's got an accusation. "Alan Cohen is a bastard." "Dominic Orr is screwing over his employees." Unless those guys stop over and try to sell me directly, I really don't care. I have a job to do.

See, folks, this is what hopping on the bandwagon is like. Did Meru break the standards? I am the only one who even yapped about the details of TXOPs. Is anyone else interested in whether that is breaking the standard, or how? Seems to me like this Molta fellow got suckered into a Cisco special, only because this Bob O'Hara guy (from the Network Computing responses) wrote a book on 802.11. Explains why they published a he-said-she-said.

I find it dang hard to believe that Cisco had a bug that made their AP break. But why didn't Meru take more than half the air time, as their Chief Technologist points out? Cisco says they should take more than half, if anything they said makes sense. It's a zero-sum game, so why are things not adding up? No collisions, just how much airspace you occupy. Show me sniffer traces, or withdraw the article if you are not going to let me judge.

And nuclear arms races for standards? Cisco sells CCX, which is not the standard, no matter what their reps tell me. The way I see it, CCX is cheating. Or explain why Network Computing ran a test for voice and didn't use WMM, then complains when one of the two vendors implements parts of WMM? Smells worse than a wet dog on a hot day.

I don't know Meru from Adam. But if they have something that takes Cisco toe to toe, and causes this much of a ruckus, I want to learn more. I want to know what Cisco is fighting so hard against. I want to know why Cisco is offering 80% discounts on wireless to any potential customer that Meru is talking to, from what my guys say. Anything worth conspiring against is worth finding out what you're missing, folks.

The more you guys pick on Meru, the more I'm gonna be fascinated about what they must be doing right.
lbknick123 12/5/2012 | 3:33:40 AM
re: Futzing the Protocol Reading Meru's network computing test defense piece:

meru claims that their products don't kill adjacent, other-vendor AP's performance, as shown in the network computing test. They say it is a Cisco bug that forced the Cisco AP to perform poorly in the presence of a meru AP.

Then, they say the following:
"Almost all our customers have legacy access points that interoperate perfectly with Meru, and Network Computing's tests were performed with standard clients."

So .. I'm sorry .. doesn't cisco own around 75% of the wlan market? They've been selling for years, and the vast majority of install base out there must be cisco. Thus -- majority of meru's customers probably have cisco APs.

So do Cisco APs work with meru, or don't they?
wii 12/5/2012 | 3:33:40 AM
re: Futzing the Protocol Prescient,

You should argue/comment against the technical aspect of the article or technical aspects of Meru. You are boring everyone with your same Meru Management rant (that too, about an old management team which is no longer part of Meru) irrespective of what the article is talking about.

Regarding why Meru has not filed for patent; do you know coca-cola http://en.wikipedia.org/wiki/T... has not patented the GÇ£secret sauceGÇ¥ which drives their whole core-business since so many years? That is because there is something called trade secret also and it is up to the innovators and company to file patent or hold on to trade secret.
CleanSheet 12/5/2012 | 3:33:39 AM
re: Futzing the Protocol First, talking about Meru's management (past and current) is relevant. It helps employees, customers, investors already associated or considering associating with Meru to make informed decisions.

Second, it is disingenous for Meru to trot out its "secret sauce" explanation. Secret sauces work say, with the corner shop selling falafels. They don't have to comply with protocols, standards, inter-operability, and such expected in the networking industry. Meru may be innovating in the business model by being the first tech company to use "secret sauces" instead of patents but nobody is buying it.

If that secret sauce and trade secret is valuable, why did the recent VP-Eng and CFO leave, not long after they were hired with fanfare? That secret sauce sure must have been strong enough for Cisco and Juniper to curl up their lips with disgust and walk away. If Meru is that confident in its trade secret what explains their fear of third party tests, competitive bake-offs? their problems interoperating with other vendors who have no problems interoperating between themselves? how come no customer, other than a few small school districts, bought into Meru's story of its secret sauce? no indepdendent industry analysts or evaluators?

Those that tout secret sauce and trade secret as Meru's competitive advantage are naive or in denial. Or a paid shill. For Meru to engage in that strategy confirms the CEO's inexperience mentioned earlier and the BS of the current and founding managers. Meru's DNA is partly derived from its founders, Ujjal Kohli and Nicholas Mitsakos, and that explains why Meru is opaque, disingenous, and paranoid.

Trade secrets/secret sauces can be reverse engineered. In this competitive industry, with so many sharp minds at work, with so many disgruntled ex-Meru engineers and executives, with the likes of Cisco and others playing hard and winning, why hasn't Meru's secret sauce been reverse engineered? Because...there isn't any secret sauce and reverse engineering a protocol violation is a laughable idea.

As to why Cisco offers steep discounts to those considering Meru, refer to the many companies that bit the dust under those circumstances. Meru is in a death spiral. To BS about secret sauces and trade secrets...
lrmobile_GTHill 12/5/2012 | 3:33:38 AM
re: Futzing the Protocol According to the test, in which the methods were sound, showed that when a Meru AP and a Cisco AP were sharing channel space, the Meru got much more access to the medium. Meru is blaming Cisco, but with Cisco vs any other vendor the "problem" doesn't exist.

My take: Meru is not playing by the standards.
<<   <   Page 2 / 5   >   >>
Sign In