Tahoe Takes Wireless to the Edge

Not all edge routing devices are created equal. At least that's what Tahoe Networks is trying to convince the networking industry.

The Silicon Valley startup is one of four new startups that is developing an edge aggregation router specifically designed to link wireless networks to wireline IP backbones.

Today, the company announced that it has entered marketing and technology agreements with Xacct Technologies Inc., an accounting and billing provider and Kabira Technologies, a software company focused on integrating personalized data services to improve deployment and provisioning times (see Tahoe Announces Partnerships). By integrating software from these two compannies, Tahoe says it will be able to help its wireless carrier customers bill and better manage IP data services like wireless Internet access and instant messaging.

The company, launched in January of 2001, was founded by Anthony Alles and Arthur Lin, two former Cisco execs who also founded Shasta Networks, an IP service switch company that was bought by Nortel Networks Corp. (NYSE/Toronto: NT) in 1999. Shasta was developed to provide network-based IP services like VPNs and firewalls to wireline networks. Alan Cohen, vice president of marketing for Tahoe Networks, says the Tahoe founders took the expertise they developed in their years at Cisco and Shasta to create an edge router focused on the problems of the wireless market.

Edge routing players have also seen the need for a new class of product, but most have taken a different approach. Instead of building new platforms, companies like Cisco Systems Inc. (Nasdaq: CSCO), Juniper Networks Inc. (Nasdaq: JNPR), and Unisphere Networks Inc. are adding enhancements to their current platforms. Juniper has teamed up with Ericsson AB (Nasdaq: ERICY), and the two have recently announced a jointly developed product called the J20, a wireless edge router that is based on Juniper's M20 20-Gbit/s routing platform (see Juniper Unveils 'Wireless Router'). And Nokia Corp. (NYSE: NOK) purchased Amber Networks last year for $432M (see Nokia Nabs Amber for $421M).

So what are the requirements for a router connected to a wireless network? Cohen says that ASIC-based high-speed packet forwarding is not necessary for wireless routers, where current users connect at data rates of 56 Kbit/s or below. What such routers do need is much more software-processing power to do things like manage highly volatile user activity, provide rich IP services like IPsec VPNs, network-address translation, and unique accounting information, and integrate into operations systems of both wireless providers and IP data providers.

"Juniper and Cisco have built routers to forward packets fast," he says. "And that isn't what's needed in the wireless edge. The requirements are different." Cohen says it is important for a wireless aggregation router to be able to handle thousands and tens of thousands of simultaneous sessions because of the nature of wireless connections. These routers must also be able to handle a rich suite of IP services like virtual private networks (VPNs) based on IPsec, because as these networks handle tens of thousands of users, it will be necessary to provide security to its customers. And finally, integration into the wireless operational systems will be crucial as service providers try to track and manage usage on their networks.

Kevin Dillon, director of product marketing for Juniper, agrees that wireless applications require different features. But he says that it is still important to be able to handle large amounts of traffic.

"Individual wireless subscribers may be using lower rates to connect to the network," he says. "However, that device is supporting large numbers of those subscribers and it's important for them to have the capacity to aggregate that traffic."

He also says that through Juniper's joint development with Ericsson, the company has been able to focus on offering services that wireless providers find important, such as IPsec VPN.

Tahoe isn't the only company trying to build a router specifically for the wireless market. There are currently at least three others focusing on this area: Cambia Networks, Megisto Systems, and WaterCove Networks. WaterCove has raised $50 million and Megisto $60 million. Tahoe has secured $50 million from top VCs like Redpoint Ventures and Accel Partners. Cambia is the only one that seems to be struggling. It has only announced an $11 million round, which closed January of 2001.

Analysts following the space say that wireless router startups are definitely companies to watch.

"I think these companies may be on to something," says Hilary Mine, executive vice president of Probe Research Inc.. "There are different requirements for wireless and they're trying to address them."

- Marguerite Reardon, Senior Editor, Light Reading
Page 1 / 3   >   >>
edgecore 12/4/2012 | 10:40:21 PM
re: Tahoe Takes Wireless to the Edge

cruiser 12/4/2012 | 10:40:18 PM
re: Tahoe Takes Wireless to the Edge they have the money, the team looks good. the question is whether the wireless operators will wake up to opportunities in ip services and if they do, will they build efficient ip networks from the ground up so they can actually make money on those services? probably some market risk here but that's the case with any new emerging business. as a user, i'm happy to know at least the technology to facilitate ip services to the handset is getting off the ground. i'm tired of carrying around a bulky pda with a wireless module running cdpd, plus my cell phone and not really getting a lot out of either. i have a $300 phone and can't look at the web without long waits and an arcane, colorless ui.

anyway, these companies look cooler than optical right now.
edgecore 12/4/2012 | 10:40:14 PM
re: Tahoe Takes Wireless to the Edge Where would this box sit exactly in the network?

When I drive by these cell antennas and see the big huts at the bottom, what gear is in there?

lightsmith 12/4/2012 | 10:40:12 PM
re: Tahoe Takes Wireless to the Edge I heard Cambian has been closed.


John Casper 12/4/2012 | 10:40:08 PM
re: Tahoe Takes Wireless to the Edge Virtually all wireless start-ups will die a natural death.

The reason is simple - they neither have resources nor even a fractional understanding of wireless technology that can compare with likes of Ericsson, Motorola, Lucent, Siemens, Nokia, Philips, NTT and Nortel etc.

No one can ever convince that a 150 persons startup has a devine vision that those mega corporations don't have, AND can do what those giants can't do with all the technical and finacial resources in their hands.

A wireless start up has zero chance of succeeding simply because wireless industry is based on well defined standards, and every one knows those standards well in advance (because all come through years of ITU/IS process). It is impossible for any one to introduce a revolutionary technology in wireless unless it has got through standardization process. Wireless and cellular industry works very differently compared to simple routers or landline business. You buy turnkey networks instead of piece-meal routers and bridges.

Before people start hyping - give me a single example of successful start up in wireless industry (except Qualcomm which got supported by US government, and also had patented technology that received tremendous help from the government in getting it passed as IS-99). Qualcomm, as you know, now only makes CDMA chips. Even, Qualcomm realized that wireless equipment business is a very hard game despite the fact that you own all the necessary patents.

Good for dreaming, but there are laready gazillion implementations of GGSGN and SGSGNs. Industry isn't dyeing to have more.

This not only applies to Tahoe it also applies to virtually all wireless startups - a small player has no chance in the field crowded by super-heavyweights. Here, you don't have one CISCO, but dozen Ciscos-++es to take care of.


cruiser 12/4/2012 | 10:40:07 PM
re: Tahoe Takes Wireless to the Edge wow jc, this is pretty strong stuff. you're right though, it's risky. there have been probably even more failed startups in wireless than in optical over time. a question for you: do ggsns do subsriber management and routing? are these systems complementary to ggsns or do they replace them?

i would think these guys are all acquisition bait because it's hard to imagine any of them going public on one product unless they go to market with big partners and those big partners will want to buy the startups vs. share revenues and mindshare with future competitors. same dilemma as the startups in wireline i should think. someone will figure out how to make this formula work though. there's money to be made so someone will figure out how to unlock that business opportunity because sprint can't offer 4,000 minutes for $39.99 forever, really voice only, and stay competitive.
flanker 12/4/2012 | 10:40:07 PM
re: Tahoe Takes Wireless to the Edge I.M.O.
h 12/4/2012 | 10:40:06 PM
re: Tahoe Takes Wireless to the Edge how does flarion and winphoria doing compared
to other wireless startups ?

last I heard winphoria was building a soft MSC.
any thoughts on how they are doing ?
lightsmith 12/4/2012 | 10:40:05 PM
re: Tahoe Takes Wireless to the Edge John, you raised an interesting point regarding standard and startup.

But the same arguement can apply to virtually all telecommunication areas, SONET, DWDM, IP, ATM, MPLS etc. Afterall, we have to follow the standard in order to achieve interoperatablity.

The reverse side of this arguement is that a proprietary implementation will raise the entry barrier for non-incumbent players, which most of them are startups.

I'm not expert on this but I beleive 3G is a very well defined standard. In theory service provider can buy base station from one vendor and GGSGN from another and put those together. Of course the real world is not that simple.

Just my two cents.

willywilson 12/4/2012 | 10:40:04 PM
re: Tahoe Takes Wireless to the Edge I'm not expert on this but I beleive 3G is a very well defined standard. In theory service provider can buy base station from one vendor and GGSGN from another and put those together. Of course the real world is not that simple.


You know what would really be cool? If any of the 3G stuff actually worked.
Page 1 / 3   >   >>
Sign In