Optical/IP Networks

Sprint Nextel Surfs the Pseudowires

Sprint Corp. (NYSE: S) is considering a rollout of pseudowires for its wireless network and has issued a major request for proposal for the technology, Light Reading has learned. Pseudowires allow carriers to carve up bandwidth into small, packetized virtual channels, helping reduce the need for additional T1 lines to backhaul traffic from cell sites to the network backbone.

While the technology has been deployed in several international markets and in pockets of the U.S., a Sprint deployment would give pseudowire a major stamp of approval. (See Wireless May Boost Pseudowire.) The impetus? It's a money-saver.

"This would definitely be huge for the market," says Scott Clavenna, chief analyst at Heavy Reading.

A pseudowire is an emulation of a native service such as Asynchronous Transfer Mode (ATM), Ethernet, or Frame Relay over a packet switched network. Despite the promise, the technology has its drawbacks: Standards are still immature, so vendor interoperability is an issue. And it isn't bandwidth-efficient: It actually expands the number of bits in a data stream. But it's considered a less expensive alternative to T1 lines for cellular operators -- who find themselves going from supporting one or two T1 lines per base station to supporting four or more, as they try to keep up with data services, Clavenna says.

"If you use pseudowires, either right at the base station, or at some aggregation point in the network, you convert the TDM backhaul signals to packet and send them the rest of the way as 'virtual leased lines,' " Clavenna explains. "In a nutshell, you can greatly reduce the cost of backhaul of 3G data traffic and support a transition to Ethernet and IP."

A Sprint spokesman confirmed the desire to invest in pseudowire and additional backhaul technology.

"We use primarily ILEC T1 for backhaul," says Bruce Hoffman, director of technology development planning at Sprint Nextel in Overland Park, Kan. "There's a lot to be said for using ILEC T1. It's a well understood technology. The challenges are around the scaleability that's going to be necessary in the future -- scaleability both from a capacity perspective and also from an economic perspective."

"We'd need to have something that would emulate T1, which we might be considering," Hoffman says. "And logically we'd look at something like pseudowire. The RFP is in response to that. That's all I'll say about the RFP."

Companies dealing in pseudowire technology include networking giants like Alcatel (NYSE: ALA; Paris: CGEP:PA) and Cisco Systems Inc. (Nasdaq: CSCO), smaller companies like RAD Data Communications Ltd. , and specialists such as Axerra Networks Inc. and Hammerhead Systems Inc. Axerra and Cisco are believed to be among the favorites for Sprint Nextel's pseudowire deployment, according to several sources. RAD Data is in the running as well.

"We are involved in this bid," says Larry Jacobs, vice president of marketing at RAD Data. "We are involved in virtually every one of the [pseudowire] bids that are out there."

Other carriers are expected to jump into the pseudowire fray, too. "All large carriers from Tier 2 up have an RFI [request for information] or RFP out on this," says a source at a large gateway vendor, who is familiar with the RFP. "It's a big wireless-to-wireline convergence savings opportunity that anyone with both networks will have to do." Sprint's Hoffman says that one reason for the technology deployment is the carrier's upcoming EV-DO Rev A deployment. That will drive the need for backhaul bandwidth.

Sprint's EV-DO Rev A services will be available to initial customers at the beginning of 2007. Plans call for reaching a possible 220 million customers with Rev A by the end of the third quarter of 2007. Rev A promises data download speeds up to ten times faster than those of DSL, according to Sprint officials.

— Carmen Nobel, Senior Editor, Light Reading

Scott Raynovich 12/5/2012 | 3:51:56 AM
re: Sprint Nextel Surfs the Pseudowires Anybody know the scope of this RFP? And what are Cisco's chances?
dav_006 12/5/2012 | 3:51:54 AM
re: Sprint Nextel Surfs the Pseudowires Well the implementation that Juniper has for both ATM and Ethernet Psedowires is very good as it is very easy to use .. but Cisco is not far behind. I would put Juniper ahead of Cisco and Cisco ahead of Alcatel GǪ.
airfiber 12/5/2012 | 3:51:53 AM
re: Sprint Nextel Surfs the Pseudowires Not relevant which vendor is ahead at this stage. Sprint Nextel is just fishing. This could be better described as an RFI.

Sprint and the other carriers are still two years away from seriously dealing with the mulitiple upgrade issues for backhaul.
lolla 12/5/2012 | 3:51:47 AM
re: Sprint Nextel Surfs the Pseudowires I agree with the two years!
PSW does not seem a reality today for the mobile backhauling.
alchemy 12/5/2012 | 3:51:46 AM
re: Sprint Nextel Surfs the Pseudowires lolla writes:
I agree with the two years!
PSW does not seem a reality today for the mobile backhauling.

What makes you say that? From where I sit, it's likely that this application is going to get launched this year with fairly large volumes next year. I know companies like Cox already have some of this gear in their cable network. Sprint already has lots of intertwined relationships with the cable companies. Sprint has, what, 20K+ towers? I'd suspect that most of them are within a few hundred meters of an MSO HFC plant with DOCSIS in it. CableLabs has released a pseudowire spec that uses the high quality clock from the CMTS and DOCSIS QoS mechanisms and there are a number of vendors participating in the initiative including DOCSIS-savvy Cisco and Arris. There are also a number of startups in the space so there will be significant price competition. I'd wager that a large fraction of Sprint towers will at least use a cable DOCSIS pseudowire connection as a disaster recovery mechanism and that Sprint will use this as a weapon to continue to drive down costs for digital leased lines from the ILECs.

What kind of dollars are associated with this? I'd guess 5,000 pizza boxes at something like $1K/box and a much smaller number of big aggregator boxes at the mobile switching centers. Given how cluttered the market is with Pseudowire startups and Israeli shops, I don't see this as being particularly profitable for any equipment vendors. One vendor will end up getting the bulk of the business. I'd expect that the vendor would end up being either Cisco or Arris because of their cable DNA.
airfiber 12/5/2012 | 3:51:45 AM
re: Sprint Nextel Surfs the Pseudowires It's good to see that bubble mentality beginning to surface once again... Now let's go back to reality.

"MSO HFC a few hundred meters from cellsites"!? Tell me, have you ever been to a cell site? Cable MSO outside plant primarily go to neighborhoods (where the people are, fancy that!) HFC plant is used to deliver video, internet, and voice over 800Mhz of analog/linear RF until it reaches the set-top box. Tell me, what box are you referring to that will take this 800Mhz RF and produce an equivalent DS-3 to a cell site?

It's this magical thinking that drives me up the wall. I said two years, sorry I was wrong, make it three
SeaRaven 12/5/2012 | 3:51:28 AM
re: Sprint Nextel Surfs the Pseudowires Where can we get a copy of this RFP? I would like to see it. Can someone share it?

Sign In