Packet core

Juniper Challenges Cisco in the Mobile Core

Juniper Networks Inc. (NYSE: JNPR) has unveiled a suite of mobile packet core capabilities that, it hopes, will help it compete with key rival Cisco Systems Inc. (Nasdaq: CSCO) in the wireless data market.

Juniper's offering includes data offload, video traffic management, and Evolved Packet Core (EPC) capabilities. (More on those later.)

Unlike six months ago, both router vendors now have mobile packet core technology to talk about, and each hopes to take advantage of the spotlight at next week's Mobile World Congress (MWC).

The key word there, with regards to Juniper, is talk. Its Project Falcon, disclosed late last year, is still in Powerpoint mode, although the company is now talking about some of the specifics. (See Juniper Looks Inward for Wireless.)

Cisco is further down the road: It has real product following its acquisition of Starent, a deal that closed in December. (See Cisco Sews Up Starent and Cisco to Buy Starent for $2.9B.)

But while Cisco now has a foothold in the mobile packet core sector, it has plenty of work to do as it's only just starting to meld Starent's ST40 platform with the rest of the company's network.

Of course, Cisco and Juniper aren't the only companies targeting the mobile core sector, which, according to Heavy Reading, was worth about $1.5 billion last year. Alcatel-Lucent (NYSE: ALU), Ericsson AB (Nasdaq: ERIC), Hitachi Ltd. (NYSE: HIT; Paris: PHA), Huawei Technologies Co. Ltd. , Motorola Inc. (NYSE: MOT), NEC Corp. (Tokyo: 6701), Nokia Networks , ZTE Corp. (Shenzhen: 000063; Hong Kong: 0763), and Tellabs Inc. (Nasdaq: TLAB; Frankfurt: BTLA), which recently acquired WiChorus, all have some say in that market. (See Evolved Packet Core for LTE and Core Blimey! Tellabs Buys WiChorus.)

But given the Cisco/Juniper rivalry, and the fact that so many people asked what Juniper would do in response to the Starent acquisition, it's only natural to picture the two squaring off at MWC in Barcelona. (See Cisco/Starent Deal Hurts Juniper.)

Juniper's plan
Project Falcon is Juniper's answer to the Cisco/Starent combination. It involves a Juniper-built packet core and a universal edge network, the latter carrying any type of service.

The three capabilities being unveiled in Barcelona are the project's first "deliverables," Paul Gainham, Juniper's director of service provider marketing for EMEA (Europe, Middle East, and Africa), tells Light Reading, and they'll all be housed on the vendor's MX 3D "universal edge" family of routers and line cards. (See Juniper Launches New Routers.)

They are:

  • Traffic Direct (launching in the second quarter): Offloads mobile data traffic onto the Internet or landline IP infrastructure to ease congestion in the mobile packet network. Gainham says Traffic Direct is already being used in a live production network and is in trials with several other operators, but he declined to name any of the carriers involved.

  • Media Flow (launching in the second quarter): A content delivery network (CDN) capability for managing and caching video across the network. Juniper has integrated video-handling software from new partner Ankeena Networks Inc. to enable this functionality. (See Ankeena's Prabakar Sundarrajan: Conquering Content.)

  • Mobile Core Evolution (launching toward the end of 2010): This is Juniper's Evolved Packet Core (EPC) platform for 3G and Long Term Evolution (LTE) deployments. Gainham says Juniper's differentiator here is that the platform is "open," enabling third-party developers and carriers to build their own applications for the product. "That's what's been missing from the other vendors -- the ability to bring that level of innovation onto the network," boasts the marketing man.

But isn't Juniper worried that it's way behind the market? Naturally, it doesn't. Gainham points out that the LTE market is still in its infancy, and believes Juniper, by developing its EPC capabilities in-house and so avoiding any integration issues, won't be left trailing in the wake of Cisco/Starent. "We won't miss the big boat," proclaims the Juniper man.

Starent at last
Cisco, meanwhile, is eager to talk about adding Starent into the product portfolio, but hasn't had time to get much done.

Most noticeably, the Starent ST40 mobile core chassis will be renamed the ASR 5000. That gives the box superficial ties to the ASR 9000 and ASR 1000 routers, but, of course, it's got Starent in its guts.

Cisco plans to emphasize the number of things the ST40 has always packed in. It's built to handle 2G, 3G, and 4G traffic in one chassis, for instance, which Starent has said is a boon because of the massive amount of signaling needed for 3G and, especially, LTE networks.

On a bigger level, though, Cisco's marketing will start to include messages about the mobile Internet, a creature that will require the wireless and wireline networks to work more closely. "It will call for tremendous scale end-to-end, not just in the Evolved Packet Core part," says Kittur Nagesh, Cisco's senior director of service provider marketing.

The only work Cisco's completed so far is in the ASR 5000's reporting capabilities. That involves creating ways to collect real-time information about what subscribers are doing -- information that could clue service providers in to new types of revenue-generating services. "It's just the first step, just six weeks after the acquisition," says Jonathan Morgan, a Cisco senior director of product marketing.

EPC developments are set to be a hot topic at MWC: Check out our Mobile World Congress Show Site for all the news and views before, during, and after the event.

— Craig Matsumoto, West Coast Editor, and Ray Le Maistre, International Managing Editor, Light Reading

For all the best news, views, pictures, and video reports before, during, and after MWC 2010, check out our Mobile World Congress Show Site.

Pete Baldwin 12/5/2012 | 4:42:58 PM
re: Juniper Challenges Cisco in the Mobile Core

As DesiEngineer pointed out on the message board last month, it's probably too late for Cisco/Juniper to bother getting into the 3G/LTE radio network.

That means they're both banking on IP, building a story where "convergence" means the integration of the wireline and wireless IP networksto create better subscriber visibility and three-screen-ness and all that.  As opposed to an end-to-end story of owning the radio network and the core network.

It's going to be an interesting debate.

OpticOm 12/5/2012 | 4:42:57 PM
re: Juniper Challenges Cisco in the Mobile Core

No way Juniper would be interested in them now.

netsalesman 12/5/2012 | 4:42:56 PM
re: Juniper Challenges Cisco in the Mobile Core

The RAN (Radio Access Network) is off limits for CSCO and JNPR. It's too crowded, too "old" (there are players in the radio space that are there since decades, starting with analog RAN's) , the marginality is too low and it's a market not (only or mainly) product-quality driven (as the Nortel story proves).  The Mobile IP core on the other side is relatively new, much more open and...requires expertise (and footprint!) both vendors have. We shall see... 

desiEngineer 12/5/2012 | 4:42:51 PM
re: Juniper Challenges Cisco in the Mobile Core


There's tons of money to be made in an LTE network, and not the least is the IP infrastructure.  Juniper and cisco make it sound like they have "mobile" solutions, but in reality, they have IP solutions that can fit the, ugh, shall I say it, ecosystem.

Frankly, I think there's more money (profit) to be made in the backhaul and core IP network than in the RAN.  So hats off to cisco and Juniper for sticking to the LTE sidelines and having a profitable business plan.


quicktime 12/5/2012 | 4:42:51 PM
re: Juniper Challenges Cisco in the Mobile Core

I agree that it is too lat for Juniper/Cisco to get in based on the following:

1) Do they have enough expertises  on wireless?

2) Do they have enough experiences on wireless?

3) Was the market new or established?  If established,

    How could they compete with current players in the arena

    for years?

I thought for both companies, they are strategic rather than real direction.

Wireless is not that simple, and owning edges on IP does not means success

 in the future converging.


Sign In