As Mobile Video and other bandwidth-hungry mobile applications dominate the wireless communications infrastructure, wireless backhaul — the transfer of traffic from a wireless base station back to the core network (and ultimately the Internet) — become critical. In the past, T1s and E1s were used to do this, but increasing bandwidth demands encourage the deployment of fiber, which can be expensive. The latest approach is to use point-to-point wireless backhaul technologies, or a combination of wireless and fiber. In this six-part series, we take a look at what’s happening with wireless backhaul today.

 

The next player we shall examine is Ciena Corporation, which specializes in high-performance network systems, software and professional services to transition today’s networks so they can deliver more profitable services.

 

Ciena’s Taylor Salman, Solutions Marketing Director handling Mobile Backhaul and MSO markets, says, “Obviously, with the advent of HSPA [High Speed Packet Access] and 4G systems coming down the pike, some of them are already deploying with WiMAX (News - Alert) in the U.S. in particular, and there is a huge ramp-up in data services being provided over the mobile networks. Clearly, the systems that were designed to backhaul TDM voice and even some of the earlier data systems, can’t keep up with the newer bandwidth demands. When you look at some of the new phones that are appearing, such as the 3G iPhone and the Samsung Instinct, where they’ve solved the user interface problem, you can see that they’re really driving the usage of data services that have already been available and thus make good use of the new bandwidth available on these new systems. So, the cost is going up on the backhaul network, and leading to the desire by the operators to drive down costs.”

 

“Clearly, the choice in the marketplace is to adopt an Ethernet backhaul model,” says Salman, “from the existing TDM systems, whether they be PDH or ATM. One of the biggest surprises we’ve seen is the uptake for migrating to Ethernet. It’s much faster than in the past. I spoke with one network operator the other day, who is looking to rollout Ethernet over their existing SONET network, and were expecting to do it at a somewhat leisurely pace. About three months after our initial conversation, they said they had scrapped those plans, and they were now felt compelled to move to an all-Ethernet network immediately. I think they were surprised by the huge demand. We’re seeing quite a bit of RFIs, RFPs and RFQs coming from operators in this space.”

 

“Once the operator has made the determination that they want to adopt Ethernet backhaul,” says Salman, “the questions they ask themselves include not just, ‘How do it do it?’ but ‘What do I end up once I get there? What will my Ethernet network actually look like? What capabilities do I need?’ In the rush to move to Ethernet, network operators are scrambling to figure out what it all means. And that’s one of the areas with which we can help them.”

 

Salman drills down: “The first issue is how do you connect to the tower? There’s an existing infrastructure out there that they might want to leverage. Different towers have different traffic demands. There are differing geographies. Therefore, there’s no one-size-fits-all solution. You’ve got fiber, microwave and copper options. If you look at what operators would like to do, they’d like to have fiber to every tower. In fact, there’s one local operator I know of who says, ‘We will have fiber to 100 percent of our towers.’ I don’t know whether I believe they can do it, but it’s a legitimate claim, given the identity of the operator and the other current plans for fiber deployment throughout the United States. Nevertheless, once you get fiber installed, you potentially have unlimited bandwidth to-and-from the base station tower, and you can quickly deploy Ethernet services on that fiber. That’s the first choice. Now obviously there are construction issues and cost issues associated with fiber. Not all towers will immediately have demand for that type of bandwidth. But fiber deployment is pretty much straight forward, with the exception of perhaps what we at Ciena would call ‘True Carrier Ethernet’ capabilities over fiber, of which I will speak about in a moment.”

 

“From what we’re seeing in the marketplace, the second option is clearly microwave,” says Salman. “That’s because of the ability to scale to high data rates, to a hundred megabits per second per customer. Plus, microwave has a relatively rapid deployment time as oppose to fiber. And it can be a longer-term solution than a mid-band copper-based solution, such as a 1.5 Mbps T1. One of the exciting things that we see in the microwave space is the emergence of new microwave equipment vendors. Of course, microwave is not a complete panacea. There are licensing issues, interference issues, and, like it or not, maintenance on microwave isn’t cheap. You do have to do a truck roll and a tower climb ever time there’s a problem. But the nice thing about microwave is that it allows a balance for some operators between CapEx and OpEx expenditures. By that I mean the difference between leasing an Ethernet service from another provider, a landline provider, and building out their own network. They could build out their own, increasing the CapEx side, but some operators are doing both to balance their budgets. The CapEx on microwave can be pretty expensive but once you’re done, you’re really done. You just have the maintenance OpEx to deal with.”

 

“You’ve got legacy microwave vendors out there,” says Salman, “that would marry microwave with PDH or SONET, and have been trying to add Ethernet into those capabilities. Those systems typically both an indoor and outdoor unit associated with them. And then you’ve got some newer vendors out there who are doing pure Ethernet microwave which are just based on an outdoor unit and they provide point-to-point microwave services natively with Ethernet. That’s pretty exciting for us because what we provide with our Ethernet switch portfolio is an Ethernet switch that acts as an indoor unit for those type of systems, and now instead of doing point-to-point microwave hops, you can actually do end-to-end Ethernet services over microwave that are traffic-engineered, provide protection and have automated provisioning capabilities that we would assign to what we call ‘True Carrier Ethernet’. This would be carrier-grade OAM [Operation, Administration, Management] capabilities. Now when you do that, when you provide the traffic engineering piece combined with the automated provisioning over the microwave link, you can now start to look at new architectures that are much more cost-effective than traditional microwave architectures. In particular, microwave architectures for backhaul have been fully redundant tree topologies, for the most part, usually spanning off of some type of ring system. To protect that type of architecture, you’ve got to have dual links between each hop, so that if one goes down, you can hook to the other one on a hop-by-hop basis.”

 

“Now, if you can put and Ethernet switch with traffic engineering at each of those microwave hops, you can now migrate from a tree-based architecture to partial mesh architecture,” says Salman. “And we’ve developed some business cases in that sense that can show cost reductions of 40, 50 and even 60 percent, depending on the case you’re looking at. And that’s done just by reducing the number of connections and being able to provide protected paths over a mesh-type architecture. Obviously, once you do that, you can reduce maintenance costs because you have fewer microwave systems. And now you’ve got an end-to-end Ethernet service that’s monitored and you’ve got carrier-class OAM on that and you can reduce SLA-compliance costs. You can also do automated provisioning on that type of system, so you can additionally reduce some OpEx costs associated with turning up new circuits. And if you’re looking at a massive deployment such as say, Xohm and Clearwire (News - Alert), with their WiMAX technology, trying to do a nationwide deployment over a span over several years, well, that requires turning up a lot of circuits, and the OpEx savings in such cases can make a huge difference.”

 

“Some of these systems handle 800 megabits per second,” says Salman, “and even with LTE (News - Alert) [Long-Term Evolution] you’re still within the capable range of some microwave backhaul systems. Of course, as I said, there are licensing issues, geography issues and interference issues.”

 

“The third choice for backhaul would be mid-band copper solutions,” say Salman. “I seem more interest in that approach in Europe, such as Ethernet over DSL. On the other hand, almost all of the operators I’ve examined in the U.S. are looking at a hybrid fiber/microwave system. Microwave over relatively short distances brings the traffic to a point close enough to the core so that high-bandwidth fiber can take it from there.”

 

 

“The key from a vendor’s perspective is having what we call a toolkit to apply to backhaul that’s comprehensive, can be managed and, most importantly, enables the transition from TDM and ATM to Ethernet in a smooth, seamless fashion, so they can manage multiple networks as one, very easily.” says Salman. “But also, when they get to Ethernet, they’re getting to what we call True Carrier Ethernet. It’s not just some slapped-together system that just provides, say, pseudowires over Ethernet and basic capabilities. No, I’m talking about being able to give them the same type of operational capability that they had with their SONET and ATM systems, but doing it over Ethernet. That means providing things like automated provisioning. So being able to have a picture of your network, click on two endpoints in the network, define the parameters of the Ethernet service, and then automatically go and configure all of the boxes within the path of that service. We’re talking here about being able to provide primary and secondary protections paths with sub-50 millisecond failover. We accomplish that with a layer 2 solution using Provider Backbone Bridging - Traffic Engineering [PBB-TE] and I think there are many cost advantages with a layer 2 solution, as well as a lot of operational advantages in a layer 2 solution, just from a simplicity perspective.

 

"Then there’s the matter of it being able to provide carrier-class OAM - being able to use devices that incorporate 802.1ag and 802.1ah and 1731 capabilities in them, which are already in all of Ciena’s boxes,” says Salman, “and then being able to manage all of those boxes from a single management suite which we do with our OnCenter platform. We can provide that next-generation Ethernet capability, thanks to our acquisition of WorldWide Packets. We also can provide pseudowire capabilities over Ethernet, which we have in our 3180 and 5060 solution. This is interesting because our 3180 is our access box and it has provides TDM and ATM pseudowires over Ethernet as well as native Ethernet services. The 5060 is the head-end box, a pretty unique device in the marketplace. The 5060 can terminate the pseudowires from a 3180, can take in native Ethernet services, and it can accept native TDM and ATM traffic, aggregating it all at one point — usually at the controller — and provide a single head-end for all of those networks while the operator is going through that transition period from different network types to Ethernet, and it can do with ‘six nines’ [99.9999 percent] reliability. So that’s a pretty impressive box.”

 

“In this business you’ve got to be flexible,” says Salman. “There are two pieces to it. You have to be able to support the transition to Ethernet, but then once you’ve enabled the transition, you’ve got to help the operator manage it. Once the transition is totally complete, they’ve got to be on an Ethernet network that’s carrier grade. And our solution set at Ciena provides all of that."