Take 5 With Michael Howard

CEN Feature (Jun 5 2012) Take Five With

1. 

   Michael Howard, co-founder and principal analyst for Infonetics, does a lot of research about carrier Ethernet, particularly as it pertains to mobile backhaul. Carrier Ethernet News caught up with him recently to ask about what trends he’s seeing in the market these days.

   Q: You’re expecting to see some wireless carriers deploying voice over LTE soon. How will that impact their Carrier Ethernet backhaul needs?

   A: Operators buying backhaul network equipment want to make sure that 2G, 3G and 4G is supported on the equipment they buy now. It’s more important for transport providers than mobile operators because they need to have the broadest offering.

   About 90% of carriers’ backhaul spending is on Ethernet, primarily for HSPA+ and LTE. The backhaul networks are becoming packet and Ethernet ready – and VOLTE is voice over packet.

   Another indication that backhaul networks are ready to support VOLTE is that most operators – and we’ve done surveys about this – want transport providers to provide clock transport SLAs [service level agreements]. And those have the same requirements, if not a stricter requirement, than voice. It’s not just a new need sitting out there to transport VOLTE. They need a clock for 2G voice. It doesn’t work without a synchronization clock.

   Q: So what’s driving the need for synchronization is that carriers want to move all their backhaul traffic onto Ethernet? In the past they might have put data traffic on Ethernet but kept voice on TDM, correct?

   A: Yes. It could be 2G or 3G. It needs clock and timing. The protocol that keeps track of voice requires timing.

   Q: Are carriers looking to use SyncE or 1588v2 for timing?

   A: Some carriers like 1588v2 because you can add software at each end to implement 1588 but for SyncE you need hardware. Of course you need a physical clock, but these already exist for their Sonet/ SDH networks.

   SyncE is essentially TDM married to packet so you have the same characteristics of a T1 or E1 circuit – and some carriers see that as an advantage. Some operators are saying, “We have a TDM circuit running here. We can get in lockstep with the master clock.” But SyncE doesn’t give you the time of day you would get with 1588v2. So they want both.

   Q: Are carrier Ethernet equipment manufacturers ready to support both?

   A: Most have 1588v2 and most but not all have SyncE – and if they don’t have SyncE, they will get left out of a lot of RFPs [requests for proposal].

   Q: The Metro Ethernet Forum earlier this year came out with profiles for a variety of different services, with recommendations about how to set specific parameters such as CoS to support each service. Are you seeing carriers embracing these profiles?

   A: They like the idea but most of the operators already have everything set up and they’re not going to move to something new.

   A different item that the MEF implemented is the ability to have two separate streams with different CoS [class of service] on the same backhaul connection. That’s what multiCOS is. The basic driver for it is you want to put Internet traffic from the handset to the Internet on more of a best-effort type of SLA, while timing, clock synchronization and voice need a higher level.

   Q: There’s been a lot of activity lately involving service providers using WiFi to offload traffic from their wireless networks in high-traffic areas. Will that bring new Carrier Ethernet requirements? Will it share infrastructure with existing cellular backhaul?

   A: When you have WiFi hotspots, many times it’s on DSL or sometimes E1 or T1. The route to the mobile packet core is likely to take a different route than mobile backhaul for the macro cellular network. And for WiFi there is a different gateway function that traffic has to go through before it gets to the mobile packet core.

   There were a slew of announcements recently about small cells with side-by-side WiFi. That’s a little more complicated. The backhaul from outdoor small cells with both WiFi and cellular will be a short distance – a short hop to the nearest macro cellsite, not to the base station itself. Those outdoor small cell backhaul links to the macrosite will hop on the existing macro backhaul network.

   For the small cell backhaul to the macrocell site, there may have to be some sort of microwave, millimeter wave or licensed non-line of sight aggregation box, a device sitting at the macro cellsite that talks to each outdoor small cell below. Each microwave box will be talking to four to five picocells and take their traffic over Ethernet to the existing cellsite gateway or router, which is connected to the backhaul network.

   Q: Is there a new product opportunity there?

   A: There is a new product opportunity. There are some constrictions that allow for specialized equipment to be built.

   The environment of small cells is that you only need to cover a short distance – maybe half a kilometer to a kilometer at the most. You would be aggregating four to five or eight microwave connections to a single box so you need aerials pointing in different directions.

   Operators are looking at millimeter wave, licensed non line-of-sight and regular microwave. Millimeter wave and licensed non line-of-sight are the two newcomers that make a lot of sense in this situation.

   Q: Are you seeing any other interesting trends involving Carrier Ethernet?

   A: We keep getting signals from operators and equipment manufacturers that the move away from TDM to Ethernet is quickening. The classic drivers of Ethernet are finally picking up steam in a lot of parts of the world. Small and medium businesses are saying T1s or multiple T1s are too limiting because each time they upgrade bandwidth they have to buy new equipment and wait for installation and testing by the operator – and service providers are saying the same thing. 

   Bookmark or Share this article

Recent Comments

Login to comment