Sunday, August 16, 2009

Interview: Ericsson's Hakan Eriksson on HSPA+, LTE, and Wireless Trends

Video interview with HÃ¥kan
Eriksson, Chief Technology Officer, Ericsson.

1. How fast will you be able to push
HSPA+?  How does this compare to LTE?

"HSPA and LTE are both mobile broadband technologies that have similar capacities in terms of bits per hertz, but LTE can operate with higher bandwidth and thanks to that you can offer higher peak rates. It is also possible to increase the peak rate with HSPA, which is called HSPA+. If you use all codes, you can go up to 14 Mbps. Then you can increase the modulation up to 64 QAM for another 50% boost, giving you 21 Mbps. Then you can employ two antennas at each end -- called MIMO -- and you will double the peak rate again to 42 Mbps. And then you could run multi-carriers and double the peak rate again to over 80 Mbps. So there are many steps you can take to get the peak rate up very high on HSPA. But if you want to go above 100 Mbps -- which is the definition of 4G -- then you will want LTE, which you could bring up to 300 Mbps."

2. Will a broad uptake in HSPA+ subscriptions delay the rollout of LTE?
Why do you need LTE?

"No, I don't think we will see a strong interdependency between HSPA and LTE in that way because HSPA and LTE will operate in different spectrum bands. You want to have a certain harmony in the world with certain bands in most countries used for HSPA. So 2.6 GHz will be very much an LTE band and 2.1 GHz will remain an HSPA band. Operators will invest in both HSPA and LTE. So why do you need LTE? We will see applications coming that will demand even higher bit rates as we roll out fiber to the home. 2G was almost matching your dial-up. 3G is matching DSL and LTE/4G will come to match your fiber."

3. Will the lack of backhaul capacity be an obstacle to HSPA+ and LTE rollouts?

"Not a big issue if you have prepared for it. If you have fiber, or if you have good mini-LINK or microwave connections, to your sites, that is not a problem. Of course if your sites are CDMA or wideband CDMA and had only T1s for voice backhaul and now you upgrade to HSPA or LTE and get this tremendous capacity in the air interface, then you risk that the backhaul becomes the bottleneck. But if you've planned for that, it shouldn't be a problem. Operators have rolled out HSPA and have catered for fiber or microwave away from the sites, they have had no problems."


4. What are the prospects for LTE voice?  

"We have to remember that LTE was not invented for doing voice again. LTE was invented to do even higher peak rates -- for being more like a "wireless fiber", so to say. When will we have voice there (on LTE)? I think voice will come when you need it for OPEX, by bringing voice and everything together on LTE. The goal is pure IP... a pure packet network. Voice today is still predominantly run on a circuit switched basis in the mobile world because it is very efficient to do it that way. The air interface is a shared medium already, so if you don't talk or don't transmit there is no interference and somebody else can use the spectrum. Since it is a shared medium already, there is nothing to be gained. So in a sense it is already packetized even though there is a circuit-switched core model. So there is no reason to go to voice-over-IP for that reason. Verizon explained its strategy saying it will move its data usage over to LTE but will let voice stay on CDMA. I think that is a wise move to go that way, but of course one day they will realize that they are running two separate networks and that they have built a pure IP backbone, so they will decide to move everything over from an OPEX
point of view. When that will happen is up to Verizon to say. But it will happen. The standard is ready. 3GPP 7 already has a very efficient standard for VoIP but there is not yet a need to deploy that."

5. What role do you see for
Bluetooth, WiFi and WiMAX?

"I am a firm believer in mainstream technologies.. those that achieve economies of scale are winning ones. Bluetooth, WiFi and WiMAX address different radio areas. Bluetooth serves the personal area network linking all the devices you may have in your pockets or briefcase. And there Bluetooth is the winning technology. WiFi is really a cordless technology. It is not really an access technology, like DSL is an access technology. And there WiFi is by far the most entrenched technology and will be there. WiMAX is addressing the wide area, but here it is GSM, HSPA and LTE that are the dominant technologies. That is why WiMAX will have a much tougher time because it is not the mainstream winning technology. We think the 3GPP standards -- HSPA and LTE -- will have about 90% of the market, CDMA will have a little bit less than 10%, and WiMAX will only have about 1% of the wide area market. So there will be 3GPP standards in the wide area, WiFi in the cordless and Bluetooth in the personal area."

6. Are the energy efficiency
improvements for mobile technologies really making a difference for our
carbon footprints?

6. Are the energy
efficiency improvements for mobile technologies really making a difference
for our carbon footprints?

"If you compare the 3G systems of today with those of 2001, we have reduced the power consumption by 80%, so we are much more efficient today. Which of course is important because it has to do with how much carbon dioxide comes out of our industry. The whole IT industry contributes about 2% of the man made carbon dioxide, while the other industries contribute 98%. So there are two things to think about from that fact. One is that we have to do what we can do reduce our 2%, and we are doing that by reducing the carbon dioxide by 80% over the past 8 years. Then, we have to consider how we can use mobile broadband to avoid traveling or traveling in more efficient ways. If you look at the other 98%, our calculations show that we can reduce those by 20% by using IT. So there are two things: first of all, reduce our own 2%, and think about how we can use our technology to reduce the other 98%. As an example, consider that an annual mobile subscription is equivalent to driving your car for about one hour. Have you in the past year avoided at least one hour of car driving by using your mobile phone? Yes, you probably have. You've probably saved ten or twenty hours. There is a big payback this industry can create in carbon dioxide savings. This is good."