From a network perspective, LTE is of course the topic of the Mobile World Congress. This morning I met with Julius Robson of Nortel, who is chairman of the LTE/SAE Trial Initiative, or LSTI for short. I’ve been following the activities of the LSTI for some time now as their aim is not to repeat the mistakes made with UMTS in the past by not properly testing networks and mobile devices with each other before pushing them out into the market.
For the congress first results from the proof of concept phase are becoming available from a lot of vendors and I will report more on the details in the coming weeks. Here are some personal insights I gained during the meeting:
I’ve been a bit sceptical so far of the chances of TDD-LTE (Time Division Duplex) becoming a big technology outside China as most network operators in Europe, Asia and the Americas mostly have FDD (Frequency Divison Duplex) spectrum. To my surprise, it seems there are more than just a few network and mobile vendors who have reported results back to the LSTI on first TDD-LTE tests. They still seem to be a bit behind FDD-LTE but it’s not the tiny trickle I was expecting. Julius told me Nortel has worked on TDD-LTE as well and since everything above the physical layer is pretty much the same in FDD and TDD, only a few people were required to come up with the additional base station software for TDD in quite a short time. Quite a difference to TDD-UMTS which is a completely different game to FDD-UMTS used today.
Also, some mobile operators, for example in Europe, have TDD spectrum in the 2.1 GHz band that they bought together with the FDD spectrum during the UMTS auctions back in the year 2000. So far that invstment lies dormant. So if mobile devices do come out early that support both FDD and TDD than I think there might actually be a chance for TDD to be deployed alongside FDD in major markets outside China. Whether it’s as easy on the mobile side to have FDD and TDD combined in one chipset is another matter and I don’t have a lot of insight into this. Maybe I’ll find out more this afternoon as I’ll talk with some silicon manufacturers then.
In LTE, there are only two radio interface states, ACTIVE and IDLE. The LTE requirements say switching between the two states should take less than 100 ms. Looks like while this sounds ambitious from a HSPA perspective, were the switching time currently is in the range of one second, the goal seems to be reached by vendors even in the early stages. That’s good news as it makes the tradeoff very small between interactivity and power conservation from the mobile device point of view. So it will be interesting in the future to observe if Discontinuous Transmission (DTX) will be used a lot while in ACTIVE state to conserve energy or if the mobile is just set into IDLE state right away.