ravi_patent 28 Report post Posted June 6, 2011 (edited) http://www.thehindubusinessline.com/todays-paper/tp-eworld/article2079720.ece?css=print In India, operators have sunk a lot of money into 3G, and even before they recover money there is already talk of LTE (Long Term Evolution). Is it too early for India to be talking about 4G? I don't think so. It's easy to see the progress from 2G to 3G, and now 4G, as sequential but I don't think that's the way the world works. You deploy certain technologies in certain spectrum. So there is 2.6 gigahertz for LTE spectrum, 2.3 gigahertz for BWA (broadband wireless) — which, for a while in India was called the WiMax spectrum; now it's back to the BWA spectrum, though. So none of these is 3G spectrum, they are not good for HSPA (High Speed Packet Access) either. Nobody is making equipment in that spectrum. It is 2.1 gigahertz where people make HSPA equipment and nobody really makes LTE equipment in that band. And then, 900 megahertz is where people make GSM equipment. So the sequencing is based on which order the regulators give out licences for these spectrums and, of course, when the standards are set out. But it doesn't mean you operate in that way; you actually operate in parallel — to use an analogy, just like a grandfather, father and son lead parallel lives. There is so much of congestion in voice in India that the 3G spectrum will just be consumed by voice. So you will need LTE for data. India is one of the few countries where there is very, very little strategy on spectrum. India has specific issues related to less spectrum availability and rural roll outs. What kind of a role do you think Ericsson can play in such a scenario? Well, one good thing that has happened is that you can now deploy all the technologies simultaneously on single-base stations; the Ericsson 6000 Series base stations allow both GSM and HSPA to be deployed on them … you can buy a base station to deploy 3G and 4G technologies instead of going separately. Then, we also have scenarios for network sharing where you share the radio access network and it branches out to your core network. Those kind of solutions need to be discussed in India, because there are so many operators! No country has so many operators per megahertz. Is the WiMax versus LTE debate over? Well, there will always be a little WiMax, but I have always said that this is not really about technology. We have seen that LTE is a better technology than WiMax but that is not the main thing. The main thing is economies of scale. We assessed that WiMax, at best, will end up with a one per cent market share, and that is why we didn't see it worth investing in. We are now being proved right, and many companies are veering round to the same view. But I wouldn't certainly go out and declare that the debate is over. Why were there no takers for WiMax? Is it a technology, business or ecosystem issue? It was predominantly an ecosystem and scale issue. We didn't see enough benefits on that track. There was 3G, HSPA and then WiMax, but it wasn't attractive enough to derail everything else going on in the industry. Is the network keeping pace with devices that are getting incredibly smarter? I think we are keeping pace. Today, the average person talks 300 minutes on phone a month, which corresponds to 20 megabyte of data. So if all the six billion people on earth talk simultaneously, they are consuming 20 gigabytes. On the other hand, people who use data heavily use around 4-5 gigabytes each. So there is a challenge. What is the future of wireless technology? What comes after LTE and LTE Advanced? If you simplify, for the short range or indoors, we will see the evolution of WiFi. For wide area coverage, we will see the evolution of LTE. It will be a question of finding the shortest way to some antenna, and from the antenna to optic fibre to the cloud. A smart card will keep track of what's happening and charge you per usage, and that money can then be reinvested in building an even better network. But I don't see a 5G technology coming because we have already reached the ‘channel limit' of what you can do. As data usage increases, will fixed wireline networks make a comeback or do you think wireless technologies will be able to meet the demand? Nobody wants to be completely connected with the wireline. I think, in the end, everyone will have only wireless — that is why people have WiFi at home. A revival of the fixed line will depend on regulatory things, but from a technical point of view there is no reason for that to happen. Edited June 6, 2011 by ravi_patent Share this post Link to post Share on other sites
city02 63 Report post Posted June 7, 2011 there is something wrong with the math: Today, the average person talks 300 minutes on phone a month, which corresponds to 20 megabyte of data. So if all the six billion people on earth talk simultaneously, they are consuming 20 gigabytes. 6 billion x 20 Mb = 120,000 TB [is that 120 QB? i don't know the next higher unit] and not 20 GB! 20 gb would be for 1024 people talking together... also, 20 Mb / 300 min = 68 kb/min = 1.1 kb/sec which seems a little low for voice. i would have expected it to be at least 2-3 kb/s instead. is that a typo or does the Ericsson CTO not know middle school arithmetic? Before anyone replies with any helpful clarifications, I'm aware that his point does not change - voice is orders of magnitude more used than data ... Share this post Link to post Share on other sites
sgiitk 32 Report post Posted June 7, 2011 There is so much of congestion in voice in India that the 3G spectrum will just be consumed by voice. very true. Also, the crazy auctions lead to crazy pricing. The state may get its revenue but then ..... Share this post Link to post Share on other sites
anandjm 3 Report post Posted June 7, 2011 (edited) My math agrees that there is a problem with the math. According to my math, if 6 billion people are talking together, they would consume 2.79TB per second. Assume an average of 8kbps (kilobits per second or 1kBps) as the data rate for mobiles. If 6 billion people are talking with each other and assuming discontinuous transmission (DTX) with minimum bitrate in DTX mode, we can assume that that there are 3 billion pairs of people talking. So 6billion/2 * 1 = 3000000000 kbps. This is 2.79TB per second assuming 1MB=1024KB and so on (not the hard disk non-standard of MB, GB and TB). Also according to my math if one person speaks for 300 minutes, the consumption is 17.57MB. Should be around 20MB if two people are talking with each other, making it on average 10MB per person. (This is assuming DTX). 6 billion x 20 Mb = 120,000 TB [is that 120 QB? i don't know the next higher unit] and not 20 GB! 20 gb would be for 1024 people talking together... also, 20 Mb / 300 min = 68 kb/min = 1.1 kb/sec which seems a little low for voice. i would have expected it to be at least 2-3 kb/s instead. is that a typo or does the Ericsson CTO not know middle school arithmetic? Before anyone replies with any helpful clarifications, I'm aware that his point does not change - voice is orders of magnitude more used than data ... Edited June 7, 2011 by anandjm Share this post Link to post Share on other sites
