Talk:3GPP Long Term Evolution

From Wikipedia, the free encyclopedia

Contents

[edit] Objectives

What I'm trying to do is give readers a flavour of what the LTE group seems to be thinking. I must admit though I neither have the time to plough through everything referenced on the LTE page, nor necessarily the right knowledge to be 100% sure I'm understanding it correctly. We could ideally do with more information on the various layers of the system, for example the protocols over the AIPN and how they will work. Squiggleslash 19:18, 4 January 2007 (UTC)

[edit] Removal of IMS link

I've removed the link to IP Multimedia System on the article page. While the topic itself is relevent, the the article concerned needs serious clean up, for reasons I've given on the associated talk: page, and in its present form is worse than useless. There seems to be general agreement on that point, but no consensus on how to proceed largely, from what I can see, because a group of people (for whatever reason) want to turn standardized SIP into some kind of political issue.

I revisited it today and it's worse than ever. Much of the page is essentially vandalism. In my view, it needs to be deleted. If this were an external link, it would probably never have been linked to, but it seemed worth linking to at the time because it was the "official" Wikipedia page on the subject. What I may do over the next few days is add a (small, summary) section to this article explaining what it is and how it relates to the LTE initiative. --Squiggleslash 14:31, 1 March 2007 (UTC)

[edit] Slices of spectrum

Too low a spectrum slice, and the efficiency and maximum bandwidth per handset suffers. Too high a spectrum slice, and there are deployment issues for operators short on spectrum.

This sentence keeps being misinterpreted, so let me try to explain the issues. If someone wants to reword it in a way that keeps the same meaning then I'm all ears (especially as hopefully that'll prevent it from being changed again.)

  • Mobile phone operators are allocated blocks of spectrum. For example, they may have 10MHz of spectrum in a particular area.
  • Each mobile phone system they operate (GSM, D-AMPS, CDMA2000, UMTS, etc) needs to have a part of that spectrum allocated to it exclusively. For example, if I were running a GSM and a D-AMPS network (which until recently was a common configuration in the US), I might allocate a 6MHz slice of that 10MHz to GSM and 4MHz to D-AMPS. I can't allocate the 10MHz to both systems at once.
  • Now, for many GSM operators in the US, they want to operate both GSM and UMTS networks. UMTS uses 5MHz slices. So if the GSM operator has 10MHz of spectrum, they can run both GSM and UMTS by allocating 5MHz of it to GSM, and 5MHz to UMTS.
  • This is a problem because many GSM operators actually only have 5MHz of spectrum. UMTS requires the whole of that spectrum be allocated to it, or none at all.

3G UMTS uses fixed sizes of slice (set at 5MHz) because of implementation issues regarding CDMA. It's big so that UMTS can maximize the amount of efficiency and bandwidth (data throughput) per handset. However, this gives US operators the above headache.

So those developing LTE and considering CDMA for LTE have an issue - they can either reduce the efficiency and capabilities of the system they're developing, but ensure it can co-exist with legacy standards by reducing the spectrum slice requirement, or they can make it almost impossible for operators to deploy in countries where spectrum is at a premium.

(LTE does away with CDMA, using an air interface technology that doesn't require fixed sizes. This means that operators are free to decide how big a slice of spectrum to allocate to LTE in each market, making it easier for it to co-exist with legacy standards.)

Now, if you're thinking of changing the above sentence, please read the entire paragraph it's in and consider the context and what it means within that context. It might be better for you to rewrite the paragraph than change the sentence - the sentence means what it says, and it's hard to see how it can be reworded and still carry the same meaning. --Squiggleslash (talk) 19:06, 29 November 2007 (UTC)

Squiggleslash - My sincere apologies; I had indeed read the paragraph, and did not (do not) see the difference in meaning between my revision and the original (which is understandable, since this is far from my area of expertise). The sentences are not grammatically correct as written and are tremendously ambiguous to a non-expert, which is most likely why they are so often changed. I would very much like to work with you to fix these issues, since you clearly understand the nuances better. I think I understand the issues at hand, as explained above - my confusion stems from the phrase "too low/high a spectrum slice." What is changed by rearranging this to "When the spectrum slice used is too low/high"? Would a different verb be preferable? Since "low/high" are referring to the size of the slice (as seems to be the case from your explanation) rather than its location, might adjectives like "large/small" or "wide/narrow" be more suitable? Thanks for your patience. St3vo (talk) 02:46, 30 November 2007 (UTC)
How about "If the size of the spectrum slice chosen is too small/large" - does that adequately preserve meaning? St3vo (talk) 03:00, 30 November 2007 (UTC)
The problem is the word "chosen" completely throws the meaning of the sentence, implying that this is about decisions the operators make rather than the characteristics of the technology and the consequences of how it is specified. As far as "wide" and "narrow" (or small/large) go instead of "low/high", those seem reasonable changes and I think would help readability. --Squiggleslash 15:16, 4 December 2007 (UTC)
I see what you mean re: chosen. The changes you made are a definite improvement on readability for layfolk such as myself. I don't mean to nitpick, but for grammatical purposes, there needs to be a verb in the clause - what about "if the spectrum slice is too narrow..."? St3vo 15:53, 4 December 2007 (UTC)

01:18, 8 May 2008 (UTC)01:18, 8 May 2008 (UTC)~~ —Preceding unsigned comment added by 98.134.170.213 (talk)

re: "Slices of spectrum" - The correct term would be "bandwidth". Earlier in the discussion, the term "bandwidth" was misused when "throughput speed" would probably be a better choice. 98.134.170.213 (talk) 01:22, 8 May 2008 (UTC)arri_guy@yahoo.com

[edit] Sub-5ms latency for small IP packets. On-line gaming?

Pardon my ignorance, but doesn't this sound like latency dependent real-time applications would finally become feasible on mobile internet? I'm referring to things such as real on-line gaming (the World of Warcraft kind, not Uno), which are currently unthinkable on mobile network infrastructure.84.254.9.136 (talk) 12:19, 17 May 2008 (UTC)