Edge vs Gprs

Robo_Mike

I am currently connected to the internet through the vodafone huawei e220 modem with full gprs coverage( the best internet we could get due to living in the back of beyonds) My max download speed is about 48kbps.

However on the o2 network we have full edge coverage. Would this be much of an improvement over our current connection?

I am thinking of switching to o2 for a year(min contract) on a student deal for €13.05 a month for 10GB, versus our current €30 a month with vodafone.

We will be covered by 3 with the NBS within the year( they are looking at the possibility of putting a mast on our land as it would cover a large area).

Basically is it worth switching in terms of speed???

drunkmonkey

Edge is quicker but it won't be much, maybe a 100k at best, i'd swith if it's going to save you money while your wating for 3 to come on line...

watty

GPRS is about 70Kbps max, ever.

Edge is upto 240k today and could be upto 1.2Mbps simply by O2 Software upgrade on their mast.

jor el

watty wrote: »

simply by O2 Software upgrade on their mast.

Have you any idea if they're planning to do this? Seen any press releases or statements at all? 'twood be nice if they did, for when I drop out of HSDPA cover.

watty

Not by O2. But by their vendors.

It makes me shake my head all this talk of swapping GSM bands to 3G. While the voice quality on GSM is a bit poorer than 3G, EDGE is more spectrum efficient than HSDPA.

With a 14.4Mbps 3G mast the speed per person can drop to less than 200kbps for 7 out of 10 users when 10 are downloading at once.
Basic EDGE can do 1 to 10 users at 240k without slowing, 120k for 20 users.
upgraded EDGE can do 1 to 10 users at 1200k without slowing, 600k per user for 20 users.
EDGE2 can do 1 to 10 users at 2Mbps without slowing, 1000kps each for 20 users.

I-HSPA 14.4Mbps mast can be doing under 118kbps for 75% of 20 users.

EDGE "up to" speed is limited to about 1/10th capacity. 3G "upto" speed is sexy sounding at 14.4Mbps, but that's marketing hype. You need to be the sole user of the mast and practically able to count the rivets (under 200m range often) to get that speed.

3G/HSDPA uses CDMA. So when there are 20 to 40 users the cell can shrink to 1/2 coverage or less. GSM/GPRS/EDGE uses FDM/TDM and has virtually no cell breathe.

GSM was a technical decision. 3G was obsolete from day1 as it is a wideband (5MHz) version of the old 1.25MHz USA CDMA-One from same era as beginning of GSM.

CDMA-one tech was cheap compared to more complex GSM.

With LTE the political stupidity of 3G will be gone again. However it needs new modems, new bases, new handsets, new spectrum as it is not at all compatible with anything else (except in backoffice equipment that support the Mobile Systems).

The 2.4Mbps EDGE2 likely needs a FW update or new modems as it is 3GPP rel 7, Most Modems/Handsets claim 3GPP rel 5 (6 will be skipped). Some may work anyway.

Robo_Mike

So basically your saying that the way forward is EDGE2 as 3G just doesnt have the capability and LTE is going to be too expensive.......(In the short-term anyway) If this is so why arent the networks moving on it instead of focusing on 3G???

watty

The vendors tell the operators what to do. Operators employ very few if any real Communications Engineers, who don't make policy.

We will probably get 3G on 900/1800 (w-cdma900, w-cdma1800) instead of Edge2, in the next year or so as that makes more money for the vendors if the regulators agree.

Then some LTE some places next year (trials this year). Big LTE rollouts in 5 years maybe. Give the operators time to pay off the 3G investment.

Mobile WiMax may get a couple more sells and then die as LTE rolls out. Fixed Wimax will replace Fixed Wireless Broadband gear as it wears out or new areas licenced.

3G "sounds good" Ericsson did a demo of "upto" 42Mbps HSPA+. EDGE2 can't really ever do more than 2Mbps. In reality with 20 users on a sector with one channel you might get 0.15Mbps. The 42Mbps is a parlour trick. It uses TWO channels. Almost nobody has enough 3G spectrum to give two channels per sector. It uses the sort of QAM only relible in your living room WiFi or Cable. The speed could be still 3.6Mbps (peak, only a solitary user) at 1/2 distance between mast and cell edge. Which is over 75% the cell area.

It used to be said there are Lies, Dammed Lies and Statistics. Now we have Mobile Data speeds too.

With headline demos like Ericsson's 42Mbps HSPA+ on 3G, who is going to invest in "better" EDGE2?

rebel.ranter

watty wrote: »

The vendors tell the operators what to do.

watty wrote: »

Not sure I agree with that. I would say it is a two way street.

watty wrote: »

Operators employ very few if any real Communications Engineers, who don't make policy.

Definitely don't agree with that!

watty wrote: »

We will probably get 3G on 900/1800 (w-cdma900, w-cdma1800) instead of Edge2, in the next year or so as that makes more money for the vendors if the regulators agree.

Agree with the first part, not sure 3G makes more money, what it does do is protect their GSM networks, the current (albeit dwindling) cash cows.

watty wrote: »

Then some LTE some places next year (trials this year). Big LTE rollouts in 5 years maybe. Give the operators time to pay off the 3G investment.

Mobile WiMax may get a couple more sells and then die as LTE rolls out. Fixed Wimax will replace Fixed Wireless Broadband gear as it wears out or new areas licenced.

LTE will probably be affected by this global economic downturn (giving you the 5 year lag), you're right operators will not want to invest more cash without any clear revenue source. 3G was over hyped & the revenue potential exaggerated. The 5 year lag might ensure a better end product.

watty wrote: »

3G "sounds good" ?

Kind of agree, still think mobile internet has a place as a complimentary product to fixed BB.

watty

EDGE/HSDPA/LTE are complementary to real Fixed Broadband, yes.
You need them for MOBILE or Nomadic access

3G on 900MHz/1800MHz will make almost no difference to operators and when Capex is included will make a loss for most. It will however be gravy for the Vendors of base stations.

Originally Posted by watty View Post
Operators employ very few if any real Communications Engineers, who don't make policy.

Definitely don't agree with that!

So why have we 3G at all instead of a better system. An OFDM/TDM hybrid is up to x4 better than CDMA and could have been the standard and based on GSM structure, 1MHz, 2MHz, 4MHz, 6MHz and 8MHz channels more sense than 5MHz too.
80 carriers per MHz, GSM has 5 carriers per MHz, by dividing GSM FDM channels by 16 approx. Mix of OFDMA and GSM TDM. 1MHz up to 3Mbps, 4MHz would support up 12Mbps, 6MHz 18Mbps and 8MHz 24Mbps. It would have been planing ahead for US & Europe analogue shutdown and use of WhiteSpace in 6MHz & 8MHz channels. GSM is often allocated in 4MHz blocks. 1MHz would allow N=4 reuse on 4MHz GSM blocks. 3G is a sick joke. Basically a 5Mhz version of the 1.25MHz original brain dead CDMA that was obsoleted by GSM.

I can think of at least one Mobile operator that appears to have no engineering expertise in Mobile or ISP matters.

I'm not talking about Site Planning guys. I'm talking about REAL engineers that could design a protocol, an RF to baseband front end or ignore vendor hype and analyse sector throughput metrics from first principles. Or design a base station without a vendor list of parts.

rebel.ranter

watty wrote: »

3G on 900MHz/1800MHz will make almost no difference to operators and when Capex is included will make a loss for most. It will however be gravy for the Vendors of base stations.

I believe it will be used more as a capacity tactic than indoor penetration tool. Or perhaps it will be used

watty wrote: »

So why have we 3G at all instead of a better system..

Money & influence. Why Bluray? Same reason, biggest players dominate the next step.

watty wrote: »

I can think of at least one Mobile operator that appears to have no engineering expertise in Mobile or ISP matters.

I'm not talking about Site Planning guys. I'm talking about REAL engineers that could design a protocol, an RF to baseband front end or ignore vendor hype and analyse sector throughput metrics from first principles. Or design a base station without a vendor list of parts.

Would be interested to hear which operator you have formed that opinion on

Also most guys that are involved in cell planning in my experience come from a wide variety of backgrounds & experience. A proper approach to cell planning & optimisation requires REAL engineers to get the best from any plan. That would include detailed analysis of metrics, right down to raw counter level sometimes. It would also include Layer 3 analysis for troubleshooting, etc. I believe/assume it would be standard practice for any one working on the access networks side of things to be very familiar with GSM/UMTS Protocol. Companies independent of the vendors like APIS (among others) provide non specific vendor training on items such as these. I would imagine most operators would ensure that their people are fully versed on these topics.
Also anyone who cannot identify the required generic components of a base station and their function should not be involved in mobile networks.

Finally a successful mobile operator will also have very capable consultants or engineers looking at various aspects of the network, roadmaps of new technologies, developing future strategies, etc. If they don't have these types of people in place they will not last long.