GSM 1900Mhz Vs. CDMA 1900Mhz

06-10-2002, 8:48 AM

Can anybody tell me which type of wireless system, GSM 1900, or CDMA 1900, requires more towers in order to provide good service/signal levels?

Is it true that GSM needs more towers than CDMA? I seem to see more Voicestream towers around Minneapolis/St Paul than I do Sprint icon

PCS towers. Voicestream's seem to me to be spaced closer together in some areas than Sprint icon

's.

Anybody know if this is true, and why?

Brian

**bobolito** · 06-10-2002, 11:24 AM

A lot of Voicestream towers were installed in North New Jersey in the past two years, much more than SPCS. Also, the existing Voicestream sites were upgraded to handle more capacity. I think this has to do with the aggressive expansion of GSM more than anything else so they are preparing for more subscribers.

06-11-2002, 6:43 AM

The frequency efficiency of CDMA is better than the efficiency of TDMA/GSM. Therefore, you can get more simultaneous CDMA users into the same bandwidth than you can TDMA users. You theoretically need fewer CDMA towers to serve an area than you would TDMA towers.

Whether this alone has any influence on the number of towers that an individual company builds.......

**Matt** · 06-11-2002, 8:13 AM

It's the spectal efficiency, not the frequency efficiency, that I belive is better for CDMA. You can handle more calls per slice of spectrum. But that's not the frequency (wavelength). Higher frequency means shorter waves and weaker penetration into buildings, forested areas, etc. All things being equal, the type of technology doesn't matter when on the same frequency. (I think [img]i/expressions/face-icon-small-wink.gif[/img] )

**FlyBoy** · 06-12-2002, 5:47 AM

The military spent a LOT of money developing the spread spectrum technology that is used by CDMA. It is specifically designed for better penetration in adverse conditions than non-spread systems. And CDMA is designed for a higher signal to noise ratio than other systems, so it should be able to recover more usable signal under the same circumstance.

With "everything else being equal" -- especially the power icon

of the signal -- I think I agree with the second anonymous poster more than Matt (ignoring the "frequency" vs "spectral" vocabulary problem) since the original question dealt with operation on the same frequencies. But both have valid points! And "everything else" is never equal, so this is a question that can never be answered.

**bobolito** · 06-15-2002, 8:22 AM

No matter what technology is used, signals will reach the same distance. Remember that regardless of the case we are dealing with RF, whether its GSM, CDMA, TDMA or analog. Provided that they operate in the same frequency band and the same atmospheric and geographic conditions are provided, they all have the same coverage range per tower. What makes the difference is that CDMA can handle more calls per tower, you'll notice that their towers have less antennas in a stack and still be able to serve more calls. In reality, a CDMA signal doesn't travel farther than any other. The fact is, under weak signal conditions, a CDMA phone can decode more useable voice than TDMA. If a certain byte of signal was lost, a CDMA phone can actually grab it from another tower and put the signals together into one conversation. That's why CDMA range per tower seems more extended. In the case of TDMA, the carrier has no choice but to put their towers closer together.

**CDANMA** · 06-16-2002, 6:18 PM

FLyboy very good explanation and post..... could not have said it better....

**theyallstink** · 06-16-2002, 9:56 PM

First off, how do you know that all the towers you see going up are in fact Voicestream? Few, if any, carriers mark their towers (for fear of sabatoge amongst other reasons). Also many towers are owned by third parties and space to hang antenna arrays is leased to carriers. If you are making the determination based upon the style of antenna array, thats not a good reference. Its possible that Cingular and AT&T icon

use the same GSM arrays.......

Regarding CDMA vs. TDMA/GSM as far as penetration is concerned, the FACT is that in theory CDMA should have slightly better penetration and what that means is that a CDMA phone has a slightly better chance of ringing and making a call out in very weak signal areas than TDMA/GSM. That does not mean that 1900 CDMA is a stronger signal than 1900 TDMA/GSM. Its just that a well designed CDMA phone should work slightly better in fringe areas.

Regarding CDMA vs. TDMA/GSM towers, CDMA tower location is much more critical than TDMA/GSM due to the fact that ALL CDMA towers contain the same frequencies for a given carrier. If CDMA towers are too close to each other that could create a number of problems, including echo, noise, dropped calls, etc... TDMA/GSM towers in any given area use different frequencies and one will not interfere with the other. Obviously frequencies are repeated, but at a tower far enough away so as not to interfere with another.

GSM is a good thing, but I would recommend you wait until Cingular and/or AT&T icon

have a nationwide network. Voicestreams network is too fragile. Also no GSM phone has AMPS roaming so if you go to the boondocks, you've got no service!!!

**bobolito** · 06-17-2002, 6:48 AM

Coming up next!

We'll see one of our dear posters defending Voicestream....

**Matt** · 06-17-2002, 8:21 AM

<< First off, how do you know that all the towers you see going up are in fact Voicestream? Few, if any, carriers mark their towers (for fear of sabatoge amongst other reasons). Also many towers are owned by third parties and space to hang antenna arrays is leased to carriers. If you are making the determination based upon the style of antenna array, thats not a good reference. Its possible that Cingular and AT&T use the same GSM arrays....... >>

If you have a Nokia phone (not sure which models are OK to use) the Netmonitor program I believe can do a search of available networks, enabling you to determine if a particular carrier is on a tower (At least for GSM I have on idea if it works for others...)

**bobolito** · 06-17-2002, 9:03 AM

The problem is Matt, you can't tell which tower the phone is picking up at a particular moment. You have to assume. Just being in front of the tower and having a full signal doesn't always mean that's the tower.

In my town, I've seen cell sites just a couple of street blocks from each other (they are from different carriers) and whether you place yourself in front of one or the other you get between -51dB and -55dB RSSI. The tallest building here is only 12 stories and that is our "skyscraper" (lol)...so we don't have a shadow effect like in big cities. There are other times where you see one site but you don't see the other because is hidden right behind a tree or a small building and you may think your phone is picking up the one you are looking at.

**Matt** · 06-17-2002, 11:18 AM

<< The problem is Matt, you can't tell which tower the phone is picking up at a particular moment. You have to assume. Just being in front of the tower and having a full signal doesn't always mean that's the tower.
>>

True, but it's probably a fair assumption in most cases. Of course there are exceptions to every rule. But since the carriers aren't going to tell us where their towers are.....it might be worth it to some. Sometimes you can get new tower location info by getting access to your local Gov't planning committe a/o supervisors' meetings. Meeting minutes are online in some places in VA.....that's how I ofund out VS was putting a stealth tower on a shopping center clock-tower near my house....

**mattwhit** · 06-20-2002, 12:33 AM

One thing you are not taking into consideration... carriers use different antennas with different gains and shape their signals... they may put up an antenna with a high gain to cover a stretch of intersate, low gain for highly populated areas. RF engineers work hard to design systems to cover population densities and tough terrain. Also, digital technologies allow for phone to "talk" to as many as 5 towers at the same time...

**IdiOTeQnoLogY** · 06-20-2002, 1:08 PM

from what i know cdma technology allows a phone to use up to 6 towers at a time.......though cdma carriers only use 3 at a time.....and also cdma uses soft handoff technology

06-20-2002, 6:04 PM

Idiot....,

It would be VERY BAD network design to have a single point where a phone would routinely be in contact with 6 towers. Draw the hexagons and see for yourself.

Likewise, almost any well designed network has many, many points where a phone could/would be in contact with 3 towers.

The CDMA protocol has no limit on the number of simulataneous messages that can be received -- although, it would be impractical to allow too many.

At any given time, a single cell phone is under the control of a single tower. It may be possible for many towers to hear the same phone and for the same phone to hear many towers, but if everything is working as intended, that time is short and should only occur as the phone is passing from one cell to another (the "handoff").

It is a terrible waste of bandwidth to have a single phone communicating with so many towers.

Charlie J.

**bobolito** · 06-20-2002, 10:53 PM

In TDMA, the phone communicates with one tower at a time, but the phone can monitor the signal of up to 12 towers simultaneously to decide which one is the best to hand-off depending on signal strength. I've seen as many as 9 towers overlapping in certain areas here although probably half of them have a signal so weak at that point that they are unusable.

06-21-2002, 4:39 AM

This seems to be one of those never ending discussions here -- "How far does a cell signal travel"?

Radio waves travel FOREVER in a straight line from the tower to ..... The earliest cell phone signals have made it past Alpha Centari by now! They weaken as they go, so the real question is the EFFECTIVE range of the signal -- which should not be much further than the next tower..

The comment about 'listening' to multiple towers simulataneously is true, but the phone should only be listening to the control channel. Likewise when it broadcasts its periodic 'query', that message can be heard by all cell towers reasonably closeby. Indeed, part of the cellular protocol is for the towers to use the relative signal strength of every phone they are currently hearing to determine where that phone is. The tower with the strongest signal should be the tower which then controls that phone.

There are doubtless hundreds -- if not thousands -- of places in the US where a large number of tower signals overlap. Network design is not perfect, but such situations are not desirable. The more overlap, the more loss of usable bandwidth. The upper stories of urban skyscrapers are notorious for getting ambivalent cell signals since a phone several hundred feet in the air can see such a large area and many towers. This becomes a significant problem (at least for the subscriber) when some of the towers are outside of their subscribed region and they are handled by a tower which makes them 'roam'. (This is partly the reason for the FCC rule against the use of cell phones in airplanes.)

**bobolito** · 06-25-2002, 9:27 AM

Last time I was in the WTC twin towers I noticed how my phone couldn't "decide" which tower to stay with. It kept flipping from one to the other.

**Rasputin** · 08-13-2002, 11:53 AM

theyallstink you have it a bit wrong. TDMA systems are much more critical in freq planinning and cell site design, than CDMA. CDMA is not all on one freq. they use 1.25Mb wide channels, TDMA phones if they here the same freq/time slot the door is open to multipath and interference. CDMA systems are only concerned with pure noise floor issues, they can cancel out other signals to pick up your own signal. CDMA systems can add COWS almost anyway on the same freq, to add coverage for events, it is not like a TDMA system where you need to move freqs around to make it work. CDMA systems are a system designers dream, TDMA systems make for more job security because you are always tweaking and moving freqs around. TDMA systems are build to have little or no overlap on the cell sites. CDMA systems are made to have quite a bit of overlap, this provides for less drops and better handoffs

**bobolito** · 08-13-2002, 8:04 PM

Rasputin, you seem to know a lot about CDMA. Can you tell me why CDMA dual-band phones are only manufactured with retractable antennas? Please don't tell me that it's Verizon icon

's policy or that internal antennas are inneficient because this topic has been debated extensively here and it is proven those are not the reasons.

**Rasputin** · 08-13-2002, 10:00 PM

My only knowledge on this has to do with Bell Atlantic not having Nokia phones past their test, the loaded coil antennas used by the Nokia phones did not offer enough RX sensitivity to work well with CDMA networks. I always thought that CDMA systems need very good Rx sensitivity, I have done measurements of antennas I know that loaded coils dropped the RX sensitivity down about 5-10 DB, the location of that antenna closer to your head probably does not help either. It could also do with the fact that the TX freq varies on the CDMA phone and it is easier to have the 1/4 wave standard. The nokia phones with the antenna built in must lose over 15 DB of sensitivity with you hand and head in the way of the antenna.

My niece's Uncle is a head scientist at Qualcomm, and her Aunt is a lead patent lawyer. I have been given the pleasure to see some of Qualcomm's inner workings.

**CDANMA** · 08-13-2002, 10:41 PM

Your head is known to give you 8 Db of loss....

**Rasputin** · 08-14-2002, 10:46 AM

I would rather have the antenna a little bit away from my head, and at least with the viable tx power icon

of a CDMA phone most of the time you TX under 50% of full power. Any little bit helps

**Rasputin** · 08-14-2002, 10:51 AM

CDMA phones only communicate with more than one tower when the signal strength is necessary, from a loading situation it is not a problem with CDMA phones being on multiple cell sites at once, the capacity is there. Overlapping cellsites and the same freq soft handoffs are a big advantage of CDMA. How come Anonymous posters always seem to give odd information, if you have an opinion put your name on it.

**Rasputin** · 08-14-2002, 10:56 AM

On that same note of CDMA/GSM signal strength, the 1.25 wide channel and the spreading of info across the spectrum is supposed to give you 20% more coverage. I bring this up because that was the reason the I believe it is called Telestra the Aussie telecom and cell system went with CDMA when they got rid of their analog system instead of going with GSM, GSM was not able to offer the coverage of the CDMA system especially in the rural areas where the Analog system was being used. I have not done the tests before myself, but that was always the theory behind using the wide band channels, I have not read up lately on the books of the CDMA Qualcomm gurus, so do not hold me to that theory.

**CDANMA** · 08-14-2002, 11:34 AM

Soft handoffs are great with CDMA but the interferance problem sucks, Signal to noise ratio are sometime hard to control with multiple site right together without having a primary PN. Drop calls with good RSSI but bad EC/IO are a common problem with CDMA, thats why it takes a good RF Engineer to setup a good network.

**bobolito** · 08-14-2002, 12:52 PM

So, how come coil or internal antennas are not a problem with TDMA or GSM phones? I've tested TDMA phones with retractable and internal antennas side by side and the reception (RSSI) is exactly the same. Sounds to me like internal antennas are more efficient since they don't need to be extended to pick up the same signal strength.

**Rasputin** · 08-14-2002, 1:20 PM

I have never done true cellular testing ( in a test enviroment, with my service monitor and test set), but I did have to test 2-way radio portables, and I know that the loaded coil does affect RX sensitivity, where you tend to notice it is in the weak signal areas. There is also a loss on the TX side. I will have to ask some of the qualcomm people I know, or I have a friend who is a Verizon icon

engineer for the true answer.

**Rasputin** · 08-14-2002, 1:27 PM

CDMA systems are not that hard to setup at all, there are some great software programs used by Lucent, and I believe Nortel, that do most of the work for you, all you have to do is to take the noise measurements, at the Rx Antenna at the site, and in specific areas were the cell sites are overlapping. You need an incredible noise floor to cause dropped calls, that is the magic behind the technology to begin with you could not just jam a military signal to stop it CDMA systems can handle quite a high noise floor, I would have to read up in my engineering books to find the magic numbers. The planning for TDMA and GSM is much more complex, you really have to shape you cell sites coverage, and you are always reworking the freqs. Cell sites do not TX/RX in Octagons, you have to be very careful with too much overlap in TDMA/GSM sites.

**bobolito** · 08-14-2002, 10:01 PM

What problems can too much overlapping cause on TDMA/GSM systems? I know areas around here where there's quite some overlapping. RSSI is only in the upper -70's or low -80's when the handoff to a new site occurs. However, call quality is very reliable in those parts of the city.

GSM 1900Mhz Vs. CDMA 1900Mhz

Thread Tools

Search Thread