Cellular Phone Forum / General / GSM / October 2005
Duplexing and SIM
|
|
Thread rating:  |
(Kane) kane@nospam.net - 26 Oct 2005 01:42 GMT I have just spent a few hours reading about the difference between GSM and CDMA and read some things that were a couple years old, so I don't know if they hold true.
1. With a GSM phone I read there is no duplexing, which means when one person is talking, it will cut out the voice of the other (if both are trying to talk at once). True or false?
2. CDMA allows people to talk over each other. True or false?
3. GSM uses 2 watts power versus 2mW used by CDMA. This means GSm phones emit more radiation and the batteries don't last as long. True or false?
4. A GSM phone can use a SIM card. This means if you buy a really expensive GSM phone you like, you can use it even if you change networks (e.g. from Cingular to T-Mobile), as long as you unlock the phone. In this case you can use the new T-Mobile SIM card and keep your same phone. True or false?
5. When using a GSM network on the go (as in a car), as you travel out of the range of one GSM cell base and into the range of another, the call must be dropped and reconnected in the new ase range. True or false? (CDMA has a "soft hand-off" with no loss of signal. On GSM I read it is a "hard hand-off" with loss of signal in the switch.)
Thanks so much. Kane
DevilsPGD - 26 Oct 2005 07:53 GMT >I have just spent a few hours reading about the difference between GSM >and CDMA and read some things that were a couple years old, so I don't [quoted text clipped - 3 lines] >person is talking, it will cut out the voice of the other (if both are >trying to talk at once). True or false? False.
>2. CDMA allows people to talk over each other. True or false? Manors do not, but CDMA does.
>3. GSM uses 2 watts power versus 2mW used by CDMA. This means GSm phones >emit more radiation and the batteries don't last as long. True or false? False.
>4. A GSM phone can use a SIM card. This means if you buy a really >expensive GSM phone you like, you can use it even if you change networks >(e.g. from Cingular to T-Mobile), as long as you unlock the phone. In >this case you can use the new T-Mobile SIM card and keep your same >phone. True or false? True.
>5. When using a GSM network on the go (as in a car), as you travel out >of the range of one GSM cell base and into the range of another, the >call must be dropped and reconnected in the new ase range. True or >false? (CDMA has a "soft hand-off" with no loss of signal. On GSM I read >it is a "hard hand-off" with loss of signal in the switch.) GSM can hand off within a network, but with a few very minor exceptions, not between networks without dropping the call.
 Signature I should have gotten off in crackton -- Lisa
Evan Platt - 26 Oct 2005 16:10 GMT >>3. GSM uses 2 watts power versus 2mW used by CDMA. This means GSm phones >>emit more radiation and the batteries don't last as long. True or false? > >False. But imagine if a GSM phone WAS 2 watts.
Hold a bag of microwave popcorn while talking, and after a 5 minute call, you'd have a nice snack.
 Signature To reply, remove TheObvious from my e-mail address.
matt weber - 27 Oct 2005 05:16 GMT >>>3. GSM uses 2 watts power versus 2mW used by CDMA. This means GSm phones >>>emit more radiation and the batteries don't last as long. True or false? >> >>False. > >But imagine if a GSM phone WAS 2 watts. A handled GSM in the 900Mhz band usually can operate with 2 watts instantaneous power, but the transmitter duty cycle is only 12.5%, so average power is .25 watts even at 2 watts.
>Hold a bag of microwave popcorn while talking, and after a 5 minute >call, you'd have a nice snack. Good luck. I doubt the popcorn would even be warm with 2 watts for 5 minutes.
John S. - 01 Nov 2005 00:33 GMT >>>3. GSM uses 2 watts power versus 2mW used by CDMA. This means GSm phones >>>emit more radiation and the batteries don't last as long. True or false? [quoted text clipped - 5 lines] > Hold a bag of microwave popcorn while talking, and after a 5 minute > call, you'd have a nice snack. Bull sh.t!
matt weber - 27 Oct 2005 05:15 GMT >I have just spent a few hours reading about the difference between GSM >and CDMA and read some things that were a couple years old, so I don't [quoted text clipped - 3 lines] >person is talking, it will cut out the voice of the other (if both are >trying to talk at once). True or false? False, although echo suppressors may make it seem that way some times. Channel timing assignements are offset, so that timing slot 1 is not the same time on the up link as it is on the downlink. I.E. Transmitter and receiver are not actually operating simultaneously. In addition, if there is nothing to transmit, nothing is transmitted. The background noise you hear is actually generated in the phone by a 'noise' descriptor provided by the network.
On an AMPS phone, Duplexing is a problem, because it is tough to have the receiver listen for picowatts, while the transmitter is operating at watts and you are sharing the antenna. D-AMPS/TDMA avoids it because like GSM, the timing slots are offset, so you don't actually have to transmit and recieve at the same time.
>2. CDMA allows people to talk over each other. True or false? FALSE.
>3. GSM uses 2 watts power versus 2mW used by CDMA. This means GSm phones >emit more radiation and the batteries don't last as long. True or false? False, both use power loop control. (however GSM at 1800 and 1900Mhz is limited to 1 watt in hand helds), mobile installations (cars) are actually defined up to 20 watts, but I don't think I have ever seen one with more than 5 watts.
However the power loop control is crucial to the operation of CDMA, whereas the major benefit in GSM is battery life. The down side for CDMA is network capacity is actually defined not by the number of users, but by the total power in the channel. CDMA is the digital analog of synchronous detection. The receiver decides whether this particular piece of data is part of your call by intergrating the output. If you remember your calculus, the intergral of the product of two signals is zero over time unless the two signals are the same frequency. So by applying the spreading code, you are in effect engaged in synchronous detection. Any spreading code other than one the signal was transmitted with, will integrate over time (even very short times) to ZERO, so it is discarded. The problem is that sudden changes in signal level can cause it to be non-zero for brief periods. To be polite, this is NOT good.
As a result, CDMA adjusts output power by +/- 1db 960 times per second, come hell or high water, and because total power in the band determines capacity, CDMA uses as little as it can get away with. In part this works because of the Shannon Theorem. Which says that the capacity of a channel is related to both it's bandwidth, and the signal to noise ratio. If you make the channel wide enough (GSM is 200Khz, CDMA is 1.23Mhz IIRC, CDMA can operate with much lower signal to noise ratios than GSM can because it uses a much wide channel.
The other good news is CDMA has no implicit range limitations. GSM does. There are 8 x 217 time slots per second, and not much time in between. 1 microsecond at the speed of light is 300 meter, or about 1000 feet. You get more than a mile or so away the Base, is nothing is done, you will slide into the next time slot. So GSM incorporates someting called timing advance. The BTS is able to determine the distance from the phone to the BTS based upon the timing it sees on the control channel, and advances the start of transmit cycle in the time slot, The adjustment ranges from 0-63, which allows the signal to reach the BTS within the assigned slot at distances up to 35km.
Beyond that something has to give. If you are even in light aircraft, or high in a building, a GSM phone may be able to see a BTS that is more than 35 km away, but you cannot use it because you can never get messages to the BTS within your assigned time slot, so you can never register. While in Bahrain for example, In the upper floors of a hotel, you can pick up BTS's in Saudi Arabia, but you cannot place calls with them because you are more than 35km away.
In some parts of Rural Australia, the adjacent time slot is 'sacrificed', reducing the channel capacity from 8 to 4, to allow signals to be received in the adjacent time slot, allowing the 35km limit to be exceeded, however I am unaware of this being used anywhere else.
GSM doesn't drop calls unless a priority call forces you out, or you are handed off to a BTS at full capacity. If you try to initiate a call in a cell that is at capacity, the call simply cannot be established. As a result, dropped calls on GSM networks are rare.
In a CDMA network, just changing how you use the phone, or a new call being established can cause other calls to be randomly dropped. A CDMA network sees all calls other than yours as noise. If the background noise level gets too high, your signal can no longer be reliably decoded, and the call is dropped. The other killer can be something called fast Rayleigh fading, which is a type of constructive/destructive interference from multipath reflections. If it happens faster than the power loop can deal with it, the call will be dropped. A fair number of drivers with CDMA in the PCS band and a heavy foot have had this happen, and all they know is the call was dropped. GSM deals with fast rayleigh fading by havig much higher signal to noise ratios to start with.
>4. A GSM phone can use a SIM card. This means if you buy a really >expensive GSM phone you like, you can use it even if you change networks >(e.g. from Cingular to T-Mobile), as long as you unlock the phone. In >this case you can use the new T-Mobile SIM card and keep your same >phone. True or false? True, however the lastest CDMA phones support SIM's as well. It makes International Roaming a lot simpler.
>5. When using a GSM network on the go (as in a car), as you travel out >of the range of one GSM cell base and into the range of another, the >call must be dropped and reconnected in the new ase range. True or >false? (CDMA has a "soft hand-off" with no loss of signal. On GSM I read >it is a "hard hand-off" with loss of signal in the switch.) Wrong, the handoff occurs in a single time slot, about 600 microseconds. YOu will never talk to two BTS's at the same time, but the period of non-contact is very short, and is transparent to the point that you are unlikely to be able to detect it. The Phone doesn't have to find the next BTS, it is told where it is, and simply asks the new BTS to assign it a new channel and time slot. Meanwhile, data is held at both ends for the few milliseconds it takes to do this. The most you are likely to notice is a click.
However if you happen to move into a cell with no available time slots, the call will indeed be dropped because it cannot be handed off. That is a very rare event, because it is bad for business.
>Thanks so much. >Kane (Kane) kane@nospam.net - 27 Oct 2005 07:22 GMT Thanks to everyone for their input. I appreciated it. :)
Kane
(Kane) kane@nospam.net - 27 Oct 2005 07:28 GMT Btw, I continue to read from many sources that CDMA is faster and also does provide far less radiation. This is not the best link to give but it's an example of what I've read many places:
http://www.skytel.mn/cdma.shtml
Just FTR.
John Navas - 27 Oct 2005 08:25 GMT >Btw, I continue to read from many sources that CDMA is faster and also >does provide far less radiation. This is not the best link to give but >it's an example of what I've read many places: > >http://www.skytel.mn/cdma.shtml That's propaganda from the CDMA camp. Not surprisingly, propaganda from the GSM camp has a different perspective. Here's the reality:
* Comparable spectral efficiency (as measured in Erlangs)
Chris Pearson, Executive Vice President, 3G Americas:
Many CDMA operators are currently in the midst of deploying 1XRTT, AN INTERIM STEP TOWARDS 3G that promises to use spectrum more efficiently. Time will tell whether that is the truth but the fact is that, based on best-case data from CDMA vendors, 1XRTT with EVRC handles up to 156 Erlangs per sector. Bearing in mind that GSM with AMR handles 142 Erlangs, it is a great stretch to argue that 1XRTT has a major advantage over GSM. GSM operators also can deploy dynamic frequency and channel allocation (DFCA), which assigns calls to channels based on conditions such as signal and interference. With AMR and DFCA, GSM can handle 170 Erlangs per sector an improvement on 1XRTTs 156. [emphasis added]
In the near future, 1XRTT operators will probably be able to deploy a technology called selective mode vocoder (SMV), which could provide 20% more capacity over EVRC. The catch is that SMV-like methods can be applied to GSM to produce almost identical capacity gains. Thus, while one technology may have slightly higher capacity gains at one point in time, another technology is always preparing to leap-frog over it.
* Comparable data speeds:
CDMA: 1xRTT, EV-DO, EV-DO Rev A GSM: GPRS, EGPRS(EDGE), UMTS, HSDPA
* Comparable radiation (as measure in SAR)
See <http://www.sarshield.com/english/radiationchart.htm>.
* Bottom line: It's not about the technology -- it's about the carrier.
 Signature Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES: John Navas <http://navasgrp.home.att.net/#Cingular>
matt weber - 28 Oct 2005 05:05 GMT >[POSTED TO alt.cellular.gsm - REPLY ON USENET PLEASE] > [quoted text clipped - 6 lines] >That's propaganda from the CDMA camp. Not surprisingly, propaganda from the >GSM camp has a different perspective. Here's the reality: As I pointed out, the advantages mostly depend upon the set of assumptions you make for each case.
matt weber - 28 Oct 2005 05:02 GMT >Btw, I continue to read from many sources that CDMA is faster and also >does provide far less radiation. This is not the best link to give but [quoted text clipped - 3 lines] > >Just FTR. It all depends upon what set of assumptions you use. For example I can probably run 30 calls on single channel, as long as it is voice. Voice doesn't produce a lot of channel occupancy, so the time average transmitted power is low, and that keeps the total energy in the channel low. If those 30 calls suddenly changed to data, and each wanted the 19.2kb available for 100% duty cycle, 20 of those calls would probably be dropped, as it would cause a massive increase in power in the channel, and the signal to noise ratio would no longer be adequate to sustain the calls.
In that same 1.23 Mhz, I can put 41 AMPS calls into that bandwidth, and 123 D-AMPS or IS-136, however even using IS-136 for data transmission, the capacity would be pretty unimpressive all up.
If I use the 1.23 Mhz for GSM, I can put 48 call into it (and then sum) I can have 576000 bps using GPRS.
The main advantage you get with CDMA is that adjacent cells sharing bandwidth are a nuisance with CDMA, they reduce capacity in other cells by increasing the energy in the channel. With AMP/D-AMPS/IS-136 and GSM, if you don't keep significant geographic distances between adjacent cells using the same channels, ugly ugly things happen to calls...
However from my perspect, broad band Celluar phone data ( Hundred of Kilobits per second), no matter what technology you are using is always going to be very expensive relative to Wi-Fi, and Wi-Fi can accomplish almost everything people want to do with 3G, at a tiny fraction of the cost. That is why you see whole cities building Wi-Fi infrastructure.
You can very high channel capacity with CDMA, but if you do, the processing gain (in effect the gain you obtain from very wide bandwidth transmission according to the Shannon Theorem evaporates.
John Navas - 28 Oct 2005 08:02 GMT >If I use the 1.23 Mhz for GSM, I can put 48 call into it (and then >sum) I can have 576000 bps using GPRS. Actually 57,600 with 4 time slots (Class 10) and CS-3 coding. Speed varies by class and coding.
>However from my perspect, broad band Celluar phone data ( Hundred of >Kilobits per second), no matter what technology you are using is >always going to be very expensive relative to Wi-Fi, and Wi-Fi can >accomplish almost everything people want to do with 3G, at a tiny >fraction of the cost. That is why you see whole cities building Wi-Fi >infrastructure. WiMAX makes more sense.
 Signature Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES: John Navas <http://navasgrp.home.att.net/#Cingular>
matt weber - 29 Oct 2005 02:52 GMT >[POSTED TO alt.cellular.gsm - REPLY ON USENET PLEASE] > >>If I use the 1.23 Mhz for GSM, I can put 48 call into it (and then >>sum) I can have 576000 bps using GPRS. Actually it is a mathematical error. The true capacity is 48 14.400. There are 6 channels in the 1.23Mhz band, each channel has 8 time slots, each of which can suport 14,400 with GPRS, so there are only 48 time slots available, making true capacity is about 691,200 bits per second.
John Navas - 31 Oct 2005 12:43 GMT >>>If I use the 1.23 Mhz for GSM, I can put 48 call into it (and then >>>sum) I can have 576000 bps using GPRS. [quoted text clipped - 4 lines] >are only 48 time slots available, making true capacity is about >691,200 bits per second. That's (commonly used) CS3 coding, which actually has a raw speed of 15.6 kbps. Other codings are slower or faster, with differing error correction and range.
User Data Rate CS1 CS2 CS3 CS4 1 Timeslot 9.05 kbps 13.4 kbps 15.6 kbps 21.4 kbps 8 Timeslot 72.4 kbps 107.2 kbps 124.8 kbps 171.2 kbps
Thus the raw GPRS capacity of 6 channels ranges from 434.4 to 1027.2 kpbs, depending on coding. Net GPRS capacity (subtracting overhead) is of course sonewhat lower. In practice, no more than 4 timeslots are used for a given GPRS connection, limiting speeds to half these rates.
The gross bit rate (including overhead) of a GSM channel is 270.833 kbps, or 1625 kbps for 6 channels.
 Signature Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES: John Navas <http://navasgrp.home.att.net/#Cingular>
|
|
|