The Fundamentals of CDMA and GSM
April 5, 2012
This term paper is going to compare CDMA and GSM cellular technologies. It will first introduce both technologies with a brief terminology and history overview. The paper will then delve into each technology individually. The first focus will be on GSM. This work will explain the fundamentals behind how GSM works and will assess its strengths and weaknesses. Following suit, the paper will do the same with CDMA. Finally, the two technologies will be compared, contrasted, and summarized to complete the paper.
The terms CDMA and GSM are often used vaguely to reference different methods for transmitting digital signals over the air. CDMA stands for Code Division Multiple Access. As the name suggests, it is a method for dividing radio frequency (RF) waves such that multiple users have access to the same wireless spectrum. A very similar term to know is TDMA – Time Division Multiple Access. As one can see, the fundamental difference between the two strategies is that TDMA uses time to divide and control wireless spectrum access whereas CDMA uses pseudo-random codes.
GSM stands for Global System for Mobile communications. In a nutshell, GSM is an overlay of TDMA; it uses TDMA as its core technology, but has evolved to add features that help reduce the potential problems inherent in the system (Sutherland, 2006). Keeping up with terminology, a Base Transceiver Station (BTS) is another name for a cellular tower or cell site. This paper will use the words “handset” to mean any cellular mobile device and “call” to refer to any type of invoked communication between a handset and a BTS.
GSM has been commercially available for longer than CDMA. In the early 1980s, Europe was seeing handsets grow rapidly, and it was apparent that the old analog technology wouldn’t be able to support the rising demand forever. To make things worse, each country in Europe had been developing its own wireless systems, and, unlike in America, nothing worked together. This meant traveling between countries in Europe required that people must purchase dissimilar handsets and plans from different wireless carriers in each country. There were no roaming capabilities like in America.
The Conference of European Posts and Telegraphs (CEPT) formed a study group in 1982 named Groupe Spécial Mobile – this is where the GSM acronym began before eventually being reassigned to mean Global System for Mobile communications. The group’s goal was to build a European-wide mobile system that allowed people to cross country borders more easily in what was becoming a more unified Europe. The European Telecommunication Standards Institute (ETSI) took responsibility of GSM seven years later in 1989 and published phase I of the GSM specifications in 1990. Mobile carriers adopted the standards and began rolling out the technology in mid-1991. In just two short years GSM claimed 36 distinct networks spanning 22 nations and several continents (Scourias, 1997).
CDMA’s roots can be traced back to World War II. In 1940, a Hollywood actress by the name of Hedy Lamarr and co-inventor George Antheil patented an idea known as frequency-hopping – later renamed to frequency-hopping spread-spectrum (FHSS). With the war looming, their goal was to control torpedoes using multiple radio frequencies by generating random patterns to frequency hop between channels. The inspiration was to conquer radio jamming technologies. While it was an intriguing concept, frequency hopping was ultimately dismissed by the United States Navy as unfeasible (Sutherland, 2006).
Nothing more became of the idea for another seventeen years until engineers at the Sylvania Electron Systems Division (SESD) espoused it in 1957. The SESD engineers in Buffalo, New York had a similar goal as the patent holders, but instead of guiding torpedoes, they were working on a way to secure radio communications. The Lamarr-Antheil patent expired 1960, and SESD was able to successfully implement their revised technology during 1962 Cuban Missile Crisis (Sutherland, 2006).
The United States was pleased, and the FHSS technique became an integral part of government communication security for the next two decades. It was adopted, tweaked, and used exclusively by the military until the mid 1980s when the Government eventually declassified CDMA. Qualcomm, which was then just a start-up, patented CDMA and is still the current owner of the technology. CDMA was first tested for commercial use by wireless carriers in 1991, but didn’t officially launch until 1995 where it made its commercial debut in Hong Kong (Sutherland, 2006).
GSM operates at the 850 MHz and 1900 MHz bands with channel spacing of 200 kHz. It divides access to this spectrum into units of time called burst periods. The BTS and handset sync together and transmit both the call and the overhead controlling data during that handset’s assigned burst period. If there are several handsets communicating with a BTS, each handset takes turns transmitting and receiving one after the other in a loop fashion. While a handset is waiting to transmit, there is nothing happening and therefore no noise. Many people used to mistakenly hear the dead space in the off periods and think the call was dropped. To alleviate this confusion, many wireless carriers now insert what is known as “Comfort Noise” during the time between bursts. Comfort Noise is synthesized sound designed to mimic the structure and volume of the actual background clamor. This manufactured static assures the caller that the connection is still intact (Punter, 2004).
Each GSM burst period is approximately 15/26 ms long. A frame consists of 8 bursts – meaning each frame holds data from 8 different calls. The 200 kHz traffic channels can carry up to 26 frames at a time, giving a total timeframe of 120 ms per cycle (National Instruments, 2009). Because the frequency bands are divided into predefined time slots, determining GSM’s total capacity is a straightforward process. Spectral efficiency or capacity is rated as “calls per BTS per MHz.” Doing a little math, one will find that GSM has a spectral efficiency of 5, meaning it can handle 5 calls per MHz of bandwidth (Punter, 2004). Having predefined time slots is pleasant because it confidently ensures a predictable quality of service by guaranteeing a minimum level of bandwidth; the drawback is that the capacity cannot scale to accommodate spikes in demand.
GSM handsets use Subscriber Identity Module (SIM) cards to identify themselves on a network. SIM cards have minuscule storage, but enough to hold data such as a person’s phone number, wireless subscriber plan data, and even contacts. SIM cards are also portable. If a person wants to travel out of their home network coverage area into a different network, she can simply replace her current SIM with one from the network she plans on entering. Because of this, a person can use the same phone on any wireless network so long as she has a SIM from that wireless carrier (Natarajan, 2006).
When a GSM handset is on the move, it will eventually travel out of range of the BTS it is currently transmitting to, and the handset will have to lock onto whichever available tower has the strongest signal. Like legacy analog systems, this handoff is performed by a switch internal to each BTS. Unlike the aged analog systems, however, the handoff isn’t done without prior knowledge of the device location. The handset continuously monitors the signal strength of every tower within range and reports back to the switch without the user knowing. The switch then takes this information and decides which BTS is best suited to catch the call when the time comes. GSM does what is known as a “hard handoff.” That is to say that when time comes for the handset to switch towers, the handset is first disconnected from the original BTS then “caught” by whichever tower was chosen by the switch to be the best candidate. Unfortunately, that means the call is dropped if the candidate BTS has reached capacity and has no more time slots available (Punter, 2004).
Networks that run GSM are capable of using FHSS, although not to the same effect as CDMA. Frequency hopping literally means that a call is able to switch between channels very quickly and at random. This decreases the likelihood of RF problems on any given channel causing errors that could potentially corrupt the call data. This also helps keep the workload between channels equal which helps curb co-channel interference, multipath interference, and fading. While frequency-hopping is possible with GSM, it is not mandated, so not all wireless providers implement it (Punter, 2004).
CDMA operates in the 850 MHz, 1700 MHz, and 1900 MHz frequency bands; and it is quite a bit more advanced and complicated than GSM. As mentioned earlier, it uses codes to divide access to the spectrum. The key thing to understand here is that CDMA allows handsets to broadcast continuously and at whatever time they want; every handset in range of a BTS has access to the entire spectrum all of the time. The calls are essentially layered, one on top of the other, and unpacked at the back end by the cellular tower using the randomly generated codes to identify the and manage the call (Herrman, 2010).
CDMA always implements FHSS technology, allowing CDMA signals to occupy a much greater bandwidth than necessary for transmission; this helps the signal avoid interference and jamming while allowing for multiple users. CDMA networks broadcast all data three times over separate channels to ensure quality. For a bit error to occur, the redundantly transmitted bits must be damaged to the extent that they cannot be compared and repaired. Unlike GSM, the call transmission and controlling overhead data are spread separately of each other. Since the data that controls the call is smaller than the actual call data, it always reaches the handset first; the handset then uses this data to extract and demodulate its call (Hendry).
Because CDMA uses codes instead of time to modulate call data, it makes great use of wireless spectrum. It is much harder to calculate efficiency in CDMA than it is in GSM, however, because in CDMA there is no maximum number of handsets allowed; a CDMA tower will scale to accommodate an infinite number of users. CDMA can theoretically allow an astounding 45 users per MHz under prime conditions, but even using pessimistic numbers one will find CDMA to be twice as efficient as GSM (Punter, 2004).
It is important to note the two types of data links and codes used to channelize handsets. The forward link is the stream of data traveling from the BTS to the handset. The reverse link is the stream traveling from the handset to the BTS. Walsh codes are used on the forward link. They provide the means to uniquely identify each handset. The Walsh codes are used by the handsets to recover their call data. The handset synchronizes to the Walsh code to recover the data. On the other side of the coin, the reverse link uses Pseudorandom Noise (PN) codes that are generated by the handset. PN codes appear to be random, but they are not. CDMA is so secure because the PN codes are able to yield 4.4 trillion different combinations. These PN codes are what the cellular towers use to interleave the calls they handle (Qualcomm Learning Center).
The forward link consists of four channel types: Pilot, Sync, Paging, and Traffic. The BTS constantly transmits on the Pilot and Sync channels. The Pilot channels are used by the handsets to acquire the system and adjust the power needed to transmit on the reverse link. The Sync channels provide the mobile handsets with the system time and BTS identification numbers. Once a handset has synchronized with a BTS, the Sync channels are ignored, and the handset starts listening on the Paging channels. These channels are used to transmit the overhead information as well as send commands and pages to the handset. These pages are used to indicate things like incoming calls. Finally, the forward link uses the Traffic channels to serve calls (Qualcomm Learning Center).
The reverse link is much simpler since it is instantiated by the handset and not the BTS. It only consists of two channel types: Access and Traffic. Access channels are used to register the handset with the network, originate calls from the handset, respond to pages and commands from the BTS, and transmit overhead messages back to the BTS. Access channels are only used when a handset is not assigned to a Traffic channel. Like in the forward link, the Traffic channels in the reverse link are used to propagate the call data (Qualcomm Learning Center).
Contrary to GSM SIM cards, CDMA uses what is called a Mobile Equipment Identifier, or MEID to store user information. This MEID is unique for every handset, but it is not a physical piece of equipment that a user can simply swap between phones. MEIDs are stored in a carrier database, so changing phones requires the carrier to reassign the MEID; one advantage here is that a MEID cannot be lost like a SIM card (Natajaran, 2006).
When a CDMA handset is on the move and has to switch cell towers, the transition is handled differently than in GSM networks. Transferring a call between towers in a CDMA network is called a “soft handoff.” This is because in CDMA networks, the network looks at multiple alternate BTS sites that could be candidates to handle the call. The handset is then able to pick and choose between which BTS it wishes to use at its own discretion, giving the phone control of switching. When a CDMA user moves from one BTS to another, the handset actually broadcasts to each BTS and combines the received data. The call is never disconnected, but is instead served by both cellular towers during the transition. By allowing the handset to choose which BTS it wants to transmit to, the need for a central switch like in GSM networks is eliminated (Punter, 2004).
When comparing CDMA and GSM networks, there are several things to consider. Each technology has its own unique advantages and disadvantages. For people who travel overseas, GSM is the network of choice. Since GSM was commercially available two years earlier than CDMA, and because it is a free global standard instead of proprietary technology, it managed to overwhelm the market share. Worldwide, roughly 80% of carriers support GSM networks (Herrman, 2010).
GSM networks also have an advantage over CDMA in the fact that they allow simultaneous transmission of voice and data, so if a user wants to browse the internet while making a phone call, he can. GSM networks also offer a predictable and reliable call quality since the number of users each BTS can manage has a fixed maximum number of time slots. CDMA, conversely, has no fixed number of allowed subscribers, and therefore the quality of service fluctuates with the BTS load. GSM handsets offer a longer talk time than CDMA – this is due to the burst nature of GSM transmissions as compared to the nonstop broadcasting nature of CDMA (National Instruments, 2009).
Unfavorably for GSM, hard handoff techniques cause more dropped calls when moving between cell towers. On top of that, CDMA cellular towers are able to disburse signals over further distances than GSM networks. GSM has a maximum cell site range of 35 km; CDMA has no maximum distance, making it more suitable for rural areas, but the cell site shrinks as the subscriber number increases. CDMA is also more secure than GSM by nature of the code division design. CDMA accommodates at least twice as many users per MHz than GSM. To add to this, CDMA gives a higher-quality transmission even under weaker signal strengths (National Instruments, 2009).
Herrman, John (2010, September 14). Giz Explains: What’s the Difference between GSM and CDMA? Retrieved from http://gizmodo.com/5637136/giz-explains-gsm-vs-cdma
Sutherland, Ed (2006, July 25). GSM Vs CDMA. Retrieved from http://www.m-indya.com/cdma-vs-gsm.php
Scourias, John (1997, October 14). GSM History. Retrieved from http://www.privateline.com/mt_gsmhistory/
Punter, Steve (2004, April 15). CDMA vs TDMA. Retrieved from http://www.arcx.com/sites/cdmavstdma.htm
Natarajan, Sabareeshwar (2006). CDMA versus TDMA. Retrieved from http://www.infosecwriters.com/text_resources/pdf/TDMA_versus_CDMA_SNatarajan.pdf
Hendry, Michael. Introduction to CDMA. Retrieved from http://www.bee.net/mhendry/vrml/library/cdma/cdma.htm
National Instruments (2008, Feb 27). IS-95 (CDMA) and GSM (TDMA). Retrieved from http://zone.ni.com/devzone/cda/tut/p/id/7107
Qualcomm Learning Center. Introduction to CDMA. Retrieved from http://www.cdmauniversity.com/prodtech/cdma/training/cdma25/m5/m5p01.html