Today the mobile phone has evolved from a simple voice device, to a fully-functional personal organiser conveniently containing important information including contacts, calendar appointments and notes. It can be used to access the Internet while on the move, especially for email, and is also a portable radio, music player, camera and photograph album.
The same technology (the Internet-connectivity part at any rate) can also be used in a data card or USB device in conjunction with a laptop computer, or can even be embedded within the PC itself. It is also used in other, newer devices, such as portable wireless access points and cellular terminals.
So what do all the acronyms surrounding this subject stand for? What is the difference between HSDPA and EDGE?
In this article I will look at the development of mobile data technologies from their first inception to the present day and beyond.

In the beginning…

Before networks became “cellular”, people who needed mobile communications could use a vehicle-mounted radio telephone services by one large central antenna in a major city. This communication was not “personal”: you simply broadcast on a particular frequency and anyone listening on that same frequency could hear you. These networks typically covered approximately 50 miles.

A cellular network employs a number of radio masts and can theoretically service an infinite area. The first generation, or 1G, of mobile networks was entirely analogue and only serviced a small number of users with low growth projections. Networks were usually state-owned, coordination between countries was minimal and roaming between countries with the same handset was impossible.
Several different technologies emerged:

• The United States opted for AMPS, the Advanced Mobile Phone System
• The UK implemented TACS, the Total Access Communications System
• The Scandinavian countries implemented NMTS, the Nordic Mobile Telephone System

The systems all worked on different radio frequencies, devices were bulky with limited battery life and coverage was, at best, adequate.
These networks could transmit data, but demand for such services was virtually non-existent.

Second Generation Systems (2G)

2G mobile networks were launched in Europe in 1992. The motivation in Europe was to adopt a single technology for the EC, replacing the patchwork of incompatible analogue systems with one digital system that would enable European citizens to make and receive calls anywhere in Europe.

The technology that was adopted was GSM, the Global System for Mobile telecommunication.

Digital technology provided improved call quality, security and data capability at 9.6kbps!

GSM networks were also rolled out in the US, however deployment was slow due to the prevalence of the AMPS system. It was cheaper for operators in the US to upgrade the existing infrastructure from AMPS to D-AMPS (Digital Advanced Mobile Phone System), also known as TDMA, or Time Division Multiple Access, which I will look at later.

Why is digital better than analogue?

Both analogue and digital systems use analogue radio waves to transmit information from the mobile device to the nearest radio mast (or cell tower). However, digital systems used the analogue wave to send binary information in 1s and 0s rather than an analogue radio signal.
Analogue signals have a tendency to lose their integrity because of “noise” or interference. The amount of noise picked up during transmission is known as the “signal to noise ratio”. When an analogue signal is passed through an amplifier to boost the signal, the noise that has been picked up along the way is also amplified. The further a signal travels, the more noise it will pick up until the original signal is indistinguishable from the noise. This is known as “signal attenuation”, and is the reason why long-distance telephone calls used to sound so much worse than local calls.
Whereas analogue signals can represent a whole range of values, digital signals can only represent two: 1 or 0. Instead of amplifiers, digital signals are regenerated using a repeater. The incoming signal is examined, the repeater determines whether the value is 1 or 0 and transmits a new, perfect, signal. Therefore any noise that is picked up along the way is removed.

GSM was originally licensed for use on the 900 MHz band, but was soon expanded to include 1800 MHz as demand grew. The US implemented GSM networks using 1900 MHz instead. Thus you often hear of phones being either “dual-band” or “tri-band”. South American and Asian networks have subsequently deployed GSM networks on 850MHz, and modern phones are thus referred to as being “quad-band”.

GSM was not readily adopted in the US and other nations outside Europe, large operators selecting an alternate technology instead: CDMA, which I will look at later.

GSM network infrastructure

A GSM network is made up of a number of key components: the mobile device (this could be a phone, datacard, USB dongle, embedded module, etc) is referred to as the mobile station (MS).
Radio signals are sent to and received from the nearest cell tower, or base transceiver station (BTS).
BTS units are normally located every few miles, or closer together in densely populated areas, or where the natural geography of the landscape requires it.
A handful of BTS units will be controlled by a base station controller (BSC), which monitors the performance of each BTS and handles the process of moving a subscriber from one tower to another (this process is known as “handover”).
The BSC units are in turn connected to the network’s mobile switching center (MSC), which is the gateway between the mobile network and the public switched telephone network (PSTN), or landline network.
Connected to the MSC is a large computer known as the home location register (HLR), which tracks and stores the location of each subscriber so that incoming calls can be routed to the correct BTS that the user is connected to at that time. The HLR also contains records on what services the subscriber is eligible for, such as MMS, Internet access, Blackberry, etc.
Linked to the HLR is the equipment identity register (EIR), which holds records on SIM numbers and IMEI numbers. If you’re not familiar with these terms, these will also be looked at later. The CDR database, or Calling Data Record, tracks how much use the subscriber makes of the network so that bills can be sent accordingly.

GSM network frequency usage

When a network is said to operate on 900 MHz, it actually uses different frequency bands for the uplink and downlink connections between the MS and the BTS. 905 MHz to 915 MHz is used for the uplink, and 950 MHz to 960 MHz does the downlink. This 10 MHz band is then further divided into 2 lots of 5 MHz, with each 5 MHz “passband” being allocated to a single network operator. Similar divisions are also made of the 1800 MHz passband. Therefore, for a long time, 4 network operators existed in the UK: 2 operating at 900 MHz and 2 at 1800 MHz. This has now changed with the licensing by the government of the frequency used by 3G, which I will look at later, and the emergence of so-called ‘virtual’ network operators, such as Tesco and Virgin mobile, who “piggyback” on the infrastructure of existing network operators.
You can find out information about the frequencies used by the different UK operators here:
http://www.gsmworld.com/roaming/gsminfo/cou_gb.shtml

The network operator then subdivides their 5 MHz allocation into 25 sub-channels of 200 KHz each. Each of these channels is known as a “carrier”. 24 are available for customer use, 1 is reserved for network monitoring and signalling. Within this small frequency range, a specific frequency will be used for the actual transfer of data, and the “surrounding” frequency is used as “spacing” to prevent adjacent carriers from interfering with each other.

No two adjacent cell towers use the same carrier frequencies. This enables operators to cover a large area with only a relatively small amount of allocated bandwidth.

This technique of frequency usage is known as Frequency Division Multiple Access (FDMA).

To further maximise the number of subscribers who can use the network simultaneously, each carrier is further subdivided into 8 “timeslots”, with one timeslot assigned to each user. Each user then transmits for just 0.5 milliseconds before use of the timeslot is passed onto the next user and so on. This process happens so quickly that no loss in quality is noticed by the user. This technique is known as Time Division Multiple Access (TDMA).

The capacity of a cell is therefore based on the number of carriers used multiplied by the number of timeslots.

One weakness of the TDMA system is that data must be transmitted at all times to keep all 8 timeslots in sync with each other. Therefore even if no speech is being transmitted, “packing” data must be broadcast instead.

So how does it work?

All mobile devices have a unique code associated with them. This is a 15-digit number called an IMEI number (Individual Mobile Equipment Identifier) and can normally be located on a sticker under the battery.
The SIM card, or Subscriber Identity Module, also has a unique number associated with it, called an IMSI number (Individual Mobile Subscriber Identifier).
When a mobile device is turned on, it will search on the control carrier for a System Identity Code (SIC), a unique 5-digit identifier assigned to each mobile network operator by the local governing body. If it cannot find any codes at all, it will display “no service” on the display, or something similar.
When it does find an SIC, it will compare it to the number programmed on the SIM card. If they match, the device will then transmit a registration request containing its IMEI number. The mobile network will then update its location database with the details of the IMEI, IMSI and the BTS and send a message to the device over the control carrier letting it know what frequencies to use for uplink and downlink.
As the device moves towards the edge of a cell, the BTS will determine that the signal strength is diminishing. Simultaneously, the BTS toward which the device is moving, will determine that the signal strength is increasing. The two BTS units coordinate with each other through the BSC until a signal is sent to the device over the control channel telling it to change frequency, thus “handing over” from one cell to another.

CDMA

CDMA stands for Code Division Multiple Access and is another 2G cellular technology, not employed in Europe, but used extensively in the US.
CDMA was developed by the US military as early as the 1940s as a robust transmission system that could withstand jamming attempts. It was deployed in cellular networks in 1993.
It is a spread spectrum technology. Simply put, this technology still divides the available frequency into “carriers”, but instead of using specific individual carriers for each base station, the entire bandwidth is used and data is sent in small packets across all of it, “hopping” from one frequency to another pseudo-randomly (the transmitter and receiver agree on the algorithm to use before communication begins, to a potential eavesdropper the sequence appears random). Rather than assigning users a ‘timeslot’ to identify them, individual users are assigned a unique code to identify them and their data stream is sent over the network simultaneously with other users’ data-streams. This allows for a more efficient use of the available frequency spectrum as there is no need to transmit “blank” data when there is no speech to broadcast. CDMA also allows for data communication at up to 14.4 Kbps.

2.5G Systems

As demand for data services grew, “bolt-ons” to the existing GSM infrastructure were deployed. The term 2.5 G is used as they were effectively stop-gap solutions on the road to 3G.
With conventional GSM, data could be sent and received at 9.6 Kbps. This was a limitation of the network infrastructure and did not depend on which operator you used, or what device you were using.
Internet connections were all “dial-up”, which required firstly that you have an account with an ISP (Internet Service Provider). To connect to the Internet you would need to dial a PSTN telephone number from your computer and then wait for up to 30 seconds for the connection to be established as the modems went through the handshaking process.
Once connected, a “circuit” between you and the ISP was opened, and left open until you disconnected. This was known as Circuit-Switched Data (CSD).
This method of connecting to the Internet was slow and also expensive as the user was billed for the length of the call placed to the ISP, even though for much of that time they would not be sending and receiving any data. It also tied up the timeslot being used for the length of time that that user was connected: not an effective use of the network’s preciously limited resources.
One of the temporary measures deployed was the introduction of High Speed Circuit Switched Data (HSCSD). This was only implemented by Orange in the UK. This involved altering the encoding mechanism used, squeezing 14.4 Kbps into one timeslot instead of 9.6 Kbps, and allowing one user to have 2 timeslots simultaneously so that they could effectively connect at 28.8 Kbps.
The other measure was the introduction of the General Packet Radio Service, or GPRS.

GPRS

GPRS uses the same physical network infrastructure as GSM, that is why it is referred to as a 2.5G technology: it is still based on TDMA and users are still assigned timeslots. However, instead of dialling into an ISP and then establishing a circuit with that provider to send and receive data, with GPRS the method of sending and receiving data is different. The network operator effectively becomes the ISP, negating the need for an account with an external provider. When the (GPRS-capable) device is turned on, it is registered on the data network immediately and is able to send and receive data instantly in “packets”. It is therefore said to be “packet-switched” rather than circuit-switched.
GPRS connections are said to be “always-on”. Data can be sent and received at any time, with data transmission being paused should a voice call need to be placed.
This has a number of advantages both for the user and the operator:

• There is no 30-second delay before a connection to the Internet is established
• The user is only billed for the data that they actually send and receive, regardless of how long they remain connected to the Internet
• As there is no “dead time” during which the user is not sending and receiving data, the network resources are used more efficiently, allowing for more concurrent subscribers.
• Data speeds are faster

GPRS networks required additional hardware to be installed within the GSM infrastructure: each BTS is upgraded to include a Packet Control Unit (PCU) to handle the routing of data packets. Each PCU sends the data packets it receives to a Serving GPRS Support Node (SGSN), which performs a similar function to that performed by the BTS for voice communication. It is effectively a router that examines each packet and determines its destination and routes it accordingly.
A Gateway GPRS Support Node (GGSN) is the router that sits between the mobile network and the wider Internet. This is also known as the Access Point Node (APN) and will record details on data transferred for billing purposes.
Each network operator will have its own APNs, some used for general Internet browsing, some dedicated purely to the sending and receiving of picture messages, etc. It is also possible for corporate customers to have their own hardware installed on the network that is dedicated to routing data from the mobile network to their own internal network. This is called a “Private APN”.
Therefore, when configuring a GPRS connection on a mobile device, rather than entering a telephone number to be dialled, you simply need to enter the name of the APN to be “attached” to.
GPRS uses a number of different means of encoding data for transmission across the radio link, which allows for faster data speeds: up to 13.4 Kbps per timeslot. Users can also be allocated up to 4 timeslots, allowing for up to 53.6 Kbps.
It should be noted, however, that the amount of timeslots allocated to data users will be controlled by the network operator: at times when there is a lot of voice traffic on the network, fewer timeslots will be made available for data usage than at times when the voice traffic is low.

3G

It is important to remember that the term “3G” does not refer to a specific technology, it refers rather to a type of service. As 2G refers to both GSM and CDMA, so does 3G refer to a handful of different technologies. Therefore, saying that you are “connecting over 3G” is not technically accurate.
3G technologies allow for the live streaming of media from the Internet, video conferencing, etc. There are several different technologies within this umbrella.
EDGE (Enhanced Data rates for GSM Evolution) is largely considered to be a 2.75G technology. It is also referred to as E-GPRS.
EDGE is a bolt-on enhancement to the existing GPRS network. It requires no hardware changes to the network, rather a software update to the BTS and BSC (and the mobile device) that enables a more advanced encoding algorithm that allows for a maximum of 129.6 Kbps (using 4 timeslots). The greater data rates are made possible by employing a technique known as Quadrature Amplitude Modulation (QAM).
The physics behind this technology is beyond the scope of this article. Simply put, data can be sent over a carrier wave by altering its behaviour to represent a 1 or a 0. The physical properties of a wave that can be altered are either frequency, amplitude or phase (or a combination of some or all of these). Amplitude is the extent of vibration of a wave, measured in terms of displacement from the horizontal axis. If a carrier wave has a normal amplitude of x, then anything less than x could be said to represent 0, and anything greater than x could be said to represent 1.
Altering the amplitude of a wave to represent binary data is known as Amplitude Modulation. This requires extremely accurate signalling and processing capabilities both in the sending device and the receiver. Quadrature Amplitude Modulation works by varying the amplitude by up to 4 distinct amounts, each amount representing either 00, 01, 10 or 11, therefore allowing for a greater amount of data to be transmitted with each wave cycle. When combined with similar alterations to a wave’s phase, yet more data can be sent per cycle.
Each wave cycle is known as a baud. When you consider that the waves we’re talking about oscillate at up to 18,000,000,000 times a second (1800 MHz), you start you see how complicated the physics is!

UMTS is the Universal Mobile Telecommunications System. This is a theoretical umbrella term used to describe any technology that constitutes a global mobile network. In reality, the term UMTS is only used to describe, and has become synonymous with, W-CDMA, or Wideband Code Division Multiple Access.
W-CDMA uses the same underlying technology as CDMA, however utilises a different frequency: 2100 MHz. the important thing is that all W-CDMA networks that have been deployed globally all use the same frequency, from the US to Europe to Japan. It can therefore, be said to be a truly global mobile technology.
Handsets that can operate on both GSM / GPRS and UMTS are said to be “dual mode”.
WCDMA networks can support data speeds up to 1.8Mbps, thanks to the greater frequency used (and hence more wave cycles), and further improvements to the modulation scheme.
HSPA is the High Speed Packet Access service, comprised of HSDPA (High Speed Downlink Packet Access) and HSUPA (High Speed Uplink Packet Access). HSPA is an upgrade to the WCDMA technology that allows for data rates of 1.8, 3.6, 7.2 and 14.4 Mbps.
HSPA achieves these higher data rates by employing 16-QAM (16 Quadrature Amplitude Modulation). As discussed above, altering the amplitude of a wave can be used to represent a 1 or a 0 per wave cycle, or baud. The higher amount of processing power contained in mobile devices today means that more precise variations of this amplitude can be achieved, allowing for more bits to be represented by each specific variation. 16 QAM allows for each wave cycle to represent 16 possible bit combinations:

0000
0001
0010
0100
1000
1001
1010
etc

The improved data rates are also achieved thanks to improvements in the control and signalling technology that the mobile device and the BTS use to synchronise the transmission of data and the “frequency hopping” algorithms.

The Future

HSPA Evolved is hoped to provide up to 42 Mbps in its first release. This technology will use antenna arrays supporting MIMO (Multiple In Multiple Out) packet transmission, effectively allowing a mobile device to simultaneously send and receive multiple steams of data to the network.
HSPA LTE (Long Term Evolution) is hoped to provide up to 200 Mbps for the downlink channel and up to 100 Mbps for the uplink channel by employing OFDMA technology (Orthogonal Frequency Division Multiple Access).
Again, the physics behind this technology is beyond the scope of this article, but essentially OFDM (Orthogonal Frequency Division Multiplexing) involves using the minimum ‘spacing’ between different frequencies possible to prevent them interfering with each other.
As was seen above with CDMA technology, specific frequency channels within the available bandwidth are identified, and the data is sent across all of it simultaneously, with individual data-streams being identified by a specific code. OFDM reduces the “spacing” used between the frequencies, freeing up more of the available bandwidth for use.
This requires very precise signalling, controlling and processing capabilities on the part of the sending and receiving equipment.
When combined with CDMA technology, allocating individual codes to individual users to identify their data stream and sending all data across the available bandwidth simultaneously, OFDMA allows for even greater data rates and even more efficient use of the network resources.
This technology is in fact already in use in 802.11 WiFi equipment, providing up to 54 Mbps data rates in the 802.11g incarnation, and is used in the 802.16 WIMAX standard, which can deliver data rates up to 200 Mbps.