Chapter 11 Mobile Telephony Mikael Olofsson 2004 Devices for wire-less two-way communication between individual users have been around for several tens of years, and they are usually called walkie-talkies. Those are mobile devices, since the communication is done by radio, but they are not telephones. The walkie-talkie is tuned to a specific radio channel, and all walkie-talkies using the same radio channel within some distance from each other will be able to communicate, and all of them will hear all the others. The communication is half-duplex, which means that a user cannot receive while transmitting, also called press-to-talk. No separate controlling unit is used. The fundamental difference between walkie-talkies and mobile phones is that the a user is able to contact a specific phone, and within some limits regardless of where that phone is. The communication is or is at least supposed to be private, and users are supposed to be able to move around while communicating. This is achieved by having several base stations, each managing the common radio channel within its cell, by using some multiple access method. The base stations are then connected in a network that keeps track of where each phone is, and assigns subchannels to the phones when necessary. 11.1 History Today, we normally distinguish between three generations of mobile phone systems, starting with systems from late 1970 s. There are several more or less differrent examples of mobile phone systems in each generation. The difference between the generations primarily has to do with the fundamental technology used (analog or digital) and what services they provide or are primarily designed for. 141
142 Chapter 11. Mobile Telephony Early Systems The first rudimental examples of mobile phone systems were introduced in the 1940 s, and they can be viewed as evolutions of walkie-talkies. Most of those early systems were based on frequency modulated half-duplex communication, and were not cellular systems. Instead there was one base station, normally located in a major city, and the phones worked within the coverage of that base station. The transmitted power was huge compared to todays mobile phone systems. In 1960, a full-duplex system was introduced by AT&T with direct-dialling. First Generation The concept of cellular systems were described in theory in 1947 by Bell Labs, but the electronics technology was not mature enough at that time. Cellular systems were first proposed by AT&T in 1968, and they were developed during the 1970 s. Experimental cellular systems were set up in the mid 1970 s, and that led to the first generation mobile phone systems around 1980. The first generation mobile phone systems are more or less translations of traditional wire bound telephone systems into mobile systems. Those systems are purely analog, and they are only designed for voice communication. Early phones for those systems were sometimes called car phones, since they were too big to carry around. A few examples of such systems are the following. NTT Nippon Telephone and Telegraph Corporation. Introduced in Tokyo, Japan, 1979. The system was very costly, and gained very little popularity. TACS Total Access Communication System A system, based on FDMA, that was introduced in Great Britain in the early 1980 s. It is around in different flavours. First, also in Great Britain, the ETACS (E for Extended) was introduced in 1985. In Japan, the JTACS (J for Japanese) and NTACS (N for Narrowband) systems was introduced in 1991 after the market was deregulated. They soon became a strong competitors to the NTT system. AMPS Advanced Mobile Phone System. This is an American system based using frequency modulation and FDMA. In 1979 the first trials were done using this system, and it was commercially introduced in 1983. There is a narrowband version of AMPS aswell, called NAMPS.
11.1. History 143 NMT Nordic Mobile Telephone (network) This system was introduced in the nordic countries in 1981 and 1982. It uses FDMA, on a frequency band around 450 MHz, and is often called NMT-450. In the remote areas of the northernmost parts of those countries, this system is still the most reliable, with best coverage. As the system grew, more bandwidth was needed. In 1986, NMT-900 was introduced, using a frequency band asound 900 MHz. Second Generation The second generation mobile phone systems are digital systems that are primarily designed for voice communication. Speech coding is used to pack the sampled data, and error control coding as well as digital modulation techniqes are used to enhance the communication quality. A few examples are the following. USDC United States Digital Cellular An American system based on TDMA, introduced 1991. It is known under several names, e.g. NADS (North America Digital Cellular), TDMA (from its access method, and IS-54 or IS-136 (standard numbers). GSM Global System for Mobile telecommunication Originally the acronym was interpreted as Groupe Spéciale Mobile, which was the name of the gruop that developed the system. The system uses frequency hopping, combined with TDMA. It was introduced in 1990, and soon was in use all over Europe. The first systems used frequency bands around 900 MHz, an on some locations around 850 MHz. Today, the system is used world-wide, and bands arond 1800 MHz and 1900 MHz are also used. CDMA Code Division Multiple Access There are several systems called CDMA, all using the access method DS-CDMA. Later, digital services were added to the systems, and those digital services are often referred to as generation 2.5. The systems called CDMA above belong to this category. Third Generation Third generation systems are being built and has just started to go commersial as this is written. UMTS (Universal Mobile Telecom System) or simply 3G is the name of those systems. They are digital systems, just as the second generation systems, but they are designed for general digital services. Voice communication is only one of those services. The design emphasis is on flexibility. There are two variants using different access methods, and they are the following.
144 Chapter 11. Mobile Telephony W-CDMA Wideband CDMA This is the European version. W-CDMA is DS-CDMA with fairly high bandwidth. This system was introduced in various European countries during 2003. TD-SCDMA Time Division Synchronous CDMA This is the Chinese version, based on TDD (Time Division Duplex) and DS-CDMA. TDD simply means that two-way communication is achieved by time division. 11.2 Cellular layout Geographically, the radio channel is divided into cells, with a base station in the center. Ideally, those cells are hexagonal. The phones within a certain cell all communicate through the corresponding base station. A subchannel that is used in one cell cannot be reused in nearby cells, since that would cause distorsion and/or cross-talk. This is called co-channel interference. By not reusing a subchannel in nearby cells, this interference is reduced. The furter away a subchannel is reused, the smaller the interference is. Therefore, cells are gathered together in clusters, and all available subchannels are divided between the cells in a cluster. Those clusters are then repeated. This results in a trade-off between the number of available subchannels per cell, and the delivered communication quality. Let D be the smallest distance between cells using the same subchannel, and let R be the radius of the cell. Then the co-channel reuse ratio, Q, is defined as Q = D/R. Let N be the number of cells in a cluster. Then it is possible to show that Q = 3N holds. Furthermore, it is possible to show that N must be given by N = m 2 + mn + n 2 for some non-negative integers m and n, provided that the cells are hexagonal. 11.3 An Example GSM As an example of a mobile phone system, let us consider GSM, the dominating system in many parts of the world.
11.3. An Example GSM 145 System Architecture As said before, the geography is divided into cells, each managed by a base station. Several base stations in an area are connected to a base station controller. That is where the actual control is done. The base station controller decides on which subchannels to assign to the users. The base station controllers are connected to the fixed telephone network via a mobile switch. Communication Details The voice signal is sampled with a sampling frequency of 8 khz and quantized to 13 bits. Those samples are handled in groups of 160 samples, which corresponds to an interval of 20 ms. Each such group is packed to 244 bits using speech encoding. This is a distorting source coding technique that is based on the statistics of spoken language. These 244 bits are then protected by a convolution code, an error control code that expands the 244 information bits to 456 encoded bits. Those 456 bits are then split into eight segments of 57 bits each. During a 20 ms interval there are four TDMA slots per user. Each 57-bit segment is then transmitted in a TDMA slot together with a segment from the previous or the next 20 ms interval. During normal operation, a TDMA slot consists of three tailbits, followed by one segment, a known 26 bit pilot sequence, the other segment, three more tailbits, and a guard interval corresponding to eight bits. The pilot sequence is used on the receiving end to estimate the channel, which in essence determines the receiver filter to use. The bits are sent bit-by-bit using BPSK. Handover A mobile phone system is based on mobility. The users are allowed to move while using the phone. Thus, a user may very well move from one cell to another. Then the communication needs to be handed over from one base station to another. In the GSM system this is done by interaction between the mobile phone and the base station controller via the base stations. The base stations send out signals that are detected and measured by the phone. The starting point is that the phone communicates in cell A via its base station A. As long as the measured signal from base station A is stronger than the corresponding signals from surrounding base stations, all is well. Assume that the user moves towards the neighbouring cell B. Then the signal strength from cell B increases, an eventually base station B s signals strength exceeds that of base station A. When that happens, the phone sends out a request via base station A for hand-over to cell B. The base station controller decides if a hand-over can be done, depending on if there are available subchannels. When that decision is made, a command is sent to the phone to switch to a given subchannel in cell B.
146 Chapter 11. Mobile Telephony Digital services Using the original GSM system, it is possible to get data communication with a maximum rate of 9.6 kb/s. However, the users have gotten used to data rates up to 56 kb/s using ordinary modems for the fixed telephone channel, and with ADSL that increased to data rates of at least 0.5 Mb/s. Therefore, methods for faster digital rates have been developed for GSM. The methods are based on using more than one subchannel on demand. The actual data rate delivered depends on the over-all load of the base station. The following two methods are around. GPRS General Packet Radio Service The data rate can be up to 115 kb/s using this method. This method uses BPSK, exactly as for voice communication. EDGE Enhanced Data rates for GSM Evolution EDGE can deliver data rates up to 384 kb/s. The digital modulation technique used is 8-PSK. The main reason for this increase compared to GPRS is the choice of modulation technique. With 8-PSK, three bits are transmitted where BPSK transmits one bit. Those methods are often viewed as a step inbetween GSM and UMTS. They are often referred to as generation 2.5.