3 First Generation (1G) Cellular Systems 3.1 Introduction As mentioned in Chapter 1, the first public mobile telephone system, known as Mobile Telephone System (MTS), was introduced in 1946. Although it was considered a big tech- nological breakthrough at that time, it suffered many limitations such as (a) the fact that transceivers were very big and could be carried only by vehicles, (b) inefficient way of spectrum usage and (c) manual call switching. IMTS was an improvement on MTS offering more channels and automatic call switching. However, the era of cellular telephony as we understand it today began with the introduc- tion of the First Generation of cellular systems (1G systems). The major difference between 1G systems and MTS/IMTS was the use of the cellular concept in 1G, which brought about a revolution in the area of mobile telephony. This revolution took a lot of people by surprise, even AT&T who estimated that just 1 million cellular customers would exist by the end of the century, instead of the many hundreds of millions that exist today. The use of the cellular concept greatly improved spectrum usage, for the reasons mentioned in the previous chapters. However, 1G systems are now considered technologi- cally primitive. Nevertheless, this does not change the fact that a significant number of people still use analog cellular phones and an analog cellular infrastructure is found throughout North America and other parts of the world. The moral lesson from this fact is obvious and has been seen in other areas of technology as well – the market does not entirely follow technological developments. However, the reason why 1G systems are considered primitive is due to the fact that they utilize analog signaling for user traffic. This leads to a number of problems: † No use of encryption. The use of analog signaling does not permit efficient encryption schemes. Therefore, 1G systems do not encrypt traffic. Thus, voice calls through a 1G network are subject to easy eavesdropping. Another problem is the fact that, by listening to control channels, users’ identification numbers can be ‘stolen’ and used to place illegal calls, which are charged to the user. † Inferior call qualities. Analog traffic is easily degraded by interference, which results in Wireless Networks. P. Nicopolitidis, M. S. Obaidat, G. I. Papadimitriou and A. S. Pomportsis Copyright ¶ 2003 John Wiley & Sons, Ltd. ISBN: 0-470-84529-5 Wireless Networks. P. Nicopolitidis, M. S. Obaidat, G. I. Papadimitriou and A. S. Pomportsis Copyright ¶ 2003 John Wiley & Sons, Ltd. ISBN: 0-470-84529-5 inferior call quality. Contrary to digital traffic, no coding or error correction is applied in order to combat interference. † Spectrum inefficiency. In analog systems, each RF carrier is dedicated to a single user, regardless of whether the user is active (speaking) or not (idle within the call). This is the reason for the inefficient spectrum usage compared to later generations of cellular systems. 3.1.1 Analog Cellular Systems A number of analog systems have been deployed worldwide [1]. These are briefly described below. 3.1.1.1 United States The first commercial analog system in the United States, known as Advanced Mobile Phone System (AMPS), went operational in 1982 offering only voice transmission. AMPS has been very successful and even today there are many millions of AMPS subscribers in the United States. Furthermore, AMPS has also been deployed in Canada, Central and South America and Australia. AMPS divides the frequency spectrum into several channels, each 30 kHz wide. These channels are either speech or control channels. Speech channels utilize Frequency Modulation (FM), while control channels can use Binary Frequency Shift Keying (BFSK) at a rates of 10 kb/s. Both data messages and frequency tones are used for AMPS control signaling. In order to combat co-channel interference, AMPS uses either (a) a typical frequency reuse plan with a 12-group frequency cluster with omnidirectional antennas or (b) a 7-group cluster with three sectors per cell. The operating frequency of AMPS consists of 2 £ 25 ¼ 50 MHz, which are located in the 824–849 MHz and 869–894 MHz bands. In a certain geographical region, two carriers (service providers) can coexist, with each carrier possessing 25 MHz of the spectrum (either the ‘A’ or ‘B’ band). 3.1.1.2 Europe In European countries, several 1G systems similar to AMPS have been deployed. These include: † Total Access Communications System (TACS) in the United Kingdom, Italy, Spain, Austria and Ireland † Nordic Mobile Telephone (NMT) in several countries † C-450 in Germany and Portugal † Radiocom 2000 in France † Radio Telephone Mobile System (RTMS) in Italy. The most popular systems are TACS and NMT, which together accounted for over 50% of analog cellular subscribers in 1995. As in the case of AMPS, all of the above systems employ FM for voice channels and Frequency Shift Keying (FSK) for control channels. Channels are spaced apart and these spacings are as follows: 25 kHz (TACS, NMT-450, RTMS); 10 kHz (C-450); and 12.5 kHz (NMT-900, Radiocom 2000). All these systems base handover deci- sions on the power received at the Base Station (BS) by the mobile one, except for C-450, which performs handovers based on measurements of round-trip delay. Wireless Networks96 3.1.1.3 Japan In Japan, a total of 56 MHz is allocated to analog cellular systems (860–885/915–940 MHz and 843–846/898–901 MHz). The first Japanese analog cellular system was the Nippon Telephone and Telegraph (NTT) system, which began operation in the Tokyo metropolitan area in 1979. The system utilized 600 duplex channels (spaced 25 kHz apart), which were realized via transmission in the 925–940 MHz (uplink) and 870–885 MHz (downlink) bands. Voice channels were again analog and control channels were 300 bps. In 1988, this rate was increased to 2.4 kbps and the number of channels increased to 2400 via use of frequency interleaving (channel spacing of 6.25 kHz). This new improved system allowed backwards compatibility, thus dual mode terminals were built that could access both the old and new system. Currently, NTT DoCoMo provides nationwide coverage in the 870–885/925–940 MHz bands. In 1987, two new operators were introduced: † ‘IDO’, which operates the NTT high capacity system discussed above, covering the Kanto-Tokaido areas in the 860–863.5/915–918.5 MHz bands. IDO has also introduced NTACS (a variant of the European TACS system) in the 843–846/898–901 MHz and 863.5–867/918.5–922 MHz bands. † DDI Cellular Group, which provides coverage outside the metropolitan areas using the JTACS/NTACS systems (a variant of the European TACS system) in the 860–870/915– 925 MHz and 843–846/898–901 MHz bands. IDO and DDI have agreed to provide nationwide service by allowing roaming between their systems. 3.1.2 Scope of the Chapter The remainder of the chapter examines AMPS and NMT, two representative 1G cellular systems. Although they might seem primitive, they were very successful at the time of their deployment and in some ways have found use as a basis for the development of several 2G systems. An example of this is D-AMPS, which is a 2G system evolving from AMPS; covered in Chapter 4. 3.2 Advanced Mobile Phone System (AMPS) AMPS [1–3] is a representative 1G mobile wireless system developed by Bell Labs in the late 1970s and early 1980s. As mentioned above, it was designed to offer mobile telephone traffic services via a number of 30 kHz channels between the Mobile Stations (MSs) and the BSs of each cell. These 30 kHz channels are used to carry voice traffic. The latter is a 3 kHz signal that is carried over the AMPS channels via analog transmission. 3.2.1 AMPS Frequency Allocations The FCC made the first allocation of bandwidth for AMPS in the late 1970 in order to enable the operation of test systems in the Chicago area. The allocated bandwidth was in the 800 MHz part of the spectrum for a number of reasons: † Limited spectrum was available at lower frequencies, which are primarily occupied either First Generation (1G) Cellular Systems 97 by FM radio or television systems. Lower frequencies are sometimes used by other systems, for example, maritime systems. † Despite the fact that frequencies above 800 MHz are not very densely used, allocation of frequencies in this bands for AMPS was undesirable due to the fact that signals in those bands (e.g. several GHz) are subject to severe attenuation either due to path loss or fading. Such deterioration of signal qualities could not easily be handled at the time AMPS was developed due to the fact that error correction techniques for an analog system like AMPS were in their infancy. † The 800 MHz band was a relatively unused band since few systems utilized it. 3.2.2 AMPS Channels The operating frequency of AMPS consists of 2 £ 25 ¼ 50 MHz, which are located in the 824–849 MHz and 869–894 MHz bands. In a certain geographical region two carriers (service providers) can coexist, with each carrier possessing 25 MHz of the spectrum (either the ‘A’ or ‘B’ band). The transmit and receive channels of each BS are separated by 45 MHz. Both traffic channels for carrying analog voice signals and control channels exist. In a certain geographical area, two operators can exist and a different set of channels is assigned to each operator. The two channel sets, A and B, comprise channels from 1 to 333 and from 334 to 666, respectively. Channels from 313 to 333 and from 334 to 354 are the control channels of bands ‘A’ and ‘B’, respectively. Thus, each operator has 312 voice channels and 21 control channels at its disposal. Each control channel can be associated with a group of voice channels, thus each set of voice channels (either of bands ‘A’ or ‘B’) can be split into groups of 16 channels, each group controlled by a different control channel. As mentioned above, traffic channels (TCs) are 30-kHz analog FM channels used to serve voice traffic. The main traffic channels are the Forward Voice Channel (FVC) and the Reverse Voice Channel (RVC) carrying voice traffic from the BS to the MS and from the MS to the BS, respectively. The network assigns them to the MS upon establishment of termination of a call. Control channels (CCs) carry digital signaling and are used to coordinate medium access of Mobile Stations (MSs). Specifically, each MS that is not involved in a call (idle MS) is locked onto the strongest CC in order to receive control information. The CCs of AMPS are summarized below: † The Forward Control Channel (FOCC). This is a dedicated continuous data stream that is sent from the BS to the MS at 10 kbps. FOCC is a time division multiplexed channel comprising three data streams: (a) streams A and B, which are identified via the least significant bit of the MS’s Mobile Identity Number (described later), with bit 0 identifying stream A and bit 1 identifying stream B and (b) the busy-idle stream, which is used to indicate the status of the RECC (described below). The use of the busy-idle stream reduces the possibilities of collisions on the RECC, as this might be used by more than one MSs. The FOCC is also used by the BS to inform a MS which RVC to use for a newly established call. † The Reverse Control Channel (RECC). This is a dedicated continuous data stream that is sent from the MS to the BS at 10 kbps. Wireless Networks98 AMPS used both data messages and frequency tones for control signaling. The Supervisory Audio Tone (SAT) and the Signaling Tone (ST) are described below. 3.2.2.1 The Supervisory Audio Tone (SAT) SAT is sent on the voice channels and is used in order to ensure link continuity and enable MSs and BSs to possess information on the quality of the link that connects them. Both the BS and the MS send this tone on the FVC and RCC, respectively, and the tone is added prior to the modulation of the voice signal. When a MS is switched on or has roamed under the coverage of a new BS, it tunes to the FOCC and reads a 2-bit field known as the SAT color code (SCC). The value of the SCC informs the MS which SAT to expect. SAT codes are shown in Figure 3.1. SAT determination is performed every 250 ms and the three defined SATs are at the following frequencies: 5.97 kHz, 6 kHz and 6.03 kHz. 3.2.2.2 The Signaling Tone (ST) The ST is used to send four signals: † The ‘request to send’ signal, which is used to allow the user to enter more data on the keypad while engaged in an ongoing conversation, T; † The ‘alert’ signal, which, once the MS has been alerted, is continuously sent on the RVC until the user of the MS answers the call; † The ‘disconnect’ signal, which is sent by the MS over the RVC in order to indicate call termination; † The ‘handoff confirmation’ signal, which is sent by the MS in response to the network’s request for handoff of this MS to another BS. 3.2.3 Network Operations Prior to describing some basic network operations in AMPS, we describe the three identifier numbers used in AMPS: † The Electronic Serial Number (ESN). The ESN is a 32-bit binary string that uniquely identifies an AMPS MS. This number is set up by the MS manufacturer and is burned into a Read Only Memory (ROM) in an effort to prevent unauthorized changes of this number. The fact that this number is stored in a ROM means that the MS will become inoperable if someone tries to rewrite the ESN. The format off an ESN is shown in Figure 3.2. It comprises three fields: (a) part 1, comprising bits from 24 to 31; this 8-bit field is the First Generation (1G) Cellular Systems 99 Figure 3.1 Mapping of SATS to SCC codes. manufacturers code (MFR), which uniquely identifies each manufacturer; (b) part 2 which comprises bits from 18 to 23 and has remained unusable; and (c) part 3, which comprises bits 0–17, which are assigned by the manufacturer to the MS. These bits are essentially the MSs serial number. When a manufacturer has produced so many MSs that 18 bits are no longer able to provide additional serial numbers for its MSs, it can apply to the FCC for an additional MFR. Thus, it can continue to produce MSs and MSs will be identified by a different MFR/serial number combination. † The System Identification Numbers (SIDs). These are 15-bit binary strings that are assigned to AMPS systems and uniquely identify each AMPS operator. SIDs are (a) transmitted by BSs to indicate the AMPS network they belong to and (b) used by MSs to indicate either the AMPS network they belong to (in cases of two collocated AMPS networks), or to determine roaming situations. † The Mobile Identification Number (MIN). This is a 34-bit string that is derived from the MSs 10-digit telephone number. 3.2.3.1 Initialization Once an AMPS MS is powered up, a sequence of events takes place. This sequence is briefly described below: † Event 1. The MS receives systems parameters in order to conFigure 3.itself to use one of the two AMPS networks. † Event 2. The MS scans the 21 control channels of the selected AMPS network to receive control messages. If a control channel with an acceptable quality is found, this is selected. † Event 3. The MS receives a message on the control channel containing system parameters. † Event 4. The message received in Event 3 provides the MS with information that is needed in order to update information that was received in possible previous initializations. Furthermore, the MS reads the SID of the AMPS network in this message, compares it to the SID of the network it belongs to and when the MS is in the service area of another network, the MS can prepare for roaming operations. † Event 5.The MS identifies itself to the network by sending its MIN, ESN and SIDS via the RECC. † Event 6.The AMPS network examines the parameters transmitted by the MS in Event 5 in order to determine whether this MS is a roaming one or not. † Event 7.The BS verifies initialization parameters by sending a control message to the MS. † Event 8.The MS enters idle state and waits for a call establishment request. During idle mode, the MS must perform operations to (a) ensure synchronization with the BS, (b) make the network aware of the MS’s location. Wireless Networks100 Figure 3.2 Structure of the 32-bit ESN. 3.2.3.2 Call Setup from a MS The procedure of placing a call from an MS can be described via a number of events. These events are summarized below: † Event 1.The MS sends to the BS a message containing the MS’s MIN, ESN and the phone number dialed. † Event 2.The BS passes the information sent by the MS to the network for processing. † Event 3.The BS indicates to the MS the channel number that will be used for the voice call. Furthermore, information related to the SAT frequency to be used is relayed to the MS. † Event 4. Both MS and BS switch to the voice channels. † Event 5.The BS sends a control message on the FVC via the SAT signal. † Event 6.The MS confirms link continuity via the SAT on the RVC. † Event 7.The call is established. 3.2.3.3 Call Setup to an MS The procedure of placing a call to an MS can be described via a number of events. These events are summarized below: † Event 1.The identification of the MS is passed to the BS. † Event 2.Control information, including the channel number to be used, is conveyed to the MS. † Event 3.The MS responds by sending its MIN, ESN and other control-related information. † Event 4.Information related to the SAT frequency to be used is relayed to the MS. † Event 5.Both MS and BS switch to the voice channels. † Event 6.The BS sends a control message on the FVC via the SAT signal. † Event 7.The MS confirms link continuity via the SAT on the RVC. † Event 8.The call is established. 3.2.3.4 Call Handoff The procedure of handoff in AMPS can be described via a number of events. These events are summarized below: † Event 1.The BS serving the MS notices a decrease in the MS’s transmission power. † Event 2.The BS sends a handoff measurement request to its MSC. † Event 3.The MSC instructs BSs in the neighborhood of the current BS to perform measure- ments of the MS’s signal strength. † Event 4.The MSC selects the best choice for a BS to serve the MS. † Event 5. The MSC allocates a traffic channel to the selected BS. † Event 6.The selected BS acknowledges the traffic channel allocation. † Event 7. The MSC sends a handoff message to the current BS. † Event 8.The current BS sends the handoff message to the MS. This message informs the MS which traffic channel to use and the power level of its transmission under the new BS. † Event 9.The MS confirms the current BS’s message and switches to the traffic channel. † Event 10.The MS starts scanning and eventually receives the new BS’s SAT. First Generation (1G) Cellular Systems 101 † Event 11.The MS confirms link continuity to the new BS via the SAT on the RVC. † Event 12.The new BS confirms the handoff to the MSC. 3.3 Nordic Mobile Telephony (NMT) NMT [4] has been deployed in several European countries. There are two versions of the system: the first operates in the area around 450 MHz and the second operates in the area around 900 MHz. These variants, are known as NMT 450 and NMT 900, respectively. 3.3.1 NMT Architecture An NMT system is made up of four basic parts: † Mobile Telephone Exchange (MTX) † Home Location Register (HLR), integrated in MTX or as a separate node † Base Station (BS) † Mobile Station (MS) The MTX and HLR control the system and include the interface to the Public Switched Telephone Network (PSTN). This interface can be made at local or international gateway levels. BSs are permanently connected to the MTX and are used to handle radio commu- nication with the mobile stations. BSs also supervise radio link quality via supervision tones. The set of BSs that are connected to the same MTX form an MTX service area, which in turn can be divided into subareas called Traffic Areas (TAs). The maximum number of BSs stations in a TA can be as high as 256. MSs can be vehicle-mounted, transportable or hand-portable. In order to set up a call to a mobile, a paging signal must be sent out in parallel from all BSs in the TA in which the mobile station resides, instead of being sent out on all BSs in the service area. The aim of this approach is to reduce call set-up time and system load. A number of network elements may also exist. These are: † Combined NMT/GSM Gateway (CGW) † Mobile Intelligent Network (MIN) † Authentication Register (AR). CGW is a gateway that can interrogate an NMT HLR and a GSM HLR. This is an optional feature for GSM MSCs that demands no new hardware. The HLR is used to store data about every subscriber, its services and location. In large networks where subscriber numbers are high, HLRs are preferably utilized as separate nodes, whereas in small networks, HLRs can be integrated with MTXs. The signaling protocol between MTSs and HLRs is according to CCITT Number 7 standard. Finally, The MIN adds intelligence to the network in order to enable introduction of new, customized services. The radio network consists of cells, each having a Calling Channel (CC) and a set of Traffic Channels (TC). In order to enable frequency reuse, adjacent BSs obviously employ different operating frequencies. The frequency reuse schemes that are typically employed divide the available frequencies among groups of 7, 12, or 21 cells. The reuse plan is then built up by repeating these groups by trying to optimize the distance between BSs that employ the same Wireless Networks102 frequency. In order to adjust to variable traffic intensities, cell size may change correspond- ingly. Radio coverage is provided in the cells by placing BSs either at (a) the center of the cell or (b) at a corner of the cell (omni cells or sector cells). The latter option gives the advantage of using one BS for several cells, thus reducing the number of BSs used and obviously deployment costs. The coverage of a BS ranges from 15 to 60 km for NMT 450 and from 2 to 30 km for NMT 900, depending on the BS placement height and the actual environment. 3.3.2 NMT Frequency Allocations Connections between BSs and MSs are utilized via full-duplex radio channels (ether in the 450 or 900 MHz band as mentioned before), which allow information to be exchanged simultaneously in both directions. These full duplex channels are utilized via a pair of uplink and downlink channels with BSs transmissions occurring in higher frequency bands than the transmissions of MSs. In NMT 450, 180 channels exist, separated via 25 kHz of spectrum. An optional extension band exists that can offer 20 more channels. With interleaved channels the system can use a total of 359 channels, which become 399 if the extended band is used. 3.3.3 NMT Channels There are four channel types in NMT. These are (a) the Calling Channel (CC), (b) the Traffic Channel (TC), (c) the Combined Calling and Traffic Channels (CC/TC) and (d) the Data Channel (DC). † Calling Channel (CC). Each NMT BS uses one channel as the calling channel. The CC is used by the BS for transmission of a continuous signal that identifies this BS to the mobiles. MSs within the cell of a BS lock onto the BSs CC. The CC is also used by the BS to page MSs under its coverage. Upon response of the MS, an additional channel, known as a TC, is allocated to the mobile. Finally, the CC may also be used for priority calls, meaning that messages over a CC can cause a user to terminate his call in order to receive one of a higher priority. † Traffic Channel (TC). The purpose of the TC is to carry the voice traffic. A TC can be in three different states: (a) ‘free marking’ state, in which the TC is mainly used for setting up calls from mobile stations; (b) ‘busy’ state, in which the TC is occupied by a voice call and (c) ‘idle’ state, in which the TC is not occupied. † Combined Calling and Traffic Channel (CC/TC). The CC of the BS can also operate as a combined calling and traffic channel. This is useful in cases where all traffic channels are occupied. In such cases, an MS can use the calling channel to set up a call. In such an event, the BS will completely lack a calling channel for some time. When a traffic channel becomes free, it functions as a combined calling and traffic channel. Thus, the BS’sCC will be used only when no other traffic channels are available. † Data Channel (DC). The DC is used to make signal strength measurements on mobile stations that are involved in a voice call on order from the MTX. The results of these measurements are used by the MTX at handover decisions. Every BS should have one CC, or some free TCs and one DC. Nevertheless, it is possible that a BS uses up to four DCs. This results in improved capacity for signal strength measurements, which is beneficial in situations characterized by increased traffic density or small cell sizes. First Generation (1G) Cellular Systems 103 3.3.4 Network Operations: Mobility Management 3.3.4.1 Paging Paging is used to determine the position of a MS. The service area of an MTX can be divided into a number of traffic areas. Paging involves sending over all CCs in the traffic area where the subscriber is expected to be (the area where the last registration of this MS was made) a page with the number of the paged MS number. Paging only the traffic area that is known to contain the MS helps reduce paging load on the system. However, if the MS is not found, then paging will be reinitiated and performed on all traffic areas of the MTX rather than only in that where the MS is expected to be. Upon reception of the page message, the MS will respond to the BS of the cell where the MS is currently located. If a certain time period elapses without a reply from the paged MS, then the page is considered unsuccessful. If the paging is unsuccessful, it is repeated once more. 3.3.4.2 Handover In order for handover to be performed, the radio connection quality is measured during the call. When the quality of the connection lowers, the BS that is currently serving the call signals the MTX. The purpose of this procedure is to investigate whether a BS with a better link quality to the mobile unit can be found. If such a BS is found and it has an available channel to serve the call, then a handover of the call to the new BS is initiated. If the handoff is to be performed, the MTX indicates to the mobile station that it must change its operating frequency to that of the new traffic channel selected in the new BS. The switch is made in the MTX at the same time as the mobile station changes its frequency. After a successful hand- over, the old channel is released. If, however, a BS with a better link quality to the mobile unit is not found, then the call continues with the current BS on the current channel and periodical signal measurements will be made in order to enable a successful handoff later. Normally 20– 30 s periods are used between successive attempts. If the handoff never takes place and the link quality continues to worsen (probably due to the subscriber moving far away from the BS) then the connection serving the call is dropped. A handover includes (a) seizing of the most suitable channel in the new BS, (b) supervision of the quality of the new channel, and (c) switching of the speech path towards the new channel. 3.3.4.3 Signal Strength Supervision The MTX also performs continuous supervision of channel quality through signal strength measurements. This operation improves call quality, as handovers will be performed at an earlier stage. 3.3.4.4 Intra-cell Handover This handover type involves moving a MS from a TC that experiences interference to another TC in the same BS. This procedure obviously improves call quality. Wireless Networks104