Wireless Networks: Where We Are, Where We're Headed

This chapter is from the book

Principles of Wireless Networks: A Unified Approach

Learn More Buy

This chapter is from the book

This chapter is from the book 

Principles of Wireless Networks: A Unified Approach

Learn More Buy

1.2 Different Generations of Wireless Networks

It is customary among the cellular telephone manufacturers and service providers to classify wireless communication systems into several generations. The first- generation (1G) systems are voice-oriented analog cellular and cordless telephones. The second-generation (2G) wireless networks are voice-oriented digital cellular and PCS systems and data-oriented wireless WANs and LANs. The third-generation (3G) networks integrate cellular and PCS voice services with a variety of packet-switched data services in a unified network. In parallel to the unified 3G standardization activities, broadband local and ad hoc networks attracted tremendous attention, and they developed their own standards. One of the major current differences between these two waves is that the 3G systems use licensed bands, and broadband and ad-hoc networks operate in unlicensed bands. The manner in which broadband local access in unlicensed bands and 3G standards in licensed bands may be integrated forms the core of the forthcoming generations of wireless networks.

1.2.1 1G Wireless Standards

Table 1.3 shows the worldwide 1G analog cellular systems. All these systems use two separate frequency bands for forward (from base station to mobile) and reverse (from mobile to base station) links. Such a system is referred to as a frequency division duplex (FDD) scheme. The typical allocated overall band in each direction, for example, for AMPS, TACS, and NMT-900, was 25 MHz in each direction. The dominant spectra of operation for these systems were the 800 and 900 MHz bands. In an ideal situation, all countries should use the same standard and the same frequency bands, however, in practice, as shown in Table 1.3, a variety of frequencies and standards were adopted all over the world. The reason for the different frequencies of operation is that the frequency administration agencies in each country have had previous frequency allocation rulings that restricted the assignment choices. The reason for adopting different standards was that at that time cellular providers assumed services to be mainly used in one country, and they did not have a vision for a universal service. The channel spacing or bandwidth allocated to each user was either 30 kHz or 25 kHz or a fraction of either of them. The 25 kHz band was previously used for mobile satellite services and the 30 kHz band was something new used for cellular telephone applications. All the 1G cellular systems were using analog frequency modulation (FM) for which the transmission power requirement depends on the transmission bandwidth. On the other hand, power is also related to the coverage and size of the radios. Therefore, one can compensate for the reduction in transmission bandwidth per user by reducing the size of a cell in a cellular network. Reduction in size of the cell increases the number of cells and the cost of installation of the infrastructure.

Table 1.3 Existing 1G Analog Cellular Systems

Band splitting obviously allows more subscribers at the expense of more investment in the infrastructure. In Chapter 5, where we address deployment of the cellular networks, we introduce a technique to improve the capacity of a cellular network with the same number of base stations using band splitting.

In the cellular industry, often 1G only refers to analog cellular systems because it is the only system implemented based on popular standards such as AMPS or NMT. However, we can generalize the 1G systems to other sectors of the wireless services. The analog cordless telephone that appeared on the market in the 1980s can be considered the 1G cordless telephone. Paging services that were deployed around the same time frame as analog cellular systems and cordless telephones can be referred to as 1G mobile data services providing one-way short data messages. In the early 1980s before the release of ISM bands and start of the WLAN industry, a couple of small companies in Canada and the United States developed low-speed connectionless WLANs using voiceband modem chip sets and commercially available walkie-talkies. These products were operating at the speed of voice-band modems (< 9,600 bps) but using medium-access control techniques used in data-oriented LANs. Because of their low speed they do not comply with the IEEE 802 definition of LANs, but one may refer to them as 1G WLAN products.

1.2.2 2G Wireless Systems

The 2G systems supported a complete set of standards for all four sectors of the wireless network industry. As we discussed in the history of voice-oriented and data-oriented networks, we have a number of digital cellular, PCS, mobile data, and WLAN standards and products that can be classified as 2G systems. In the remainder of this section, we cover each of these four sectors of 2G systems in a separate subsection.

1.2.2.1 2G Digital Cellular

Table 1.4 summarizes the major 2G digital cellular standards. There are four major standards in this category: GSM, the pan-European digital cellular, the North American Interim Standard (IS-54) that later on improved into IS-136 and Japanese digital cellular (JDC)—all of them using TDMA technology and IS-95 in North America, which uses CDMA technology. Like the 1G analog systems, the 2G systems are all FDD and operate in the 800–900 MHz bands. The carrier spacing of IS-54 and JDC is the same as the carrier spacing of 1G analog systems in their respective regions, but GSM and IS-95 use multiple analog channels to form one digital carrier. GSM supports eight users in a 200 kHz band; IS-54 and JDC support three users in 30 and 25 kHz bands, respectively. As we explain in Chapter 4, the number of users for CDMA depends on the acceptable quality of service; therefore, the number of users in the 1,250 kHz CDMA channels cannot be theoretically fixed. But this number is large enough to convince the standards organization to adopt CDMA technology for next generation 3G systems. By looking into these numbers, one may jump to the conclusion that GSM uses 25 kHz for each bearer, and IS-54 uses 10 kHz per user. Therefore GSM supports 2.5 times less the number of users in the given bandwidth. The reader should be warned that this is an illusive conclusion because when the network is deployed, the ultimate quality of voice also depends on the frequency reuse factor and signal-to-noise/interference requirements that will change these calculations significantly. These issues are addressed in Chapter 5.

Table 1.4 2G Digital Cellular Standards

The channel bit rate of GSM is 270 kbps whereas IS-54 and JDC use 48 and 42 kbps, respectively. Higher channel bit rates of a digital cellular system allow simple implementation of higher data rates for data services. By assigning several voice slots to one user in a single carrier, one can easily increase the maximum supportable data rate for a data service offered by the network. How the higher channel rate of GSM allows support of higher data rates is discussed in Chapter 9, when we discuss GPRS mobile data services. Using a similar argument, one may notice that the 1228.8 kcps channel chip rate of IS-95 provides a good ground for integration of higher data rates into IS-95. This fact has been exploited in IMT-2000 systems to support data rates up to 2 Mbps.

The speech coding technique of 2G systems are all around 10 kbps. As shown in Table 1.4, cellular standards assume large cell sizes and a large number of users per cell, which necessitates lower speech coding rates. On the other hand, cellular standards were assuming operation in cars for which power consumption and battery life were not issues. The peak transmission power of the mobile terminals in these standards can be between several hundreds of mW up to 1W [PAH95] and on the average they consume around 100 mW. All of these systems employ central power control, which reduces battery power consumption and helps in controlling the interference level. In digital communications, information is transmitted in packets or frames. The duration of a packet/frame in the air should be short enough so that the channel does not change significantly during the transmission and long so that the required time interval between packets is much smaller than the length of the packet. A frame length of around 5 to 40 ms is typically used in 2G networks.

1.2.2.2 2G PCS

As we discussed in the history of wireless voice-oriented networks, the 2G PCS standards evolved out of the 1G analog cordless telephone industry and merged into the 3G cellular systems. Figure 1.5(a) illustrates the difference between these two industries during the evolution of the 2G networks; before they were integrated into one in 3G systems. Table 1.5 illustrates a more quantitative comparison of PCS and cellular industries that at the time was used to justify the existence of two separate voice-oriented standards. The basic philosophy was that PCS is for residential applications, the cell size is small, coverage is zonal, antennas are installed on existing posts (such as electricity or telephone posts), it is not designed to be used in the car, and the complexity of the handset and base station is low. These standards preferred 32 kbps speech coding to support wireline quality and shared the same spectrum in different zones; time-division-duplex (TDD) and noncoherent modulation techniques were mostly used to support simpler implementation.

Table 1.5 Quantitative Comparison of PCS and Cellular Philosophies

Table 1.6 provides a summary of the specifications of the four major PCS standards. CT-2 and CT-2+ were the first digital cordless telephone standards; PHS, which later on became PHP, was the first and the only nationwide deployment of these systems; and PACS is the last standard developed with this philosophy. Except for CT-2+ all these standards were designed for 1.8 and 1.9 GHz frequency bands, which are commonly referred to as PCS bands; all systems use TDMA/TDD except PACS, which adopted FDD. To support wireline quality of voice, speech coding at 32 kbps is used in all of these standards. This rate is around three times higher than the speech coding rate of digital cellular systems. The carrier bit rate of 1,728 kbps in DECT is even higher than GSM, which had the highest carrier rate of all TDMA digital cellular systems. This carrier bit rate is even higher than the chip rate in IS-95, the 2G CDMA standard. This feature provides an edge to DECT in supporting high-speed data connections for Internet access. Perhaps this is the major reason why DECT is the only PCS standard that is still considered in new technologies like HomeRF. The power consumption of PCS standards is almost one order of magnitude less than that of the digital cellular standards because PCS systems are designed for smaller cells. If digital cellular systems were deployed with the same cell sizes, the average power consumption could be comparable to PCS systems. Modulation techniques used for PCS standards, GFSK and DQPSK, are less bandwidth efficient and more power efficient than modulation schemes used in digital cellular systems. These modulation techniques can be implemented with simpler noncoherent receivers reducing the size of the handset. The shorter propagation time for the smaller-cell PCS standards allows shorter packet frames which help the quality of voice in spite of the wireless channel impairments.

Table 1.6 2G PCS Standards

1.2.2.3 Mobile Data Services

As shown in Figure 1.5(b), mobile data services provide moderate data rate and wide coverage area access to packet-switched data networks. The mobile data networks emerged after the success of the paging industry to provide a two-way connection for larger messages. Table 1.7 provides a comparison among a number of important mobile data services. ARDIS and Mobitex use their own frequency bands in 800–900 MHz, terrestrial European trunked radio (TETRA) uses its own band at 300 MHz, CDPD shares the AMPS bands and site infrastructure, GPRS shares the GSM's complete radio system, and Metricom uses the unlicensed ISM bands. The early systems, ARDIS, Mobitex, and CDPD, were developed before the popularity of the Internet, and the dominant design criteria was coverage and cost rather than data rate. These systems were a wireless replacement for voice-band modems operating at data rates up to 19.2 kbps, which was the rate of modems at that time. TETRA is designed for pan-European civil service application and has its own features for that purpose. Metricom and GPRS support data rates more suitable for Internet access. The advantage of GPRS is that it is incorporated in the popular GSM digital services with large numbers of terminals all over the world. Except for Metricom, the channel spacing of the rest of the mobile data services is based on the channel spacing of cellular telephone networks with 25 or 30 kHz bands or a fraction (12.5 kHz) or a multiple of them (200 kHz). These services are designed to use multiple carriers in an FDMA format and use different versions of random access techniques such as DSMA, BTMA, or ALOHA, which are explained in Chapter 4. Modulation techniques for these systems are like digital cellular and PCS systems explained in Chapter 3. A more detailed comparison of mobile data services is provided in Chapter 9.

Table 1.7 Mobile Data Services

1.2.2.4 WLANs

As shown in Figure 1.5(b), WLANs provide high data rates (minimum of 1 Mbps) in a local area (< 100 m) to provide access to wired LANs and the Internet. Today all successful WLANs operate in unlicensed bands that are free of charge and rigorous regulations. Considering that PCS bands were auctioned at very high prices, in the past few years WLANs have attracted a renewed attention. Table 1.8 provides a summary of the IEEE 802.11 and HIPERLAN standards for WLANs. IEEE standards include 802.11 and 802.11b operating at 2.4 GHz and 802.11a, which operates at 5 GHz. Both HIPERLAN-1 and -2, developed under the European Telecommunication Standards Institute (ETSI), operate at 5 GHz. The 2.4 GHz products operate in ISM bands using spread spectrum technology to support data rates ranging from 1 to 11 Mbps. HIPERLAN-1 uses GMSK modulation with signal processing at the receiver that supports up to 23.5 Mbps. IEEE 802.11a and HIPERLAN-2 use the orthogonal frequency division multiplexing (OFDM) physical layer to support up to 54 Mbps. The access method for all 802.11 standards is the same and includes CSMA/CA, PCF, and RTS/CTS, which are described in Chapters 4 and 11. The access method of the HIPERLAN-1 is on the lines of the 802.11, but the access method for HIPERLAN-2 is a voice-oriented access technique that is suitable for integration of voice and data services. The details of these transmission techniques and access methods are described in Chapters 3 and 4. IEEE 802.11, IEEE 802.11b, and HIPERLAN-1 are completed standards, and IEEE 802.11 and 11.b are today's dominant products in the market. IEEE 802.11a and HIPERLAN-2 are still under development. The IEEE 802.11 and HIPERLAN-1 standards can be considered 2G wireless LANs. The OFDM wireless LANs are forming the next generation of these products. The last four chapters of this book are focused on wideband local access systems which describe these systems in further detail.

Table 1.8 Wireless LAN Standards

1.2.3 3G and Beyond

The purpose of migration to 3G networks was to develop an international standard that combines and gradually replaces 2G digital cellular, PCS, and mobile data services. At the same time, 3G systems were expected to increase the quality of the voice, capacity of the network, and data rate of the mobile data services. Among several radio transmission technology proposals submitted to the International Telecommunication Union (ITU), the dominant technology for 3G systems was W-CDMA, which is discussed in detail in Chapter 11. Outside the 3G standards, WLAN and WPAN standards are forming the future for the broadband and ad hoc wireless networks. Figure 1.6 illustrates the relative coverage and data rates of 2G, 3G, WLAN, and WPANs. WPANs, studied in Chapter 13, are formed under the IEEE 802.15 standard. This community has adopted Bluetooth technology as its first standard. Bluetooth is a new technology for ad hoc networking which was introduced in 1998. Like WLANs, ad hoc Bluetooth-type technologies operate in unlicensed bands. Bluetooth operates at lower data rates than WLANs but uses a voice-oriented wireless access method that provides a better environment for integration of voice and data services. The WPAN ad hoc networking technologies are designed to allow personal devices such as laptops, cellular phones, headsets, speakers, and printers to connect together without any wiring.

Figure 1.6 Relative coverage, mobility, and data rates of generations of cellular systems and local broadband and ad hoc networks.

From the point of view of cellular service providers, 3G provides multimedia services to users everywhere, WLANs provide broadband services in hot spots where a short proximity is needed, and WPANs connect the personal devices together. The telecommunications industry is a multidisciplinary industry, and it has always been difficult to predict its future. However, there are certain current trends that one may perceive as important. In terms of frequency of operation, 3G systems use licensed bands whereas WLANs and WPANs use unlicensed bands. Unlicensed bands are wider and free of charge and rigorous rules, but there is no regulation to control the interference in these bands. Some researchers and visionaries have gone too far to state that the future is "everything unlicensed," but it is safer to say that recent years have witnessed an immense growth of hope for an increase in using unlicensed bands. Wideband CDMA is the dominant transmission technology for 3G systems and OFDM is becoming very popular in broadband WLANs operating in 5 GHz. Again some visionaries also state that the next- generation systems will be based on OFDM; however, it is safe to say that OFDM appears to play an increasing role in the future of broadband wireless access. Another important evolving technology is the ultra wide band (UWB) which is expected to support a myriad of users with noiselike impulses with a spectrum spreading over several GHz [SCH00]. We address this evolving technology in Chapter 3. There are some visionaries who predict that UWB may be the next step to OFDM, but it is safe to state that it is an evolving, promising technology. Another evolving technology is position finding, in particular in indoor areas, which is becoming an integral part of the wireless networks for the next generations. We address location finding in the last chapter of this book. The FCC in the United States has already mandated the integration of position location systems with cellular systems; however, the extent and method of integration is not yet clear.