above the one of IS-95 are arranged, transmitting antenna-specific pilot signals from two BS antennas involved in diversity. This enables the MS to separate pilots of different antennas and have strictly synchronized references for the diversity signals to demodulate, decode and properly combine diversity branches.

More complex organization of cdma2000 requires stronger systematization of logical forward channels. Instead of only four in IS-95, the hierarchy of logical channels in cdma2000 contains 10 only on the first layer, adding common control channel, common power control channel, broadcast control channel, quick paging channels etc. As for the traffic channels, they include, among others, the fundamental ones (primarily used) and supplemental channels involved in multicode transmission when very high rates not achievable with only fundamental channels are on request. The supplemental channels may use either convolutional or turbo codes (see Section 9.4).

The reverse link of cdma2000 does not use the MC-DS-CDMA technique, being implemented as conventional DS CDMA with three times broader bandwidth (3.75 MHz). This corresponds to three times higher chip rate (3.6864 Mcps) of the long spreading code. With such high DS spreading rate, a number of solutions of IS-95 were revised and replaced by more relevant ones. First of all, a convolutional encoding rate 1/4 is used instead of 1/3, meaning that the raw data rate, say 9.6 kbps, after encoding becomes 38.4 kbps. Using QPSK data modulation and spreading such a codestream by a long code gives a spreading factor 2 (3:6864/38:4) 103 ¼ 192, which is not a critical loss as compared to 256 in IS-95. At the same time, the reduction of rate of a convolutional code of constraint length 9 to 1/4 returns some extra (non-asymptotic) coding gain. On this ground the designers decided to reject the orthogonal 64-ary modulation and bring the reverse traffic channel structure closer to that of the forward link.

Another distinction is the long list of logical channels, including the reverse pilot channel, which now becomes obligatory, since in the absence of orthogonal modulation the BS should have a local coherent reference to demodulate the QPSK data of MS. There are other new logical channels, some of them running simultaneously. Being channelized by the Walsh signals, they are then linearly summed, meaning that the transmitter power amplifier should be linear so involving O-QPSK is needless. Thus, an ordinary QPSK data modulator is used, where the QPSK-mapped codestream is first multiplied by the QPSK spreading code and then the real and imaginary parts of the product modulate the cosine and sine CW carrier components, as explained in Section 7.1. Note that, similarly to the forward link, reverse supplemental channels may support a reverse fundamental (traffic) channel (one per MS), providing multicode transmission at the high rates unattainable with only fundamental channels. The supplemental channels may again use either convolutional or turbo codes.

11.4 Air interface UMTS

11.4.1 Preliminaries

The UMTS (acronym of Universal Mobile Telecommunication System) is a 3G wideband CDMA standard whose development was pioneered by the European telecommunication community. Currently UMTS, along with cdma2000, enters the so-called IMT-2000 family, i.e. the list of standards declared by the International Telecommunication

Union (ITU) to be basic for 3G systems. There are two versions of UMTS: frequency division duplex (FDD) and time division duplex (TDD). As is immediately seen from the names, these systems differ from each other in the method of separation between downlink and uplink. In FDD-UMTS downlink and uplink occupy non-overlapping frequency bands, while in TDD-UMTS they employ different time slots. For brevity we limit the discussion below to only FDD-UMTS, keeping in mind that many solutions adopted in it are common to both systems.

International regulatory documents allocate for FDD-UMTS in Europe frequency bands 1920–1980 MHz (uplink) and 2110–2170 MHz (downlink) with a limitation of 5 MHz on the link bandwidth. DS spreading is a fundamental technique securing separation of physical channels, in particular multiple access (DS CDMA). The standard sets a universal and constant chip rate 3.84 Mcps in full agreement with the bandwidth limitation. Like cdma2000, UMTS is a system in which the data transmission rate may vary in a very wide range. As a consequence of this, along with chip rate invariance, the spreading factor changes with transmission rate.

In what follows we concentrate again on the physical layer of the system, i.e. solutions concerning spreading, channelization and modulation. Note that BSs of UMTS do not rely upon GPS support and operate with their own autonomous non-synchronized clocks. Although this architecture saves on BS equipment costs, it does, however, make cell search and handover procedures more complicated and is responsible for many distinctions between the physical layers of UMTS and cdma2000.

Similarly to Section 11.3, we can hardly present here anything more than a very short sketch, but many recently published books dedicated to the UMTS standard [92,104,114,115,120–122] will help the interested reader to expand his/her knowledge about this system, which is likely to become dominant among mobile telecommunications in the near future.

11.4.2 Types of UMTS channels

According to the terminology of the UMTS specifications there are logical, transport and physical channels. On the layers higher than physical the data is distributed between the logical channels on the basis of information content, but before arriving at the physical layer it is restructured into the transport channels. The criterion to distinguish transport channels is mode and format of data representation, while physical channels (which are just signals bearing messages), as in any CDMA system, are distinguished by their specific codes. The physical layer consists of two sublayers. The transmitted data arrives at the first physical sublayer from the upper layers packed into transport channels according to information content. The first sublayer includes among other things attachment of CRC for data block protection, channel coding and interleaving. Channel coding is either convolutional (constraint length 9, rates 1/2 or 1/3) or turbo coding (constituent encoders of memory 3, rate 1/3). For very high data rate uncoded transmission is possible, too. The second sublayer, i.e. radio link, covers mapping the transport channels to the physical ones (signals), and transmitting signals over the propagation medium. The signal received by MS (downlink) or BS (uplink) then undergoes all necessary reciprocal operations of the two sublayers (demodulation,