The accuracy of GLONASS is of the same level as that of GPS. Both systems are now considered as cooperative, entering the integrated Global Navigation Satellite System (GNSS). It has already been emphasized that increasing the number of processed satellite signals improves the positioning precision, so joint use of both constellations is obviously profitable. In addition, scenarios are not rare, where some satellites over the horizon are obstructed (e.g. by an aircraft wing), so that the total number of available ranging signals within only GPS or GLONASS is not sufficient for positioning. Then again, joint processing of GPS and GLONASS signals may give a considerable gain in positioning integrity. A great number of receiver models presently on the market or in development are capable of combined processing of signals of both systems.

11.2.6 Applications

The role of satellite-based global navigation systems in the modern and future world can hardly be overestimated. Just simply naming the areas of their involvement forms rather a long list, including traditional navigation of ships, aircraft and terrestrial moving objects (cars, trucks etc.), transit systems, mapping utilities (e.g. pipelines), monitoring forestry and natural resources, farming, civil engineering, geodetic surveying, seismic forecasting, airborne mapping, seafloor investigations and many more. Not being able to go deeper into this fascinating topic, we direct the interested reader to the sources [117–119] and their references.

11.3 Air interfaces cdmaOne (IS-95) and cdma2000

11.3.1 Introductory remarks

The first interim specifications of the 2G CDMA cellular telephone of standard IS-95 (presently referred to as cdmaOne, too) were published in 1993–1995, and the operational phase of IS-95 networks started in 1996. Nowadays networks of this standard cover huge territories serving tens of millions of consumers. Its impressive commercial success, widely recognized high quality of service and openness to further modernizations were among the decisive factors favouring the CDMA philosophy as the basic platform for the next generations of mobile radio (3G and beyond).

Initially IS-95 was meant to gradually replace (maintaining compatibility with) an American analog standard, AMPS, operating in the 800 MHz range. The IS-95 documents set up frequency division separation of forward (869–894 MHz) and reverse (824– 849 MHz) links,1 while no limitation on frequency reuse in neighbouring cells or sectors was stipulated. The nominal bandwidth of the IS-95 signal is about 1.25 MHz, so that within the total assigned 25 MHz band an operator has remarkable freedom in carrier selection and frequency planning of the network. All the BSs entering a network are strictly synchronized via GPS to operate in a unified time scale, allowing MS easier

1 The terms ‘forward’ and ‘reverse’ links are synonyms of downlink (BS to MS) and uplink (MS to BS) adopted in the cdmaOne and cdma2000 specifications.

switching from one BS to another (handover). IS-95 and its 3G evolution cdma2000 are typical DS spread spectrum systems, which clearly manifest all the benefits of this technology. They also possess very high educational value, since they demonstrate in a lucid form practical ways of realizing many ideas studied above. In the text to follow we are going to dwell on only the most general principles of spreading, channelization, coding and modulation in the IS-95 and cdma2000 air interfaces. Readers who wish to acquire deeper knowledge may consult the sources [18,69,83,120,121] and many others.

11.3.2 Spreading codes of IS-95

The spreading sequences used in the IS-95 standard were partly mentioned in examples earlier. They are designed to provide CDMA separation of physical channels, distinguishability of signals of different BS arriving at the MS receiver and privacy of the transmitted data. Synchronous CDMA multiplexing of physical channels of the forward link served by a fixed BS is realized on the basis of Walsh sequences (see Section 2.7.3) of length N ¼ 64. The orthogonality of Walsh sequences allows separating the corresponding 64 physical channels theoretically with no MAI. The duration of a chip of Walsh sequences is nearly 0:81 ms and the chip rate is 1.2288 Mcps (megachip per second), resulting in the abovementioned bandwidth of 1.25 MHz. Certainly, the number of forward-link physical channels thus implemented is 64 and, consistent with the CDMA principle, they occupy the same common bandwidth with no frequency or time offset. All of the base stations use the same set of 64 Walsh functions, and the spreading by so-called short codes makes the signals of different base stations separable from each other in the MS receiver. There are two different basic binary short-code pseudonoise sequences, PN-I and PN-Q, used in the in-phase and quadrature-phase branches of the BS modulator, respectively. They are primarily generated as two m-sequences whose LFSR generators (see Section 6.6) contain 15 flip-flops and are defined by

are extended by one more zero symbol following after 14 consecutive zeros. This brings the lengths to N ¼ L þ 1 ¼ 215 ¼ 32 768 chips, and with the same chip rate as for Walsh codes there are 37.5 periods of the short codes per second or 75 periods over two seconds. To discriminate between different base stations every one of them employs its BS-specific time-offset replica of the basic short-code sequences. There are 512 such pairs of replicas, every pair being shifted compared to the previous one by 64 chips or about 52 ms. The network planning should assign short-code pairs to the base stations in a way guaranteeing low risk of any MS receiving a signal from an unintended BS, whose timing due to propagation delay is about the same as that of the desired signal and whose strength is sufficient for mixing them up. It should be stressed that relative timeoffsets between the base stations entering a network, once set up, remain constant forever, since all the BS use GPS receivers to synchronize their clock oscillators with each other.

One more spreading code is a long code, generated primarily as a binary m-sequence of memory 42. According to the specification a primitive polynomial of the LSFR