1、 STDoCEPT ERC REPORT 52-ENGL 1997 m 232hYL4 0023308 778 = ERC REPORT 52 European Radiocommunications Committee (ERC) within the European Conference of Postal and Telecommunications Administrations (CEPT) METHODOLOGY FOR THE ASSESSMENT OF PMR SYSTEMS IN TERMS OF SPECTRUM EFFICIENCY, OPERATION AND IMP
2、LEMENTATION Bucharest, December 1997 Copyright 1998 the European Conference of Postal and Telecommunications Administrations (CEPT) STD-CEPT ERC REPORT 52-ENGL 1997 m 2326414 0013310 32b m ERC REPORT 52 METHODOLOGY FOR THE ASSESSMENT OF PMR SYSTEMS IN TERMS OF SPECTRUM EFFICIENCY. OPERATION AND IMPL
3、EMENTATION 1 INTRODUCTION . 1 2 SCOPE OF THE REPORT 1 3 DEFINITIONS. ACRONYMS in TDMA trunked systems with a proportion of traffic between unsynchronised mobiles, the number of usable time-slots per carrier may be reduced. The mode factor N, takes into account the mode of operation. With these defin
4、ition: .t is assumed that the temporarily unused radio capacity during a conversation, e.g. the reverse channel in duplex systems, is not used for other purposes. This might not be true in particular cases, e.g. packet radio systems. In these cases, NM is increased above its conventional system valu
5、e. Without trunking only a limited percentage of the available radio capaciy can be used in practice and even with efficient trunking methods the efficiency of channel usage is well below 100%. However, trunking is applicable to all mobile radio systems and thus can be disregarded in the comparison
6、method. It should also be noted that the use of omnidirectional antennas in the base stations as well as in the mobiles and a uniform distribution of the mobiles is assumed. The interrelation of modulation bandwidth B, and carrier separation AF, should also be considered: Sf, and Sf, are the frequen
7、cy tolerances of the receiver and transmitter which are often negligible compared to the modulation bandwidth. Generally the modulation bandwidth B, is identical to the receiver modulation acceptance bandwidth B, and denotes about 98% of the transmitted power. In special cases the receiver pass band
8、width may be smaller than the modulation bandwidth but then distortions have to be expected and compensated. In other cases the receiver centre frequency tolerance is not explicitly taken into account because it is already included in the receiver pass bandwidth. B, is the modulation bandwidth arisi
9、ng from the transmitter, defined as including all modulation products attenuated by less than a certain amount from the level of the carrier. When considering channel separation, the value of B, to be taken is that related to the -60 dBc or -70 dBc points, since, particularly for simplex systems, is
10、 it important that receivers are protected from excess adjacent channel power. Using Carsons rule, the maximum possible transmission modulation bandwidth is l6kHz for a 25kHz system, derived using the peak frequency deviation at the maximum modulating frequency - 5kHz and 3kHz respectively for 25kHz
11、 systems. The -60 or -70dBc bandwidth is approximately twice the Carson bandwidth. Neglecting tolerances, equation (2) results in a carrier separation of 24kHz. In the limits sometimes the transmitters frequency tolerance may also be included. It should be noted that for constant envelope FM and PM
12、systems BM 1. In most PMR systems, the range is about 9 I N, I 19. In the case N, = 1, the frequency efficiency of interference limited systems becomes identical to that of noise or coverage limited systems. (For CDMA the cluster size is generally defined as the ratio of the maximum number of availa
13、ble channels per cell in a monocell system to the maximum number of available channels per cell in an infinite uniformly loaded multicell system. It is claimed that this ratio lies between 1.5 and 2.0) For heavily loaded systems with strong co-channel interference and a = 4, the number of channels c
14、an be expressed using instead of N,. N, is the average load factor of the interfering cells. If these belong to the same system then N, = N, can be assumed. The load factor NL = O. 1. In congested areas NL = 0.3 may be taken for non-trunked systems while an estimate of NL = 0.7 might be more appropr
15、iate for very heavily loaded trunked systems with a large number of available traffic channels. All these considerations need great care and the results may vary From case to case particularly when mixed scenarios have to be evaluated. In most PMR systems a = 3.5 is a more correct assumption but the
16、n the formula becomes much more complicated without giving significantly different results in the case of rough system comparisons. For absolute figures the formula is: NA * NM Bsjw (Al?: / 3) * (NLI) * (C / I)D NI = RTC/cell (7) It should be noted that is the carrier to interference power ratio und
17、er fading conditions including shadowing. This means that fading and shadowing, which are very dependent on the propagation conditions, have a great influence on and reuse distance and consequently on the spectral efficiency. However, if different systems are compared under identical propagation con
18、ditions then all these factors generally have only small or negligible influence. For the purpose of the calculations used in these comparisons, only fading has been taken into account, because (CA), for most digital systems is known and can be estimated easily for analogue systems 2 . Using digital
19、 transmission the spectrum efficiency q, for interference limited systems also has to take the cluster size into account: (bit/s)/(Hz, cell) Again only the net bit rate R, per traffic channel is of interest. STD-CEPT ERC REPORT 52-ENGL 1997 2326414 0013321 101 W ERC REPORT 52 Pape 11 5.4 Other syste
20、m limitations There are additional system limitations. In contrast to the limitations above which are based on hard physical facts, the limitations referred to hereafter are by nature soft facts and can be overcome with increased technical effort. Some of the limiting factors affect simulcast system
21、s more than normal systems, requiring exceptional care to be taken in such cases. Delay limited systems exhibit a poor ratio of burst to guard time which is a problem associated with TDMA but not with FDMA. For large coverage areas and long signal travelling times therefore the duration of guard tim
22、e and burst ramping time must be shortened in order to improve efficiency if the burst time cannot be made longer. The guard time can be considerably shortened if time advance methods are introduced. This means that the mobile transmits its bursts with varying time advance compared to the received b
23、ase station TDMA frame to compensate for varying signal propagation times. However, the guard and ramping times together cannot reasonably be made shorter than the delay spread as determined by the multipath propagation conditions. Dispersion limitations occur when intersymbol interference is introd
24、uced by multipath propagation conditions. This occurs when the delay spread exceeds a considerable percentage of the symbol duration. Obviously this becomes very critical when half the symbol time is approached. However, this limitation can be overcome by equalising methods where each burst contains
25、 a well-known training sequence from which the channel propagation conditions can be calculated and be used to restore the unknown message symbols. The necessary effort is generally significant. Depending on the type of modulation and the bandwidth the Dopder spread may also limit system performance
26、 if it is not negligible compared to the modulation bandwidth. Here again suitable equalising methods might be applied to overcome this problem, requiring additional effort. pandwidth on Demand .s mobile and data applications become more prevalent in mobile communication systems, the ability to supp
27、ort increased data rates will become more important. TDMA systems can provide enhanced data capabilities by allocating additional capacity to users when required to increase the data rate available within the same channel separation. For example, in TETRA, a user employing one time-slot can have an
28、unprotected data rate of 7.2 kbps, the same user however can be allocated all 4 time-slots thus providing a 28.8 kbps unprotected data rate capability within a 25 kHz carrier separation. This may also be possible with FDMA systems if they have contiguous channels. This feature requires specific term
29、inals with extra processing and transmission mean power capacity. 5.5 Mixed scenarios In many real systems, a combination of interference and coverage limitations may be observed. In this case, the appropriate measure for spectrum efficiency is a function of the type of services. For group calls, it
30、 is desirable to ensure as many members of the group as possible are in the same cell and thus coverage limited systems seem preferable; for individual calls with a fixed party, the interference limited approach seems more suitable. Moreover, with the advent of new technology, using FDMA techniques
31、to split the radio resource from 25/20112.5 kHz channels into for instance 12.5/10/6.25/5 kHz channels, or using TDMA techniques to split the radio resource into time slots can be ways to provide extra capacity i.e. a greater number of physical channels per MHz and cell than is available with conven
32、tional old technology. In addition, it may be more economical and more spectrally efficient for a small user group with low traffic requirements to subscribe to a national or regional PAMR system rather than invest in a self- provided system (PMR). 5.6 Methods for the improvement of spectrum efficie
33、ncy For a basic given system, the spectrum efficiency can be further improved. This is directly possible by the introduction of trunking techniques. Methods such as voice activity detection (VAD), discontinuous transmission (DTX), transmitter power control and in a limited sense also discontinuous r
34、eception (DRX) reduce interference directly or at least reduce its appearance in the receiver. This makes additional capacity available which can be used to carry additional traffic. STD-CEPT ERC REPORT 52-ENGL 3997 = 232b434 0033322 048 ERC REPORT 52 Page I2 Improved coding, interleaving, equalisat
35、ion and detection with improved data compression techniques will also result in improved spectrum efficiency. Since most of these methods are applicable with similar results to all systems, they need not necessarily be taken into account for the purpose of the evaluation of basic systems, for which
36、the theoretical maximum possible spectrum efficiency should be evaluated assuming for comparison purposes that one single frequency simplex channel provides the capacity of one radio traffic channel (RTC). Concerning the influence of the multiple access mode, FDMA or TDMA, on the spectrum efficiency
37、 of PMR systems, the two parameters net c-tta (or information) rate to channel separation ratio and limit of the dynamic signal to interference ratio are, in the first approach, the same for the two modes of access provided identical modulation schemes, but different symbol lengths, are used, with p
38、erhaps a small advantage in favour of FDMA which is less sensitive to distortions due to multipath propagation. Instead of the ratio net data rate to modulation bandwidth, which is a precise theoretical measure, the ratio net data rate to channel separation is more relevant for real systems because
39、this also reflects operational requirements among other things. However due to the specific configurations (relatively small coverage) and the specific services (group calls, half-duplex operation) of PMR with respect to public radiotelephone networks, the potential for achieving the largest possibl
40、e individual cell coverage is an important factor for increasing the efficiency of the radio systems and decreasing the cost of the networks. All other things being equal, in particular for the same transmitter peak power and with the same modulation and coding schemes, a FDMA system (one channel pe
41、r carrier) will provide wider coverage than a TDMA system (several channels per carrier). However, if the average power per traffic channel is the same, the coverage will be the same; provided all other parameters are kept the same. When the density of traffic is low or irregular and the system is c
42、overage limited, FDMA is more flexible and efficient than TDMA for PMR applications. 6 OTHER CONSIDERATIONS Not all of the parameters of a radio transmission system are relevant for spectrum efficiency. However they must fulfil the user needs and some of them must be taken into account when comparin
43、g systems, e.g.: - Domder effect If the Doppler degradation of a highly spectrum efficient system is bad, then this system may be useless for mobiles travelling at high speed. - C/I If theC/I of one system is much better than that of another, this may have additional benefits in a rnultipath propaga
44、tion environment, where this may permit considerable reduction of radio channel equalisation needs. - channel access It is not believed possible to increase the capacity of spectrum to the extent that radio channels can be made available on an exclusive basis in dense urban areas, thus channels must
45、 be shared. The protocols for access to shared channels will affect the overall efficiency of the use of spectrum. - adaptation to the PMR environment, robustness, ease of implementation It is necessary to examine the feasibility of implementation of new techniques in the PMR environment. Whereas pu
46、blic radiotelephone operators are able to invest in order to have good sites, PMR users generally have to install equipment without close consideration of site engineering dependent radio parameters (intermodulation due to non-linearity etc.). The technology must be easy to implement and use, whilst
47、 being robust and cost effective. - functionality PMR users historically have not usually needed elaborate functionality and features from their systems. When comparing different systems, one must be aware of the difference in functionality offered. For example, the functionality of analogue and dig
48、ital speech transmission may be very different. Advanced PMR systems make use of digital voice transmission which provides on average a superior speech intelligibility and quality compared to conventional analogue speech transmission. Digital voice transmission also permits privacy by encryption whi
49、ch can be more easily implemented and is much more secure than is the case with analogue systems. Additionally all kinds of data transmission are possible ranging from short pre-coded messages to more demanding requirements like text and data files and even pictures. For special applications, the technology allows the possibility of slow motion video with limited resolution. STD-CEPT ERC REPORT 52-ENGL 3997 2326434 0033323 T84 W ERC REPORT 52 Page 13 - digital versus analowe: other considerations For comparing analogue systems, the static C/I has to be replac
