ITU-T Q 45 BIS-1988 TRANSMISSION CHARACTERISTICS OF AN ANALOGUE INTERNATIONAL EXCHANGE《模拟国际交换机的传输特性》.pdf

1、INTERNATIONAL TELECOMMUNICATION UNION)45G134 1 G0BISTELECOMMUNICATIONSTANDARDIZATION SECTOROF ITU%.%2!,G0G02%#/-%.$!4)/.3G0G0/.G0G04%,%0(/.%37)4#().G0G0!.$G0G03).!,).).4%2.!4)/.!,G0G0!54/-!4)#G0G0!.$3%-)G13!54/-!4)#G0G07/2+).42!.3-)33)/.G0G0#(!2!#4%2)34)#3G0G0/- the nominal relative level at the exc

2、hange output port at point D is designated Lo.Fascicle VI.1 - Rec. Q.45 bis 3For the RETURN direction of transmission the input port with its nominal relative level Liis located at pointD and the output port with its nominal relative level Lois located at point A.The values of the nominal relative l

3、evels Liand Lomay be different for each 2-wire path of a 4-wireconnection through the analogue international exchange.1.2.3.2 nominal relative levels at virtual analogue switching pointsThe nominal relative levels at the virtual analogue switching points are defined to assure stability and to assist

4、maintenance procedures 3 4.The difference of the nominal relative level at the end of the incoming 2-wire path and the nominalinternational through-connecting level, which is by convention -3.5 dBr, is the stability loss T assigned to a 2-wire pathof a 4-wire circuit. By the value of this loss T the

5、 nominal transmission loss of a 2-wire path of a connection through ananalogue international exchange is referred to its virtual analogue switching point.1.2.4 Measurement conditions1.2.4.1 reference frequencyThe nominal reference frequency, on which relative levels, transmission loss, loss-frequenc

6、y distortion etc. are4 Fascicle VI.1 - Rec. Q.45 bisbased, is 800 Hz or 1000 Hz alternatively 5.Note - Since 1020 Hz is the recommended nominal frequency for techniques using digital processes thisfrequency should be preferred to harmonize into the evolving digital network 6.1.2.4.2 impedanceMeasure

7、ments shall be made under nominally matched conditions, i.e. the exchange boundaries are terminatedwith their nominal exchange impedance.1.2.4.3 test levels at exchange boundariesAt the nominal reference frequency, test levels are defined in terms of the apparent power relative to 1mW. Atfrequencies

8、 different from the nominal reference frequency, test levels are defined as having the same voltage as thetest level at the nominal reference frequency. Measurements are based on the use of a test generator with a frequency-independent e.m.f. and which has an impedance equal to the nominal impedance

9、.1.2.5 Transmission loss1.2.5.1 nominal transmission lossA connection through an analogue international exchange (see Figure 1/Q.45 bis) is established by connectingan input port located at one exchange boundary to an output port located at another exchange boundary in bothdirections.The nominal tra

10、nsmission loss of a 2-wire path of a connection through an exchange is equal to the differenceof the relative levels at the input and the corresponding output:NL = (Li- Lo) dBNote - The nominal transmission loss of the exchange may be different in the GO and RETURN direction oftransmission.1.2.6 los

11、s distortion with frequencyThe loss distortion with frequency is the logarithmic ratio of output voltage at the reference frequency, U(Ref),divided by its value at frequency f ,U(f):(See Supplement No. 1 to Volume VI, Fascicle VI.5, CCITT 6.)2 VF-parameters of a connection through the exchange2.1 Im

12、pedance2.1.1 Nominal valueThe nominal impedance at the input and output ports located at points A and D of Figure 1/Q.45 bis shall be600 ohms, balanced.2.1.2 Return lossThe return loss of one port located at point A or D of Figure 1/Q.45 bis has to be measured against thenominal impedance whilst all

13、 other ports of the connection through the exchange are terminated with the nominalimpedance.At any frequency from 300 to 600 Hz the return loss should be not less than 15 dB. The corresponding valuefrom 600 to 3400 Hz should be not less than 20 dB.Fascicle VI.1 - Rec. Q.45 bis 52.1.3 Impedance unba

14、lance about EarthThe impedance unbalance about Earth is measured as longitudinal conversion loss (LCL) according toFigure 1/O.9 16 and as longitudinal conversion transfer loss (LCTL) according to Figure 2/O.9 16 at the interfaceslocated at points A and D of Figure 1 /Q.45 bis using Z = 600 ohms and

15、ZL = 150 ohms.The measured values should not be worse than:300- 600 Hz: 40 dB600-3400 Hz: 46 dBNote - Some Administrations guided by their knowledge of local conditions may feel a need to specify a valueof impedance unbalance about Earth for a lower frequency, for instance, 50 Hz.2.2 Values of relat

16、ive levels Liand Lo2.2.1 Basic nominal valuesBasic nominal values for the input level Liand the output level Loof a connection through an analogueinternational exchange are given in Table 1/Q.45 bis. For the purpose of demonstration, these values are valid under thefollowing hypothetical assumptions

17、:- there is no transmission impairment between the points X and A and the points D and Y ofFigure 1/Q.45 bis;- the nominal relative levels Liand Loare determined by the corresponding nominal relative levels of thechannel translating equipment recommended for two cases in Table 2/G.232 7 corrected by

18、 the nominalper-channel loss of the international circuit, T = 0.5 dB.TABLE 1/Q.45 bisBasic nominal values of relative levels at the exchange boundaries of a connectionthrough an analogue international exchangeRelative level Channel translating equipmentCase 1 Case 2LiLo+ 4 dBr-14.5 dBr+ 7 dBr-16.5

19、dBrNominal values of relative levels will differ in practice from these basic nominal values by the impact ofvarious equipment being inserted and the necessary cabling to interconnect the channel translating equipment to theexchange boundaries. Due account should be taken of this impact in specifyin

20、g corresponding nominal relative levels,especially by cable length between points X and A and points D and Y in Figure 1/Q.45 bis.2.2.2 Offset of mean actual valuesThe actual value of the output relative levels depend on the tolerances of components, i.e. mainly attenuationpads, and on the routing o

21、f a connection through an exchange via the switchblock (Points B and C in Figure 1/Q.45bis).The offset of the mean value of the distribution of the actual output relative levels Loshould be very close tozero but does not need to be specified.6 Fascicle VI.1 - Rec. Q.45 bis2.2.3 Dispersion of actual

22、valuesThe dispersion of actual values of the output relative level Lois mainly due to the diversity of paths in theswitchblock. The standard deviation of a representative distribution of the actual output relative levels measured at thenominal reference frequency should be as small as practicable. F

23、or purpose of calculation a value of 0.2 dB may beassumed.In order to confirm this value, it is considered sufficient that for purposes of design and acceptance testing, thedifference between the actual relative output levels at the nominal reference frequency of the shortest and longest pathsfrom p

24、oint B to point C in Figure 1/Q.45 bis in no case exceeds 0.8 dB. For a practical assessment of the average valueof the actual relative output level, the influence of the switchblock between points B and C can be achieved using thearithmetically computed mean of the maximum and minimum actual relati

25、ve output levels.These values apply for connections routed directly, and once only, through the switchblock. If special re-entrant trunking arrangements are used, requiring the connection to pass through the switchblock twice (this may be aconvenient way to extend the availability of the switching n

26、etwork or to introduce additional equipment, e.g. echosuppressors), the distribution of the actual relative output levels will be increased to lower values. In view of this, there-entrant technique should not be used to such an extent as to decrease significantly the mean value of the actualrelative

27、 output level distribution.2.3 Basic nominal values of transmission lossIn accordance with the definition in 1.2.5.1 and the basic nominal values of relative levels quoted in 2.2.1the following basic nominal values of transmission loss result for the purpose of demonstration:case 1 : NL = +4 dB - (-

28、14.5) dB = 18.5 dBcase 2 : NL = +7 dB - (-16.5) dB = 23.5 dB.2.4 Response to frequency and input level2.4.1 Loss distortion with frequencyThe loss distortion with frequency according to the definition in 1.2.6 measured on any 2-wire path ofconnection through the exchange between points A and D of Fi

29、gure 1/Q.45 bis should lie within the following limits:300- 400 Hz: -0.2 dB to +0.5 dB400-2400 Hz: -0.2 dB to +0.3 dB2400-3400 Hz: -0.2 dB to +0.5 dB.2.4.2 Variation of output level with input levelThe actual output level measured on any 2-wire path of a connection through the exchange between point

30、s Aand D of Figure 1/Q.45 bis should follow the input level with a variation not more than 0.2 dB in the range of the inputlevel from -40 dBm0 to +3.5 dBm0, using the reference frequency.2.4.3 Group delay distortion with frequencyAccording to the definition of group delay 9, the group delay distorti

31、on measured on any 2-wire path of aconnection through the exchange between points A and D of Figure 1/Q.45 bis over the frequency band 600 to 3000Hz should not exceed 100 microseconds.2.4.4 IntermodulationThe intermodulation products shall be measured on any 2-wire path of a connection through the e

32、xchangebetween points A and D of Figure 1 /Q.45 bis.Fascicle VI.1 - Rec. Q.45 bis 7The intermodulation products to be taken into account for end-to-end multifrequency signalling and for datatransmission are those of the third order, of type (2f1-f2) and (2f2-f1) where f1and f2are two signalling freq

33、uencies.For a measurement of the intermodulation products, the two frequencies applied to an input are f1= 900 Hzand f2= 1020 Hz (see 8). With each frequency f1and f2at a level of -6 dBm0, the difference at the output between thelevel of either frequency f1or f2and the level of either of the intermo

34、dulation products at (2f1-f2) or (2f2-f1) should beat least 40 dB.2.5 NoiseFor a 4-wire international exchange, noise measurements should be performed on a connection through theexchange between points A and D of Figure 1/Q.45 bis during the busy hour 10. Each port should be terminated with600 ohms.

35、 The noise should be measured at the output port of each 2-wire path and should be referred to a point ofzero relative level. Thus in Figure 1/Q.45 bis the noise in the 2-wire path of the GO direction is measured at point Dand the noise in the 2-wire path of the RETURN direction is measured at point

36、 D and the noise in the 2-wire path of theRETURN direction is measured at point A. A sufficient variety of connections should be chosen to ensure that themeasurements are representative of the various possible routes through the exchange.2.5.1 Weighted noiseThe mean value of the psophometrically wei

37、ghted noise over a long period during the busy-hour should notexceed - 67 dBm0p (200 pW0p).2.5.2 Unweighted noiseUnweighted noise has to be measured with a device having a uniform response curve throughout thefrequency band 31.5 Hz-16 kHz 11.The mean value of the unweighted noise over a long period

38、during the busy-hour should not exceed -40 dBm0(100,000 pW0).2.5.3 Impulsive noiseFor measurement procedure of impulsive noise see Annex A of this Recommendation.Noise counts should not exceed 5 counts in 5 minutes at a threshold level of -35 dBm0.Note - Figure 3/Q.45 bis shows the maximum number of

39、 impulsive noise counts acceptable in a 5-minuteperiod.2.6 CrosstalkCrosstalk should be measured in exchanges at a frequency of 1100 Hz in accordance withRecommendation G.134 12.2.6.1 Crosstalk between different connections(Inter-connection crosstalk)In an analogue international 4-wire exchange the

40、signal to crosstalk ratio measured at points A and D ofFigure 1/Q.45 bis between any 2-wire paths of different 4-wire connections through the exchange should be 70 dB orbetter.This limit of 70 dB should normally apply to the most unfavourable case, in which two connections haveparallel paths through

41、out the exchange. It should be noted that this does not occur in practice, because normal cablinglayout in such that when, at one switching stage, two connections use adjacent switches, in the following stage the twoconnections generally use switches which are not adjacent.2.6.2 Go-to-return crossta

42、lk of the same connection(Intra-connection crosstalk)The signal-to-crosstalk ratio between the GO and RETURN 2-wire path of the same 4-wire connectionthrough the exchange should be 60 dB or better.8 Fascicle VI.1 - Rec. Q.45 bis3 Use of cables specified by the IECThe cables for telephone exchanges i

43、n accordance with IEC (International Electrotechnical Commission)publication 189 13 will meet the electrical characteristics required by the CCITT (especially as regards crosstalk) forordinary exchanges, but this may no longer hold good for larger exchanges with considerable lengths of cable.In acco

44、rdance with Recommendation G.231 14, it will be for the Administrations or the contractors to checkwhether standard cables will be satisfactory in equipping an exchange which requires telephone cables of exceptionallength.ANNEX A(to Recommendation Q.45 bis, 2.5.3)Procedure for impulsive noise measur

45、ementA.1 A test circuit should be formed by setting up a connection across the switching unit and terminating theconnection on the exchange input by the nominal impedance and on the exchange output by the impulse measuringdevice in parallel to the terminating nominal impedance. Those terminated port

46、s should be at points A and D in thediagram of Figure 1/Q.45 bis which includes the switching equipment of the exchange. Where it is the desire of anAdministration, measurements may be made at points X and Y if precautions are taken to ensure that the result applyonly to the automatic switching equi

47、pment, signalling equipment, echo suppressors, relay sets, pads and cabling of theexchange.Fascicle VI.1 - Rec. Q.45 bis 9A.2 The measurements should be made using the device specified in Recommendation O.71 15. The600-3000 Hz filter network should be in the circuit.A.3 The measurements should be ma

48、de at times when the probability of noise occurring is at its highest, that isnormally during the busy-hour.A.4 The time of observation for each test should be five minutes.Note - The number of different test circuits set up through the exchange for measuring should take intoaccount the size and com

49、plexity of the switching unit and should be representative for all various routes through theexchange. See also the documents cited in 15 and 17.References1 CCITT Recommendation Transmission performance objectives and Recommendations, Vol. III. Fascicle III.1,Rec. G.102, 3 and 4.2 CCITT Recommendation Circuit testing, Vol. IV, Fascicle IV.1, Rec. M.110, 1.3 CCITT Recommendation The transmission plan, Vol. III, Fascicle III.1, Rec. G.101, 2.1 and 5.4.4 CCITT Recommendation Loudness ratings (LR) in an international connection, V