Fascicle II.3 _ Rec. E.523 7 All drawings appearing in this Recommendation have been done in Autocad. Recommendation E.523 STANDARD TRAFFIC PROFILES FOR INTERNATIONAL TRAFFIC STREAMS The worldwide nature of the international telephone network, spanning as it does all possible time zones time zones, has stimulated studies of the traffic streams between countries in different relative time locations. These studies have led to the development of standardized 24_hour traffic profiles which, theoretically based and verified by measurements, would be useful for engineering purposes. In fact, these concepts can be applied to a variety of network situations: i)variable access satellite working where a large number of traffic streams with possibly differing traffic profiles share the pool of satellite circuits; ii) combining of traffic streams on groups of terrestrial circuits which may be either high_usage or final choice routes; iii) detour routing of traffic between origin and destination countries to take advantage of prevailing low load conditions on the detour path. In developing any such applications, account must be taken of the International Routing Plan International Routing Plan (Recommendation E.171 [1]) and of accepted accounting principles (Recommendation D.150 [2]). It must be recognized that the preferred basis for dimensioning consists of traffic profiles based on real traffic. Nevertheless, many countries have found the standard profiles presented in this Recommendation very useful where streams are too small to obtain reliable measurements or where no measurements are available. For both_way profiles both_way profiles, two equivalent methods of presentation are given in chart and tabular form. In Figure 1/E.523 hour_by_hour traffic volumes are shown in diagrammatically as percentages of the total daily traffic volume; such percentages are particularly convenient for tariff studies. In Table 1/E.523, hourly traffics are expressed as percentages of the busy hour busy hour traffic, and this is convenient for engineering purposes. Time zone differences Time zone differences are given in whole hours only. Directional profiles Directional profiles are given in Tables 2/E.523 and 3/E.523. Although tables are given for both_way and directional traffic streams, it must be emphasized that at this stage only the both_way profiles can be regarded as soundly supported by measurement. The directional profiles are theoretically based and supported by some measurements, but should be used with caution until adequate verification has been achieved. The theoretical basis for the profiles presented here is contained in Annex A. It depends on a convenience function f(t) which represents the profile of profile of local daily traffic, where of course no time zone difference exists. The function f(t) used for computation of the standard profile was derived by mathematical manipulation of measurements of the Tokyo_Oakland and Tokyo_Vancouver streams. Although these results have been supported by other measurements, it leaves open the possibility that the convenience function may vary from one country to another and that, strictly, these should be derived independently and then used to obtain a calculated profile for the international relation. It also seems that the convenience function for the country of destination should be given greater weight than that for the country of origin. These remarks suggest possible refinements, but are not quantified in this Recommendation. Figure 1/E.523 - CCITT 48101 TABLE 1/E.523 Standard hourly bothway traffic patterns Local time in the more westerly country BH 0 1 2 3 4 5 6 7 8 9 10 1112 13 1415 1617 18 1920 2122 23 % 0 5 5 5 5 5 5 1010 50 9010 9585 70 8585 7045 25 4040 3520 15 10 0 .0 1 5 5 5 5 5 5 1025 70 9510 9080 80 8580 6035 30 4035 2515 10 10 0 .0 2 5 5 5 5 5 5 2030 75 1010 9090 85 8565 4545 35 4030 2515 5 10 0 0 .0 3 5 5 5 5 5 5 2535 75 1095 1095 80 7050 6045 35 3025 15 5 5 10 0 0 .4 4 5 5 5 5 5 5 2535 65 8510 1085 60 5060 5540 25 2520 5 5 5 11 0 0 .5 5 5 5 5 5 5 5 2530 65 9510 9070 50 6060 5530 20 20 5 5 5 5 12 0 .4 10 6 10 5 5 5 5 5 2530 75 0 10 7555 60 6560 4025 15 5 5 5 5 5 13 0 .1 7 10 5 5 5 5 5 2535 80 1085 5570 65 6550 4020 5 5 5 5 5 10 13 0 .5 8 25 5 5 5 5 5 3545 95 1080 9590 75 6050 35 5 5 5 5 5 20 20 11 0 .7 9 40 5 5 5 5 5 3540 75 8010 9585 60 5535 5 5 5 5 5 2525 40 12 0 .1 10 40 5 5 5 5 5 3535 60 9510 9065 50 40 5 5 5 5 5 25 3050 55 12 0 .5 11 40 5 5 5 5 5 3025 75 1095 7055 35 5 5 5 5 5 2530 6570 60 12 0 .3 12 40 5 5 5 5 5 2035 80 1080 6540 5 5 5 5 5 20 3560 1080 65 11 0 0 .3 Note 1 _ The 24_hour profile of both_way traffic between any two countries is read from left to right from the appropriate row of the table; all time differences can be expressed in the range 0_12 hours. Each entry is expressed as a percentage of the busy hour traffic. Note 2 _ The more westerly country of a traffic relation is the one from which we can proceed eastwards to the other through time zones not exceeding 12 hours. Note 3 _ For network planning studies, UTC (Universal Coordinated Time) would normally be used so that all traffic streams are processed time consistently. Clearly if the more westerly country is W hours ahead of UTC (ignoring the international dateline), then the traffic at 0000_0100 UTC is obtained from the row corresponding to the time difference between the two countries at the column headed W. Alternatively, the first entry in the appropriate row gives the relative traffic intensity for the hour (24_W) to (25_W). Example: For the traffic stream between the U.K. (UTC + 1 hour) and the central zone of USA (UTC + 18 hours), the time difference is 7 hours and the USA is regarded as the more westerly country, hence W = 18. Thus from the table, the traffic during 0000_0100 UTC is 5 % of the busy hour traffic, and the busy hour is 1500_1600 UTC. Note 4 _ The column headed "BH %" gives the busy hour traffic volume as a percentage of the daily traffic volume. MONTAGE Time difference (in hours) between two countries TABLE 2/E.523 Diurnal distributions of eastbound international telephone traffic Local time in the more westerly country 0 1 2 3 4 5 6 7 8 9 10 11 12 1314 15 1617 18 19 2021 22 23 0 10 5 5 5 5 5 1010 50 9010 95 85 7085 85 7045 25 40 4035 20 15 0 1 5 5 5 5 5 5 1030 80 9510 90 80 8085 80 6035 30 40 3525 15 10 0 2 5 5 5 5 5 5 2540 85 1010 90 90 8585 60 4045 35 40 2520 15 5 0 0 3 5 5 5 5 5 5 4050 90 1095 10 95 8065 40 5545 35 25 2010 5 5 0 0 4 5 5 5 5 5 5 3550 70 8510 10 85 6040 50 5040 25 20 15 5 5 5 0 0 5 5 5 5 5 5 5 3040 70 9510 90 65 4550 50 5025 20 15 5 5 5 5 0 6 10 5 5 5 5 5 4045 85 1010 65 45 5555 50 3020 15 5 5 5 5 5 0 0 7 10 5 5 5 5 5 4050 90 1075 40 60 5555 40 3010 5 5 5 5 5 10 0 8 25 5 5 5 5 5 5565 10 1070 90 85 7045 35 25 5 5 5 5 5 20 20 0 0 9 50 5 5 5 5 5 4045 70 7510 10 85 5550 35 5 5 5 5 5 25 35 60 0 0 10 65 5 5 5 5 5 4545 60 9510 90 60 4535 5 5 5 5 5 2530 75 10 0 0 11 65 5 5 5 5 5 4040 75 9080 55 40 25 5 5 5 5 5 20 2580 10 95 0 12 55 5 5 5 5 5 2040 65 7050 40 20 5 5 5 5 5 20 25 7010 90 80 0 Note _ This table is based on p = 1.4, q = 0.6, i.e. greater weight is given to the convenience function of the called party (see Annex A). MONTAGE Time difference (in hours) between two countries TABLE 3/E.523 Diurnal distributions of westbound international telephone traffic Local time in the more westerly country 0 1 2 3 4 5 6 7 8 9 10 11 12 1314 15 1617 18 19 2021 22 23 0 10 5 5 5 5 5 1010 50 9010 95 85 7085 85 7045 25 40 4035 20 15 0 1 5 5 5 5 5 5 1020 60 9510 90 80 8085 80 6035 30 40 3525 15 10 0 2 5 5 5 5 5 5 1520 65 1010 90 90 8585 70 5045 35 40 3530 15 5 0 0 3 5 5 5 5 5 5 1020 60 1095 10 95 8075 60 6545 35 35 3015 5 5 0 0 4 5 5 5 5 5 5 1520 60 8510 10 85 6060 70 6040 25 30 25 5 5 5 0 0 5 5 5 5 5 5 5 2020 60 9510 90 75 5570 70 6035 20 25 5 5 5 5 0 6 10 5 5 5 5 5 1015 65 1010 85 65 6575 70 5030 15 5 5 5 5 5 0 0 7 10 5 5 5 5 5 1020 70 1095 70 80 7575 60 5030 5 5 5 5 5 10 0 8 20 5 5 5 5 5 1525 90 1090 95 95 8075 65 45 5 5 5 5 5 20 20 0 9 25 5 5 5 5 5 3035 80 8510 95 85 6560 35 5 5 5 5 5 20 20 25 0 10 10 5 5 5 5 5 2525 60 9510 90 70 5545 5 5 5 5 5 2530 25 10 0 11 15 5 5 5 5 5 1010 65 9510 80 65 45 5 5 5 5 5 25 3540 35 25 0 12 20 5 5 5 5 5 2025 70 1090 80 55 5 5 5 5 5 20 40 6570 50 40 0 Note _ This table is based on p = 1.4, q = 0.6, i.e. greater weight is given to the convenience function of the called party (see Annex A). MONTAGE Time difference (in hours) between two countries ANNEX A (to Recommendation E.523) Mathematical expression for the influence of time differences on the traffic flow A telephone call is initiated when a person wishes to call someone else, but both parties have to be on the line before the call is established. It is considered that a telephone call is made at a time which tends to be convenient for both the calling and called parties. The degree of convenience for making a telephone call is considered to be a periodical function of time t, whose period is 24 hours. When the time difference between both parties is zero, the degree of convenience is denoted by f(t), where t is local standard time. The graphic shape of the basic function f(t) will be determined by the daily pattern of human activities, and will resemble, or fairly closely coincide with, the hour by hour traffic distribution in the national (or local) telephone network. It is assumed that the hourly traffic distribution hourly traffic distribution Ft(t), when a time difference of t hours exists between the originating and called locations, is expressed as the geometric mean of convenience functions convenience functions of two locations t hours apart: where (A_1) The sign of t is positive when the time at the destination is ahead of the reference time, and negative when the time of destination is behind the reference time. The distribution of equation (A_1) represents the sum of the outgoing and incoming traffics. Expressions for the one_way hourly traffic distributions can also be obtained by extending the concept of convenience function as follows. Define convenience functions both for the caller f0(t) and for the called party fi(t).Then the one_way traffic distributions of east_bound and west_bound telephone calls, for the case of t hour time_difference, are similarly expressed as follows: (A_2) dt (A_3) where t is the local standard time of the west station and t is positive. It is natural that a caller makes a call considering the convenience of the called person, and therefore the convenience function of the called person fi contributes more than the convenience of the caller f0 to the directional distribution F. They can be written as follows: (A_4) where p > q and p + q = 2, and where k1 and k2 are normalizing coefficients to ensure that: . As to the values of p and q in equation (A_4), it has been found empirically that the convenience of the called side p is considerably larger than that of originating side q, and appropriate values are roughly p = 1.4 and consequently q = 0.6. References [1] CCITT Recommendation International telephone routing plan, Rec. E.171. [2] CCITT Recommendation New system for accounting in international telephony, Rec. D.150. Bibliography CASEY (J. Jr.) and SHIMASAKI (N.): Optimal dimensioning of a satellite network using alternate routing concepts, Sixth International Teletraffic Convention, Munich, 1970. RAPP (Y.): Planning of a junction network with non_coincident busy hours, Ericsson Technics, No. 1, 1971. CABALLERO (P. A.) and DIAZ (F.): Optimization of networks of hierarchical structure with non_coincident busy hours, Seventh International Teletraffic Convention, Stockholm, 1973. OTHA (T.): Network efficiency and network planning considering telecommunication traffic influenced by time difference, Seventh International Teletraffic Convention, Stockholm, 1973.