FI79920B - KOPPLINGSANORDNING FOER TELEANORDNINGAR, SPECIELLT FOER TELEFONFOERMEDLINGSANORDNINGAR MED EN CENTRAL ELLER DECENTRALISERAD KOPPLINGSMEKANISM FOER DATABEHANDLING. - Google Patents

Abstract

1. Circuit arrangement for telecommunication installations, espectally for telephone exchanges in which centralized and/or partly centralized information-processing switching devices with a limited capability with regard to the information-processing capacity are provided, as well as measuring facilities (B) for determining the loadability of such switching devices, which integrate the constantly fluctuating load over periods of time which are in each case great enough to ensure a determination result with adequate result accuracy, and in which a counting facility (W), serving to establish the information-processing traffic load in each case of a switching device and to detect and avert information-processing traffic overloads, supplies with as little time delay as possible an instantaneous value of the information traffic load which is current and is produced by comparison in each case of a periodically repeatedly obtained counting result with the determination result received by it and stored in it, in order to avert without great delay such overloads in the event of an increase in the continuous counting results over the determination result by the rejection of information-processing orders, and in which, after aversion of an information-processing traffic overload and after a dropping of the periodically obtained counting results and an uninterrupted remaining of the same below the determination result during a certain minimum waiting time, during which increasingly higher values are established for the determination result deciding on loadability, due to the return of the load, the generation and/or relaying and/or evaluation of the determination result - the latter by continuous comparison of the periodically repeatedly obtained counting results with the determination result - is interrupted, and in which the determination result which is last established and stored in the counting facility (W) is replaced by a lower normal value, remaining unchanged after the aversion operation, until the generation and/or relaying and/or evaluation of the determination results is reactivated due to a build-up of information-processing orders or an increase in the continuous counting results over a comparative value lying substantially above the normal value, characterized in that, upon dropping of the periodically obtained counting results, a value existing at the time the determination result was reached or undershot is stored, and in that, after elapse of the minimum waiting time, during which the periodically obtained counting results remain below the determination result, this stored value serves as normal value, and in that until a new determination result is obtained, the periodically obtained counting results on the basis of this normal value serve to establish the continuous instantaneous values of the information traffic load.

Description

1 79920

Switching device for telecommunication equipment, in particular call switching equipment with central and / or distributed information switching equipment

The presentation given at the 9th International Teletraffic Congress (ITC) in October 1979 in Torremolinos, Spain (Conference Paper ITC-9, Somoza / Guerrero 1-7) (see also IEEE Transactions on Communications ", September 1973, p. 1027 ff.) Is known for telecommunication equipment, in particular switching equipment for telecommunication equipment, in particular for call transmission equipment with central and / or part-centric information processing equipment with limited performance in terms of information processing capacity, as well as measuring capacity. to measure the load capacity, which integrates a continuously varying load for time periods which are in each case large enough to guarantee a determination result with sufficient result accuracy, and a calculator for each time the information processing traffic load of one switching equipment is to detect and overcome such overloads, whereby this calculator produces with the smallest possible time delay the current and current value of the information processing traffic, which is determined by comparing the calculation result with the determination result received and stored therein, to implement overload prevention without delay.

Thus, in this type of switching device, a distinction is made between, on the one hand, a load measurement which produces the load determination result of the central switching equipment and, on the other hand, a load analysis of the information processing traffic, which indicates the current state of the load. While the determination result articulates the short-time variations in the load, the result of the continuous information processing traffic load study should be as timely as possible, i.e. it should produce a result, 2,79920, in which continuous load variations occur. Said determination result of the load-capacity measurement indicates a measurement quantity to which the continuously determined current result of the load of the information processing traffic is compared. The result of the load measurement is a setpoint, ie a holding value for which traffic load is to be fed to the central control equipment. The result of the load on the information processing traffic is actually the current actual value of the traffic load occurring at each moment. The comparison of this actual value with the holding value is in each case an indication of whether the traffic load supplied to the central control equipment can be increased compared to the current value or whether it must be reduced.

The load capacity of each such switching equipment depends in principle on its structure (structure and software development) and on the quality and scope of the information processing events continuously transferred to it. In addition, the composition of the various information processing events may shift during the operation of the switching equipment. Thus, for example, provided that the switching equipment is the central control equipment of the call transfer equipment, it may occur that at certain times of the day mainly local connections have to be established, while at certain other times of the day mainly long-distance connections have to be established. Furthermore, it may happen that more short-term calls occur at certain times of the day and more long-term calls at other times of the day. It may still be the case that more special services are required at certain times of the day than at other times of the day. It may also be the case that the proportion of premature disconnection events at certain times of the day is higher than at other times of the day. Thus, the load capacity of such control equipment is not constant.

In order to optimally load the switching equipment, it is expedient to take into account the different and different load capacities determined due to different conditions during the different operating times. For this reason, it is customary to equip such switching devices with measuring devices for determining the load capacity, which always integrate a variable load for time periods which are in each case large enough to guarantee an analytical result with sufficient accuracy, so that short-term fluctuations can no longer be affected. Such a measuring device can operate, for example, on the basis that each control device generates a demand signal in the respective processing of all the information processing tasks that occur, which is fed e.g. to this measuring device. This measuring device calculates the number of such demand signals present in the control equipment in question for a certain period of time, for example one second or 4 seconds.

Thus, in the known case, the load measurement to produce said hold value takes place over relatively large intervals, each lasting several seconds, while the load measurement to determine continuous actual values takes place in the same way by calculation, but at much smaller intervals, each of which is only a small fraction of each interval. - If, during this period, this figure exceeds the limit value corresponding to and defined by the holding value, then all other tasks will be rejected until the end of the period. Thus, the load on the switching equipment and consisting of the individual tasks to be performed (calculations, control events, etc.) is regulated. In this case, it is necessary that the duration of each time period is also large enough so that the load adjustment does not take place in the roughest jump, but in fine enough.

The load results obtained by the calculation limit the number of tasks and thus the flow of information processed in the calculator or they can be stopped temporarily altogether. Thus, the traffic load is optimally adapted to the load capacity of the switching equipment, for example the calculator, in order to obtain the highest possible processing capacity and to avoid, as far as possible, overloads which are known to cause significant temporary operating limitations or malfunctions in such switching equipment. Thus, in the use cases presented in the presentation, this is improved by dynamically adapting the traffic load according to different types and thus repetitive information processing tasks fed to the switching equipment to the same variable load.

The latter presented adaptation of traffic load to variable load is relatively accurate in the range of full load and overload. However, if the traffic load of the switching equipment decreases as the number of information processing tasks per unit time falls substantially below the traffic load capacity, then it is entirely characteristic of the control that the reference value obtained from the load measurement increases by calculating the number of continuously processing information processing tasks per time unit. Thus, in times when there is a relatively small traffic load, a larger supply of information processing tasks is possible. This is appropriate in the sense of good utilization of processing capacity, provided that the traffic load varies greatly, as this condition can be used to compensate for variations and the switching equipment in question is always well balanced in terms of load. However, in the case of long periods with relatively low traffic congestion, an extreme increase in the reference value may jeopardize the switching equipment so that it may be completely overloaded with a sudden incoming load jump on information processing tasks before the reference value can be adjusted down accordingly. However, for certain load conditions in many switching equipment, e.g. central control equipment for call transfer equipment, it is just typical that a stronger load jump is preceded by a relatively long load with a relatively low load, e.g. when the tariff changes from a normal tariff to a low tariff.

5 79920 In order to eliminate this problem, the switching device disclosed in DE-OS 33 85 575 has provided that it is prepared for a sudden, strong and permanent increase in this load with respect to time periods with a relatively low information processing load, which is achieved by when the load decay limit is uninterrupted after the end of a certain minimum quarantine period, during which the most recently determined analytical result stored in the calculator is replaced by a constant average, and that the generation and / or transfer and / or interpretation of the analytical result is used anyone back to action when the current value triggers or responds, prepares to overload congestion. The above-mentioned constant average is a calculated and uniquely determined normal value.

The invention relates to a similar switching device, namely a switching device for telecommunication equipment, in particular for call transmission equipment, which is provided with central and / or part-centric information processing switching devices with continuous performance limited in terms of information processing capacity, and measuring devices for such switching devices. for varying load time periods which are in each case large enough to guarantee a determination result with sufficient result accuracy, and in which to calculate the load of information processing traffic of one switching device 79920 or repeatedly received calculation income as the continuous calculation results increase over the analysis result, such overloads can be counteracted without delay by rejecting the information processing tasks, and in which for the dominant test result due to load recovery, the production and / or sub-delivery and / or interpretation of the test result is cut off - the latter continuously comparing the calculation results with the accumulation of tasks or when the continuous calculation results increase substantially above the reference value above the normal value, the production and / or sub-delivery and / or interpretation of the analytical result is reactivated.

As already mentioned, in said application, the constant average is a normal value which has been calculated on the basis of traffic theory and has been determined on a unique basis. It may now be the case that the practical operating conditions differ considerably from the original conditions with respect to the conditions assumed for traffic theory calculations and determined by traffic simulations, which may of course also be due to the fact that the operating conditions may change gradually but permanently.

When the standard average of the load capacity deviates significantly from the operational reality, very unfavorable conditions may be created for initiating the adjustment of the reference value stored in the calculator and determining the control of charges. If the constant average is too high, then the reference value stored in the calculator determined by this 7,79920 causes the overload control to start relatively late. If, on the other hand, this average is too low, it may happen that this control leads to the control of incoming reservations when the congestion situation has not even occurred yet.

The object of the invention is to ensure in a switching device of the latter type that, in the event of an overload situation, control measures are initiated at the right time, but that nothing is left unused, especially when the requirements are high.

The invention solves this problem by storing a value existing at the time the assay result is reached or undershot when the periodically calculated calculation results are calculated, and that at the end of the minimum quarantine period during which the periodically obtained calculation results remain below the assay result, this stored value acts directly or by several such stored values. , and that the calculation results obtained periodically until a new determination result is generated operate on the basis of this normal value to determine the continuous instantaneous values of the information traffic load.

The drawing shows an embodiment of the invention only in parts which substantially facilitate its understanding. The description first deals with the general functions of a PCM call transfer apparatus operating in accordance with the invention. Only later will the special features of this transmission device according to the invention be explained.

The switching field formed as three switching stages RZE, R and RZA has a rather large number of switching field connections on the input side, one of which is shown and marked with the letter A. The switching field consists of several switching field parts. Each switching field interface always comprises a pair of β 79920 time multiplex lines, one time multiplexing line operating for signal transmission to the switching field and the other time multiplexing line operating for signal transmission from the switching field.

Thus, the switching field interface A includes a time multiplex line A1 and a time multiplex line A2. For each time multiplex line, the direction of transmission of the respective signal is indicated by arrows.

The switching multiples of the first switching stage RZE are combined time-station-state-station multiples, as can be seen from the symbols shown. The switching multiples of the middle switching stage R are state station multiples. The switching multiples of the last switching stage RZA are again state-station-time-station multiples. A relatively large number of multiples are arranged for each switching stage, although only three multiples of the type shown are shown per switching stage. The switching multiples of the different switching stages are connected to each other via intermediate lines as shown in the drawing. These intermediate wires are time multiplex intermediate wires. Likewise, the switching interfaces are of the time multiplex type.

Connection groups, e.g. connection line LTG1 (Line Trunk Group), are connected individually to the switching field connections, for example to the connection A of the switching field part K via time multiplex lines designed for both communication directions, for example the time multiplex line ltgl. Such an interface group can be connected on the input side with subscriber lines (analog), analog lines and PCM lines (alternatively and in combination). The subscriber station connected via the subscriber line is marked with the symbol T1. The interface group includes encoders, decoders, multiplexers, demultiplexers and the equipment needed to carry out the so-called BORSCHT function (cf. NTZ Bd. 33/1980, Heft 10, pages 646-652 and 1978 International Zurich Seminar on Digital Communications, Proceedings IEE Catalog, Nr. 78 CH 1325-0 ASST, pages B2-1, A4.1, 9 79920 also DE patent applications P 31 00 811.9, P 31 04 002.0 and P 31 06 903.7.

The drawing shows e.g. connection group LTG1. Its structure and mode of operation are described in detail in DT-OS 28 26 113 starting from the bottom of page 6. Further reference is made to the supplement to the 4th volume (1981) of the magazine "telcom report". Other explanations require this explanation to be known in advance and are limited in the present case to very essential connections.

The central control equipment ZW works e.g. to establish message connections via the switching field. It shall determine the transmission information required to establish each such connection. As is known, the transmission information of the connection to be switched on accurately indicates the passage of the connection in question via the switching field, i.e. the switching times and intermediate lines passed by the connection, as well as the channels reserved in them in each case. This transmission information is processed by means of the central control equipment, which thus performs e.g. also the functions of the path finder in addition to the reservation memory, and are transferred to the switching field control device KE of the switching field K. The data stream transmitting the transmission data passes from the central control equipment ZW to the switching field control device KE via the central data line D via a buffer device MB connected to the central control line D, which acts for intermediate storage and recoding of data and for adjusting transmission rates. The buffer device may be limited to some of these functions or only to one of these functions at a time. Such a buffer device is described in detail, for example, in DE-PS 15 37 849.

In addition to the central control equipment, a rather large number of distributed control devices GP1 to GPn are presented, which are arranged individually in each case for the connection groups LTG1 to LTGn. These control devices operate to perform all switching events 10 79920 in each of said connection groups.

These switching events include, in particular, all relay events that take place in these connection groups. Further details can be found in the aforementioned DE-OS 28 26 113.

The distributed control devices GP1 to GPn communicate with the central control device ZW via data channels which are individually connected per distributed control device via a switching field (K) to the buffer device MB. This buffer device, namely its data end ML, thus in each case leads one separate data channel to each distributed control device. For this purpose, the buffer device MB is connected at its data end ML via a time multiplex line m to one switching field interface. The buffer device is connected to one switching field terminal of the switching field in the same way as each of said connection groups LTG1 to LTGn. The time multiplex line m is thus connected to a switching field interface comprising a time-multiplex input of a first switching stage into a state-to-station-time-multiple and a time-multiplex output from a state-to-time-multiple of the last switching stage. The time multiplexer m is thus connected to a switching field interface as marked a.

Each data connection between the data end of the buffer device and a group of interfaces, for example the interface group LTG1, comprises a channel for transmitting data from the buffer device to that interface group and another channel for transmitting data in the opposite direction. Both of these data channels lead to a distributed switching device individually arranged for the interface group in question and vice versa.

The establishment and maintenance of the described data connections via the switching field takes place by means of the switching field control device KE in the same way as the establishment and maintenance of message connections, for example telephone connections. For this purpose, a switching memory is arranged in the known manner for the switching multiples of the switching field K, in which transmission data concerning one switching multiple in each case is stored. With the help of these hold memories, the necessary pass-through or counter, the necessary write events and the read events for the total memory of the time station multiples are available in the respective time stations. All other details concerning the structure and operation of the time-division multiplexer are required to be known herein and will therefore not be described in detail.

As shown, the transmission information necessary for establishing message connections is transferred from the central control equipment via the buffer device MB to the switching field control device KE. For this purpose, e.g. a data head MK arranged for a buffer device MB. For the sake of completeness, it should be noted at this point that several switching field control devices KE can also be connected to the buffer device MB. These several switching field control devices can be arranged individually for several switching field parts of a relatively large switching field. It is also possible to provide one switching field control device per switching stage in each case.

Since the above data connections must be switched from the buffer device MB via the time multiplex line m to each interface group by means of the switching field K of LTG1 to LTG, before communication connections, e.g. telephone connections can be established, a completely free switching field is generally available, i.e. all channels are unoccupied. For this reason, the establishment of data connections can take place arbitrarily according to a predetermined diagram, which remains unchanged at all times. For this reason, it is arranged that the switching field control device KE additionally contains, in the memory V arranged for it, the control information, i. i.e. transmission data for data channels. Thus, this transmission data does not have to be processed again when a new data connection is established with the central control equipment ZW in each case, nor does it have to be retransmitted via the buffer device MB. If the buffer device MB and the connection groups LTG1 to LTGn, i.e. the necessary data connections between the main control devices GP1 to GPn have to be re-established via the switching field K, then the central control system ZW only transmits a corresponding command to the buffer device MB, which it passes to the switching field controller KE and takes the transmission data stored in the memory V to perform the necessary brokerage transactions individually in succession.

As already mentioned in the aforementioned publication DE-OS 28 26 113, in addition to the communication connections between the subscriber stations and the connection line, the already mentioned control devices for the connection groups, e.g. device GP and on the other hand central control, are established via the time channel switch TSU data connections between ZW. These data connections continue to take place, as already shown, via the switching field K and the ai-camultiplex line m.

As already mentioned, the group control devices GP1 to GPn are distributed or partially central control devices compared to the central control apparatus ZW. All in all, these partially central control devices and the central control device are also called switching devices. Group control equipment and central control equipment have limited performance in terms of information capacity. A particular problem is the performance of the central control equipment.

As already stated, in addition to the central switching field K, several connection groups LTG1 to LTGn are provided. Each of these connection groups is in each case connected via a PCM line, for example ltgl, to the switching field connection of the switching field K, for example to connection A, as shown. This PCM line arranged per connection group leads there to a switching device TSU, the meaning and mode of operation of which are described in more detail in the aforementioned DE-OS 28 26 113. This switching device is a partial switching field through which both said data connections and connections from subscribers to mains are switched. . Of course, "connection lines" also means the channels of the connected PCM connection lines. The connection of these subscriber lines / connection lines and channels is presented in the literature sections already mentioned.

On the input side, various information arrives via the subscriber lines, connection lines and channels connected to the connection group, for example the call signals of the subscriber stations that want to connect, the dialing codes of these subscriber stations, but also the connection lines and channels. Subscriber interface connections for the connection of subscriber lines, line termination connections for the connection of incoming and / or outgoing busy local and long distance lines, internal contact groups, pick-up groups, billing pulse sensors (ZIGs) and the like are subscriber-specific and / or connection-specific connections.

The group control apparatus operates in a manner known per se to establish the reception of the above-mentioned incoming contact individual information. In addition, the group control equipment performs the processing or pre-processing of these incoming information, in each case arranged for the respective subscriber line, the communication line or the respective channel. In addition to this, a few such information are cached in the group control equipment in the individually arranged memory SP in each case, namely by means of the IOP of the input-output device. Furthermore, the group control equipment is responsible for transmitting signals and control signals over these lines (subscriber lines and communication lines and channels), for example call switching current pulses and listening tones via subscriber lines, as well as dial codes and line signals and the like via outgoing busy lines.

14 79920

Connections with respect to the connection establishment direction are connected via a partial switching field belonging to the connection group, both incoming from the line (subscriber line, connection line or corresponding, corresponding channel) in the direction of the central switching field K and also to such a line. for this purpose, the information required via this switching field is transferred from the group control equipment via the data connection already mentioned to the central control equipment ZW. Other connection switching takes place via an interface group (possibly the same interface group, but usually another interface group), for which purpose the respective group control equipment of the respective interface group must become operational or, respectively, the respective group control equipment of the respective interface groups must become operational.

Thus, there are various switching events for the group control equipment with respect to the connection establishment direction, namely, on the one hand, the switching of the subscriber line or the communication line (or the corresponding channel) to the switching field K and, on the other hand, the switching in the opposite direction. In addition, the through-connections via the partial switching field differ from those which pass on the other hand from the subscriber lines and to them and from the connection lines and to them (or, respectively, to the channels).

The information reception events and / or information handover events performed in connection with these respective switching-through events per connection require the respective group control equipment at least once per transit event. The quality of the traversal in question (incoming from the subscriber line or the trunk line or outgoing to the subscriber line or the trunk line) is decisive for which they receive in connection with a switching event, unchanged or pre-processed for the central control equipment common to them. Likewise, the central control equipment transmits information to the group control equipment in order to perform the outgoing connections required therein and to transmit signals and control signals via the respective lines (subscriber lines and communication lines or channels, respectively). The information to be transmitted from the group control equipment to the central control device per connection is essentially control and control information.

Thus, each call from the subscriber station side and each incoming reservation of the communication line (or, conversely, the channel) requires a certain amount of work power from the group control equipment in question. Each such work output is detected by the initial switching ID of the group control equipment in question. Such an initial switching field identifier is in connection with a subscriber line with a loop pulse output, for example a loop lock indicating a paging signal. Such an initial connection identifier in connection with the communication line is, for example, the charge signal of the c-conductor in the incoming charge. The same applies to individual channels.

These initial connection codes are thus different from each other, namely according to the quality of the line in question (subscriber line or connection line). In connection with the switching of a part of the connection to be established, they signal in each case a corresponding amount of the expected information processing traffic load for the group control equipment in question, in particular in connection with the reception, processing and / or transmission of information. Thus, for different parts of the connections to be switched, certain amounts in this respect are typical, which are similar in the case of similar parts of the connection to be connected, but different in the case of different parts. As for each group control equipment, the same applies to the central control equipment ZW. Depending on whether the switching through the field K of the connection 79920 16 is an internal connection, an outgoing long distance connection or an incoming long distance connection, the load of the information processing traffic of the central control equipment per switching is different, namely depending on the quality of the connection.

The already mentioned central control equipment ZW is the central information switching equipment. Distributed control devices GP1 - GPn are in part key switching devices for information processing. As already indicated, the data exchange between the central switching equipment GP1 to the GP on the one hand and the central switching equipment ZW on the other hand takes place via the buffer memory MB. The information to be transmitted to the central switching equipment is passed on via lines za, ze and h. The information transmitted from the central switching equipment ZW in part to the central switching equipment GP1 to GPn is transmitted via path c.

As already mentioned and as is quite well known,: * the central switching equipment has limited performance in terms of its information processing capacity. This performance relates to the number of information processing transactions to be performed per unit time. The load capacity of each central switching equipment depends in principle on its structure (structure and program design) and on the quality and scope of the information processing events that are continuously transferred to it. These are usually different. They appear confused; loadability therefore applies to the current mix of information processing tasks.

Experience now teaches that the composition of various information processing events can shift during the operation of the central switching equipment. This may be the case, for example, that local calls must be made primarily at certain times of the day, while long-distance calls must be made primarily at certain times of the day. There may still be 17,79920 so that there are more short calls at certain times of the day and long calls at certain times of the day. It may still be the case that more special services are needed at certain times of the day than at other times of the day. In addition to this, the share of connection attempts disconnected in each case before the complete connection is higher at certain times of the day than at other times of the day.

For the various reasons given above, the load capacity of the main switching equipment is not constant. In order to evenly load the main switching equipment, it is therefore expedient to take into account the different load capacities that exist at different operating times due to different conditions. For this reason, it is customary to equip such control devices with measuring devices for determining the load capacity, which integrates a continuously varying load for periods which are in each case large enough to guarantee a sufficient determination result with sufficient result accuracy, thus not affected by short-term fluctuations.

A measuring device operating for the above purpose can operate, for example, on the basis that the central control equipment produces a demand signal after the processing of all the information processing tasks that has taken place in each case, which is fed e.g. to this measuring device. This measuring device now counts the number of such demand signals present in the central switching equipment for certain equal time periods, for example four seconds. The drawing shows a measuring device B arranged on the central switching device ZW and connected to it via line n. Thus, from this key switching device, it receives a demand signal via this line n each time it occurs. The measuring device B calculates the level signals 79920 in a manner known per se. From the central timing device Zg, it receives a time stamp signal every 4 seconds via the line t1. This time stamp signal marks the end of the measurement time and at the same time the beginning of the next measurement time. The measuring device B now determines the number of demand signals arriving between the two such time stamps of the respective central switching device in question. If no information processing tasks occur during a relatively long period of time, then these requirement signals are transmitted at time intervals which, on average, correspond to the information processing event in each case. It is also possible to report, instead of or in addition to these demand signals, third information processing events from the central switching equipment ZW to the measuring device B, i.e. these third information processing events are the basis for determining the load capacity. Such third information processing events may be, for example, internal audit routine program events.

The measuring device B thus calculates the demand signals received during the respective measurement time period and forms a load value measurement value from this. In addition, there is a standard value for load capacity. This has been determined from the computationally known output values during the first commissioning of the call transfer equipment of the type presented here and stored in the memory B2 of the measuring device B. It indicates the number of requirement signals that are transmitted - on a long-term basis - from central switching equipment of the type shown here under normal load. This constant value is additionally stored in the memory B1 when the relay is commissioned. The measuring device B now subtracts from the value stored in the memory B1 the number of demand signals received in each case for the measurement time period and in turn divides the resulting difference by the value stored in the memory B1. If a value of 1 is then obtained, then it is found that the central switching equipment ZW has continuously processed information processing events during the respective measurement time period. This is an indication that the main switching equipment 19 79920 was overloaded during that measurement period. The main switching equipment should normally only be 95% loaded. If said division event results in a value of 0, then it can be seen that the central switching equipment was 95% loaded as desired. However, if a lower value is obtained, for example -0.02, then it can be seen that the central switching equipment was only 93% loaded during that measurement period, i.e. too little.

Depending on the result of the distribution event per measuring period, measuring device B either does not transmit any signal to the load counter W, or it transmits when the main switching equipment is too low or too loaded (i.e. 95% nominal load is higher or lower than the actual load at that time). switching equipment load) via the increase signal path + b or, respectively, via the decrease signal path -b.

As mentioned above, the increase signal or, i.e., the reduction signal, is input from the measuring device B to the load calculator W. A reference value to be explained later is stored therein. This is increased or decreased when the increase signal or the decrease signal is received. These signals, i.e. in each case one or the other, is fed from the measuring device B to the load counter W. All information processing tasks which pass via the line za and are passed on via the line ze pass through this load counter. These information processing tasks are partly initial tasks and partly post-tasks. The connection IDs provided by the calling subscriber to establish the connection consist primarily of an initial connection ID (paging signal) and post-switching field IDs, which arrive at intervals required by the nature and manner of the calling subscriber. In each case, in connection with the connection establishment, one initial connection field identifier and several post-connection identifiers always belong together. In addition to this, the post-connection codes can be 20 7 9 9 2 0 in connection with the connection establishment, the end-of-selection code, the response code, the end-code and the like, i.e. the line codes, as well as the payment calculation pulses. These switching IDs are received by the distributed switching equipment (partly the central switching equipment) GP; they consist of information processing tasks which must be separated according to the separation of the switching symbols, as well as into the initial switching symbols and the post-switching symbols, namely the initial switching and post-switching functions.

Successive information processing tasks now arrive via management za. These consist to a greater extent of the initial tasks and to a greater extent of the subsequent tasks. These information processing tasks arrive in disarray with respect to the various connection establishment events, i. so confused. They pass through the load counter W and are stored in the FiFo memory. This FiFo memory has a relatively large number of memory units in a known manner, in which case one memory unit in each case operates to store one information processing task. In the same order as the information processing tasks are handed over via line ze to the FiFo memory, they are further handed over from the central switching equipment ZW when called via line h from the FiFo memory to this central switching equipment.

The load counter W now acts as a second measuring device for determining the deviation of the working load of the acute information processing work of the central switching equipment from its performance. Of these, the load counter W, through which all the information processing tasks pass, calculates only the initial tasks.

There are two possibilities for the latter calculation. Another possibility is that, for example, every second the time stamp transferred to the load counter via line t2 starts or vastly stops the calculation event in the load counter by 21,79920 W, and that the calculation result obtained between the two time entries is compared with a reference value. Another possibility is that the counter is switched on and off at the same time intervals with the initial tasks, regardless of the initial tasks, and that the reference value is the number of feedback events performed in each case between the two in the same way as above.

Thus, a common FiFo memory is provided for the initial tasks (initial connection IDs) and the subsequent tasks (post-connection IDs); the load counter W selects the initial tasks from the information processing tasks (switching codes) entered in the FiFo memory to limit the calculation event only to these. As shown, the load counter W performs the calculation in successive calculation time intervals limited by said time notations; it is returned to its initial position at the end of each calculation period with a time stamp and produces a calculation result comparable to the reference value at the end of each calculation period. This benchmark constitutes a threshold. Another possibility is that the load counter W is switched forward with the initial tasks and reset continuously gradually. In this case, the load calculator thus always keeps the calculation result ready, which is subjected to a continuous, but at least once per, forward counting event and / or return event, comparison event with a threshold value. - In both cases of the special structure of the load counter W, the input of initial tasks (initial connection codes) to the FiFo memory is prevented on the basis of exceeding the threshold value, as a result of which all subsequent tasks (initial connection codes) belonging to the initial task (initial connection code) are blocked. To achieve this, a restriction device L is provided. If in the load calculator W the count value exceeds a threshold value, i.e. a reference value, then the load counter W transmits a reduction signal via line -v to the restriction device L. If the threshold value is again overridden by the count value, 2279920 counter W transmits the increase signal + v to the restriction device L. The delivery of the reduction signal or the increase signal from the load counter W to the restriction device L takes place at regular # intervals, for example 1 second.

The restriction device L now causes the connection groups LTG1 -LTGn # to reject a certain percentage of connection attempts here. This reduces the number of incoming information processing tasks. To this end, a memory Zi is arranged in each of the distributed switching equipment GP, in which the percentage of call signals to be discarded when congestion of information processing traffic is detected in order to counteract such congestion is stored. For example, percentages of 0%, 25%, 33.3%, 50% and 75% can be stored in this memory. These percentages indicate the proportion of incoming paging signals that must be rejected. Rejection means that when a paging signal arrives at the subscriber in question, a busy signal is sent and that it is prevented that the dialing codes provided from it are not received or received in the form of post-connection codes. It is equally possible to store, instead of the percentage of call signals to be rejected, also the percentage of call signals to be received, i.e. the "pass-through rate". - The rejection of the call signals by means of the rejection device can take place according to the calculation principle, in which case the calling subscriber is temporarily successful or unsuccessful, or it can also be distinguished between calls or subscribers with different priority classes.

Information on the respective percentage of initial tasks to be rejected or passed in each case is transferred from the restriction device L to the distributed switching devices GP and stored therein. Such transmission of this information shall take place either without interruption or only when an increase or decrease in this percentage has to be notified. The memory switching devices Zi arranged in the distributed switching devices by means of the percentage storage 23 79920 reduce or increase the stored percentage each time the information indicating the increase or decrease is received.

It is now arranged that the data indicating the percentage increase or the corresponding decrease are transmitted together with the data transmitted from the central switching device to the distributed switching devices, in particular the control and monitoring information, from the central switching device to the distributed switching devices and to the switching devices. For this purpose, the restriction device L is connected to the central switching device via a line d. It provides information on the current percentage through this road. The central switching equipment ZW adds this information, which in each case has a relatively small information content, to the control and control information for generating transmission events, which are thus continuously transferred from the central switching equipment to the distributed switching equipment. Thus, it is not necessary to initiate your own message transactions in order to transmit the information on the percentage in question, but this information is added to the connection control and control information. Thus, due to the different intensity of the data exchange with the different interface groups, the data indicating the respective percentage is transferred very quickly to the distributed switching equipment, which is in a very busy data exchange with the central switching equipment. Since said percentage is stored in the distributed switching equipment, during the entire duration of the load rejection phase with an equally strong load rejection, it is not necessary to transfer information about the load of the information processing traffic from this central control equipment to the distributed switching equipment. If congestion of information processing traffic occurs from the interface group, then the restriction information in question is first reported here.

24 79920

Thus, the information processing tasks (connection IDs) arrive in their respective chronological order with respect to different sets (the set always corresponds to the connection establishment event) in succession; they indicate their respective affiliation in each case by the origin address added to each switching code. This origin address indicates in each case the order in question calling the subscriber to the subscriber interface connection or connection group or to the line termination connection or the like, i.e. to one of said connection-specific switching devices.

It is provided that after the control of the congestion of the information processing requirements decreases and the amount of control percentage delivered from the restriction device L to the switching device ZW (distributed to the distributed switching devices GP for the above division) after, during which time, for the load-determining value determining the load, the higher values are always determined on the basis of the recovery of the actual load, the adjustment of the reference value stored in the memory S of the calculator W is temporarily terminated. This adjustment was made on the basis of the load measurement result obtained from the measuring device B and the division result formed therefrom. - A measuring device E is provided for measuring the above-mentioned minimum quenching time. This is connected via line e to the output of the restriction device L. Through this line e, the pointing device E receives in each case the percentage of control determined by the restriction device L. Furthermore, the indicating device E is connected via the line t1 to the timing device Zg. When no congestion of the information processing traffic has occurred for a long period of time without interruption, for example within 12 seconds, i.e. the rejection percentage has been at zero without interruption for such a long time, the load capacity measurement is deactivated. To this end, the indicating device E thus monitors from time to time from the time of the return of the control percentage to the value zero in the restriction device L from this time onwards that this percentage remains at zero.

If this condition is met, then the indicating device E transmits the corresponding signal to the calculator W via the signal path w.

25 79920

Similarly, at the beginning of the minimum quench time, i.e., whenever the rejection percentage delivered from the restriction device L reaches zero, the indicating device E passes the signal through the signal path w to the counter W, which marks the beginning of the potential minimum quench time. This signal causes the reference value currently stored in the memory S in the calculator W to be transferred to the additional memory s and stored therein. The reference value that may have already been stored in it will then be deleted at the same time. If, at the end of the minimum quenching time, the counter W also receives a signal already arranged above for this purpose via the signal path w, then this causes the reference value stored in its additional memory s to be transferred to its memory S, namely by deleting the reference value obtained in the memory S during the minimum quenching time. . Thus, during this minimum quarantine period, the rejection rate is by definition zero without interruption. Rejections are therefore not necessary during this period. In general, this is achieved by the recovery of the corresponding load. This has now ensured during the minimum quarantine period that the reference value of the counter W, which is stored in its memory S, is continuously post-adjusted with a positive trend, i.e. it has gradually obtained a relatively high value. In practice, this means that as the flow of information processing tasks or initialization IDs increases again, a relatively large number of these would first be accepted, i.e. no rejections would occur first, precisely until finally the reference value control from the measuring device B reduced this again. that, correspondingly, by again decreasing the reference values, the rejection percentage again obtains de-finitive values, i.e. rises from zero. However, this would result in a temporary overload of the central control unit 26,79920. To avoid this, it is arranged, as shown, that the reference value resulting from the decrease of the rejection percentage to zero in the memory S of the counter W is transferred to its additional memory s so that it can be transferred back to the memory S again after the minimum quarantine time has elapsed.

After deactivating the reference control, the calculator W operates at a start reference value that is higher than the reference value stored in the memory S. A new start in the adjustment of the reference value and the rejection rate data can be achieved for two reasons. If the load counter W, which calculates the initial switching symbols, again reaches a higher value, namely a count value equal to or higher than the start value similarly set in the memory S, which is higher than the maximum stored reference value in general, or if both the initial tasks and the the occupancy rate of the FiFo memory storing the tasks reaches or exceeds a certain limit value, then the interpretation of the determination result delivered from the measuring device B is restarted. The time period between start-up and deactivation forms the load adjustment phase; it has the current information processing workload in the load control area comprising the full load and overload range.

It is thus arranged that the start value stored in the memory S is set so high that start-up takes place only when, only in the case of a short-term overload, this is quite substantially greater than the load capacity. In this case, the start-up takes place only for a short time. If, on the other hand, the re-entering overload only slightly exceeds the load capacity limit, it will take longer for the start-up to start. For this purpose, a fixed value stored in the measuring device F of the FiFo memory is provided. With a relatively small but longer lasting 2? 79920 overloaded this memory was filled '' constantly, a boot occurs using this. The delay time is then almost inversely proportional to the amount of overload and greater than the response time of the counter W to a sudden large overload. The start value stored in the memory S and the fixed value stored in the measuring device F of the FiFo memory complement each other in their effect, namely in that the value stored in the memory S ensures that the high overload is detected as quickly as possible and the value stored in the measuring device F also ensures a small overload.

Measuring device B thus produces a load test result for the central control equipment, which, for the reasons explained above, always collects a slightly variable load capacity and integrates its values for time periods which are in each case large enough to guarantee this test result with sufficient result accuracy. This determination result always appears in the load measurement value stored again in the memory B1, in the increase and decrease signals further transmitted to the calculator W and thus in the reference value continuously adjusted in the calculator W.

As further shown, at the end of the load adjustment step, i.e. when the calculation results periodically obtained from the calculator W continuously decrease, i.e. from the calculation result to the calculation result, the value currently at or below the reference value stored in the memory S is stored, thus corresponding to the reference value and the most recent calculation. value, at the beginning of the minimum quenching time in the additional memory s and at the end of it is transferred from this memory to S. This value remains stored in the calculator W after the end of the load control phase until the beginning of the next load control phase. This value acts as a normal value, i.e. as an initial value for the calculator W at the beginning of the next load adjustment phase. Thus, the calculator W continuously starts its periodically repeated calculation operations on the basis of a reference value of a reference value of 28,79920. which is substantially more topical compared to the already calculated, calculated load value. This reference value present in the calculator W at the beginning of the new load control phase is that which occurred when the overload limit for non-overload was exceeded in the previous load control phase. This reference value stored in the load control phase from the load control step thus corresponds substantially more to the current operating situation than the calculated constant load capacity value. The reference value control, which is restarted after start-up, therefore starts from a value that is substantially closer to the current operating situation, the various possibilities of which were presented more closely at the beginning, than the standard value of the load capacity. This value stored from one load control step to another serves as a normal value. - In addition to this, it is also possible to determine the average of the reference values stored in the memory as shown during several load control steps and to use the corresponding average as a normal value. In this case, the memory s is furthermore an average converter which, at the end of the load adjustment step, averages the value already stored from the previous load adjustment steps and then the new reference value taken into the memory S.

Claims (2)

29 79920 1. Switching device for telecommunication equipment, in particular call transfer equipment, in which central and / or sub-central information processing switching equipment (ZW, GP1-GPn) with limited performance in terms of information processing capacity is arranged, and measuring devices (B) for determining the for continuously varying load time periods which are in each case large enough to guarantee the determination result with sufficient result accuracy and in which the information processing traffic load of each switching device (ZW, GP1-GPn) is determined to determine and prevent a value obtained by comparing the calculation result obtained repeatedly from time to time with that received and to it as the continuous calculation results increase over the analysis result, such overloads can be counteracted without delay by rejecting the information processing tasks, due to the recovery of the load, the production and / or sub-delivery and / or interpretation of the analytical result shall be cut off - the latter continuously comparing the calculation results , until due to the accumulation of information processing tasks or continuous l when the assay results increase substantially above the reference value above the normal value, the production and / or sub-delivery and / or interpretation of the assay result is reactivated, characterized in that when the periodic calculation results decrease when the assay result is reached or undershot, during which time the periodically calculated calculation results remain below the determination result, this stored value serves as a normal value, and that until the new determination result is obtained, the periodic calculation results operate on the basis of this normal value to determine continuous instantaneous values of the information traffic load. Switching device according to Claim 1, characterized in that the average value is the average of several stored values.