RU2195080C2 - Method of hybrid switching of digital communication channels - Google Patents

Abstract

FIELD: computer science. SUBSTANCE: invention refers to information-computer networks and can find use in design of digital networks of integrated service. Technical result lies in prevention of network interlocking, in raised efficiency of utilization of communication channels with fixed level of service failures and preservation of time scale while long messages are transmitted. Dialog between communication center and calling subscriber is organized at phase of connection establishment. As result of dialog length L of message is ascertained and compared with threshold value L_t.v = k•L_cr+b,. If length of message exceeds threshold value, i.e.L>L_t.v, then physical connection and message transmission are carried out under condition of channel switching and if L<L_t.v then message is memorized, divided into packages and message transmission is carried out under condition of package switching with their transmission through virtual connection. EFFECT: raised efficiency of utilization of communication channels and prevention of network interlocking. 4 dwg

Description

 The invention relates to the field of computer networks and can be used in the design of digital integrated service networks.

 A known method of hybrid switching, allowing you to switch channels in synchronous circuit mode and in synchronous and asynchronous package mode (see EPO patent (EP) A1. 0403911, class H 04 L 12/64, 1991). This method is based on the use of time sharing between subscribers, with the access time interval assigned to different subscribers in the packet switching mode, which limits the required RAM memory of the switching centers. However, this method does not ensure the efficient use of communication channels, since with such control inevitably there are pauses between the moment the message was sent and the moment the sender issued the next message.

 A known method of adaptive switching, providing the organization on the network of connections in the mode of switching channels with simultaneous transmission of messages in the mode of packet switching. In this case, the dynamic redistribution of the throughput of the network paths between the message flows transmitted in the switching modes of channels and packets is carried out (see S. Samoilenko, Adaptive Switching Method. - "Telecommunication", 6, 1981). The disadvantage of this switching method is that the messages are divided into blocks and recorded in a common memory field, regardless of the switching method. The choice of one or another switching method is carried out in a failure mode in the absence of free memory cells. In addition, different blocks of the same message can be transmitted using different switching methods, which leads to a violation of the time scale of the entire message.

 The closest in technical essence method chosen as a prototype is a hybrid switching method based on the integration of switching equipment necessary for the implementation of each switching method: channels and packets (Jenny Christian J., Kummerle Karl, Burge Helmut. Network node with integrated circuit / Packet switching capabilities. - "Communes. Networks Eur. Comput. Conf., London, 1975." Oxbridge, 1975, 207-228).

 A feature of this method is the use of a modular structure in its implementation, where the redistribution of network resources is carried out by the central processor. However, when the channel switching mode is established, the information flow is controlled without the participation of the central processor. This means that the transition from one switching mode to another occurs without taking into account the current state of the network, which can lead to either its blocking, unjustified denial of service, or inefficient use of network resources when transmitting long messages. In addition, the transmission of long messages in packet switching mode will inevitably lead to disruption of the real time scale.

 The purpose of the invention is the prevention of network blockages, increasing the efficiency of using communication channels at a given level of denial of service and maintaining the time scale when transmitting long messages.

The goal is achieved by the fact that as a result of the dialogue, the message length (L) is found out and this length is compared with the threshold value L _p = k • L _cr + b, where L _cr is the critical message length pre-set for each particular network; k is the coefficient of proportionality, taking into account the degree of employment of the amount of memory; b - a value that reflects the dynamics of the graph. If the message length exceeds the threshold value L> L _p , a real connection is established and the message is transmitted in the channel switching mode, and if L <L _p , the message is stored, divided into packets and a virtual connection is established.

 Thus, notification of the switching center about the length of the message to be transmitted in the connection phase allows preventing network collisions associated with the memory overload of switching nodes, increasing the efficiency of communication channels by transmitting very long messages in real time and reducing the total number of messages receiving connection failure due to lack of free memory buffers. Moreover, the storage of long messages is assigned to the caller, and the storage time should not exceed a certain value of τ in accordance with recommendation Q.543 of the JTU-T sector (former CCITT).

 The implementation of the proposed method is illustrated using figures 1-4.

 The structural diagram of a device that implements this method is presented in figure 1. It is supposed to use the modular structure of the switching center, including control module 1, identification module 2, intermediate memory module 3, access and interface modules for communication lines with the switching center 4, 5, the information inputs and outputs of which are interconnected via trunk 6. Incoming line 7 through access module 4 is connected to another input of the identification unit 2, the other output of which is connected to the second input of the intermediate memory module 3, the second output of which is connected via the access and interface module 5 with outgoing line 8, while the third output of the identification module 2 through the interface module 5 is connected to the outgoing line 9.

 An example implementation of the identification module 2 is presented in figure 2. It contains the first digital-to-analog converter 10, the input of which is the other input of the identification module 2, and the output is connected to the first input of the comparison circuit 11, the output of which is connected to the S-input of the control trigger 12, the direct and inverse outputs of which are connected to the first inputs of the first 13 and second 14 AND elements (logical multiplication) connected by second inputs to the input of the first digital-to-analog converter 10. The outputs of the first 13 and second 14 AND elements are connected to the other and third outputs of identification module 2, the second The second input of the comparison circuit 11 is connected to the output of the summing amplifier 15, one input of which is connected directly to the output of the second digital-to-analog converter 16, and the other input through a differentiating element 17, while the input of the digital-to-analog converter 16 is connected to the highway 6, and the R-input of trigger 12 connected to the "Installation" terminal 18.

The proposed method consists of a sequence of interrelated operations. The switching center carries out switching of channels and packets, while it is assumed that the switching equipment necessary for the implementation of each switching method is integrated. The structure of the switching center includes four types of modules that perform certain functions:
control module 1 is used to control and control the connection of outgoing and incoming lines, performs the functions of controlling the switching center and routing messages, as well as the functions of computing, logic, and others related to accounting and control of the current state of the network;
identification module 2 distributes messages along their length taking into account the current state of the network and, in essence, is a decision block for the implementation of a particular switching method;
the intermediate memory module 3 allocates buffer memory for storing packets and processes them;
access and pairing modules of communication lines 4, 5 pair the incoming and outgoing lines with the switching center.

 All modules can exchange information through line 6. This structure of the switching center allows you to manage its work with a simple processor.

 The switching center operates in two modes.

In the phase of establishing a connection, a dialogue occurs between the caller and the switching center, during which the switching center finds out the message length and the address of the called subscriber. At the end of this phase, the switching method is selected by analyzing the message length and the buffer memory occupancy. If the message length does not exceed a critical value, i.e. L <L _cr , and all the buffers of the channel set to transmit the message to the addressee are free, the message is transmitted to the output of the first element And 13, the first input of which is in the initial state at high potential, removed from the inverse output of the control trigger 12, and then comes in the intermediate memory module 3, where it is divided into packets and through the access and interface module 5 it enters the outgoing bus 8 for transmission over a virtual connection. The control module 1 provides the intermediate memory module 3 with the information necessary to form the packet headers, places the packets in the allocated part of the memory buffer, sends the buffer address to the address register of the module 5 that processes the outgoing line 8. The necessary amount is also allocated in the switching centers of the transit nodes and the destination node buffer memory for each virtual connection for sending or assembling messages, respectively.

If the message length exceeds a threshold value, i.e. L> L _cr , then regardless of the state of the buffer memory and the amount of traffic, a decision is made to establish a physical connection and send a message in channel switching mode. In this case, the message can be transmitted directly to the access and interface module 5 and without the participation of the control module 1 by supplying the corresponding potential level to the second input of the comparison circuit 15 of the identification module 2. In this case, the functions of the control module 1 are reduced only to the analysis of the address part of the message and establishing a physical connection. If part of the buffer memory is occupied and (or) insufficient to accommodate the entire message, then the decision to use the channel switching method is made in the identification unit 2 in accordance with the expression L> k • L _cr + b by applying the corresponding potentials to the first and second inputs of the circuit comparisons 11.

If the message length information (L) and the threshold value (L _p ) are set in binary code, then in the example of the implementation of the identification unit 2 in figure 2, the digital-to-analog converters 10 and 16 directly convert the code to voltages, which are compared in the comparison circuit 11. If Since the values of L and L _n are set by some other code, then before the elements 10 and 16 it is necessary to put the decoders that convert this code to binary.

где m - общий объем памяти; m_з(t) - текущее значение занятого объема памяти;

- производная по времени от занятого объема памяти; β - коэффициент пропорциональности.The code corresponding to the length of the transmitted message arrives at the input of the first digital-to-analog converter 10 at the end of the dialogue, while the input of the second digital-to-analog converter 16 receives the current values of L _p calculated by the processor during the entire operation of the network taking into account the state of the memory buffers. In an example implementation of the circuit of the identification unit 2 shown in FIG. 2, the threshold value L _p can be calculated by the formula

where m is the total amount of memory; m _z (t) is the current value of the occupied memory size;

- time derivative of the occupied memory; β is the coefficient of proportionality.

If the true message length, converted to voltage by the digital-to-analog converter 10, exceeds the threshold U _p = α • L _p (α is the transfer coefficient of the digital-to-analog converter), then a high potential will appear at the output of the comparison circuit 11, at which the trigger 12 will switch to the second stable state and at its direct exit there will be a high potential. The first element And 13 will close, and the second element And 14 will open. Identification module 2 is ready to send a message through module 5 to outgoing line 9. After establishing a through channel to the destination, the switching center initiates the transmission of the message. Otherwise, if the connection is not established, the subscriber is denied.

Taking into account the dynamics of traffic changes in the circuit of FIG. 2, it is implemented on a summing amplifier 15, on which U _p is supplied to one input, and its derivative ∂U _p / ∂t, the sign of which increases or decreases the threshold U _p , depending on whether , the number of occupied buffers at the moment increases or decreases, thereby changing the ratio between both switching modes (see Fig. 3).

 After transmitting the message, the control trigger 12 is returned to its original state, for example, by applying a "Reset" signal to its R-input from the control module 1 via terminal 18.

Thus, according to FIG. 3, for long messages and increasing traffic, the channel switching method will prevail and, conversely, if short messages are received with very pulsating traffic, it is advisable to transmit messages using the packet switching method. The use of a particular mode is determined by the choice of L _cr . If the number of occupied buffers (m _s ) and the value of b can be controlled within a given switching node or throughout the network depending on the accepted routing method, then the critical message length (L _cr ) is a design parameter and is set at the design stage of a specific network.

т. е. не должно превышать объем буферной памяти, так как, в противном случае, сообщение получает отказ или должно разбиваться на блоки.The total number of packets of a fixed length (L _pack ) that can be formed from a message of length L is determined by the ratio

that is, it must not exceed the amount of buffer memory, because, otherwise, the message is rejected or must be divided into blocks.

где ρ = λ/μ - коэффициент загрузки сети; λ - интенсивность поступления сообщений; μ - интенсивность обслуживания; Р_отк ^треб - требуемое значение вероятности отказа в обслуживании.Thus, the critical message length can be determined from relation (1):
L _cr = L _Pak • m. (2)
For a single-channel network as a queuing system with expectation, the probability of a denial of service message being received is [1]

where ρ = λ / μ is the network load factor; λ is the message arrival rate; μ is the intensity of service; R _open ^req - the required value of the probability of denial of service.

Выражения (2) и (5) позволяют определить критическую длину сообщения с учетом загрузки сети и требуемого значения вероятности отказа:

Условие выбора критической длины сообщения (5) не учитывает структуру сети и такие ее показатели, как время задержки сообщения в сети, общий трафик сети, стоимость сети и другие показатели, являющиеся исходными данными при проектировании сети.Solving equation (3) with respect to m for the limiting value P _open , we obtain

Expressions (2) and (5) make it possible to determine the critical length of a message taking into account the network load and the required value of the probability of failure:

The condition for choosing the critical message length (5) does not take into account the network structure and such indicators as the delay time of the message in the network, total network traffic, network cost, and other indicators that are the initial data in the design of the network.

где γ - общий трафик сети; С_зад - стоимость передачи единицы количества информации; k - коэффициент пропорциональности;

- оптимальное значение коэффициента загрузки; n - общее число узлов сети.Studies conducted by the inventors [2], allow a more informed choice of L _cr taking into account the above indicators. The minimum average message delay time in the network can be calculated from the relation [2]

where γ is the total network traffic; With _ass - the cost of transmitting a unit amount of information; k is the coefficient of proportionality;

- the optimal value of the load factor; n is the total number of network nodes.

T dependency curves $\genfrac{}{}{0ex}{}{\text{min}}{\text{cp}}$ = f (ρ _opt ) and P _open = φ (ρ _opt ), constructed in accordance with expressions (6) and (3), are shown in the combined graph of Fig. 4. The analysis of the curves shows that the required probability of failure P _open ^demand = 0.01 is achieved with the number of buffers m = 20 and the network load factor ρ = 0.9 (point B). Moreover, for a rather complex network containing n = 40 switching nodes and busy traffic γ = 36 packets / s, the average message delay time T _cp ^min = 7 s. (point A) and does not depend on the cost of the network. In accordance with expression (5), the critical message length is L _cr = 20480 bits with a packet length of L _pack = 1024 bits.

 Thus, the analysis confirms the advisability of transmitting long messages by the channel switching method, since this not only ensures the preservation of the time scale, but also serves to achieve the stated objectives of the invention.

LITERATURE
1. Ventzel E.S. Operations research. - M .: Nauka, 1989 .-- 552 p.

 2. Lipets G.I., Fomin. L.A., Zdanevich S.N., Pavlenko N.A., Budko P.A. Synthesis of a data network with limited network resources. // Sat scientific tr : Information processing systems. - Kharkov: NASU, PANI, HVU, 2000. - Issue 1 (7). - S. 65-71.

Claims (1)

A method of hybrid switching of digital communication channels, in which, in the phase of establishing a connection, a dialogue is established between the switching center and the calling subscriber, as a result of which information about the address of the called subscriber is received and processed and a connection is established in the channel or packet switching mode, characterized in that, for the purpose preventing network blockages, increasing the efficiency of using communication channels at a fixed level of denial of service, and maintaining the time scale when transmitting long messages messages, as a result of the dialogue, find out the message length L, which is compared with the threshold value L _p = k * L _cr + b, where L _cr is the critical message length pre-set for each particular network; k is the coefficient of proportionality, taking into account the degree of employment of the amount of memory; b is a value that reflects the dynamics of traffic changes, and if the message length exceeds a threshold value, i.e., L> L _p , then a physical connection is established and the message is transmitted in the channel switching mode, and when L <L _{p the} message is stored, divided into packets and message transmission are carried out in packet switching mode with their transmission over a virtual connection.

Images