SU1431043A1 - Random process generator - Google Patents

Abstract

The invention relates to a pulse technique. The circuit of the invention is the extension of functionality due to the formation of random processes that are explicitly specified. The formation of a random process is provided by pre-recording in the blocks 2.1-2.N of the memory of the matrix of probabilities of the transition, in blocks 3.1-Z.M.

Description

W (L

four

Ca

14:

WITH

143

(M + 1) memory - binary codes for scale factors. By impulse by fault 7 Starting into register 8, a code is written that provides for the control inputs of the controlled random code generator (HSC) 1 elements of the transition probability matrix from blocks 2c, 1-2.N of the memory, and for the control inputs of the controlled HSC 4 - corresponding codes from 3.1-ZM memory blocks. HSC 1 is interrogated by the same impulse delayed by delay element 6. Additionally, the same impulse delayed by delay element 9 interrogates HSC 4, from whose output it goes to the inputs

: one

The invention relates to the pulse: technology and can be used IB information-measuring tech- nique I The aim of the invention is to increase the functionality of the generator by forming random iHbtx processes that are explicitly defined:

G FIG. 1 shows the structure-; on the generator circuit of the random pro-; process; in fig. 2 shows an example of a controlled random code generator.

The random process generator (Fig. 1) contains the first controlled generator 1 of a random code, the corresponding groups of control inputs of which are connected to the outputs of the corresponding blocks 2.1-2.N of the memory of the first group 2, the address inputs of blocks 2.1-2 The .N memory of which is connected to the second groups of address inputs of the corresponding blocks 3.1-3. (M + 1) of the memory of the second group 3, the outputs of the 3.1-ZM modules of the memory of which are connected to the corresponding groups of control inputs of the second controlled oscillator 4 of the random code whose outputs are connected according to The current information inputs of the transducers 5 code is the time interval, the output of which is connected to the first delay element 6, 1st 7, the Start with the register synchronization input 8,

43

Converter 5 code - time interval. The latter forms at its output signals that organize the further operation of the device. Codes from the output of GSK 1 are fed to another group of address inputs of blocks 3.1-3. (M + 1) memory, the contents of the last of which controls the operation of converter 5 of the code-time interval, making the transition from the discrete distribution of the waiting time (stay) of the process in this state to non-continuous to form a discrete Markov process or Markov chain. An example of the implementation of GSK 4. 1 GoP f-ly, 2 ill.

A delay element 9, the output of which is connected to the polling input of the second controlled oscillator 4 of a random code and the synchronization input of the converter 5, is the time interval, the control inputs of which are connected to the corresponding outputs of block 3. (M + 1) memory of the second group 3, second groups of address inputs of blocks 3.1-3. () the memory of which is connected to the corresponding outputs of the register 8, the corresponding information inputs of which are connected to the first groups of the address inputs of blocks 3.1-3. (M + 1) of the memory of the second group 3 and to the outputs of the first controlled generator 1 of a random code whose polling input is connected to the output of the first delay element 7 and to the input of the second delay element 9

The controlled generator 1 (4) of the random code contains the generator 10 of a random stream of pulses, the output of which is connected to the direct input of the BANNER element 11, the output of which is connected to the input of the switch 12, the outputs of which are connected to the inputs of the corresponding converters 13.1-3.N the code is the intensity of group 13, the transducer outputs are 13.1-13.N the code is the intensity of which is connected to the corresponding inputs of the first register 14, the outputs of which are connected to the corresponding inputs of the encoder 15, the outputs of which are connected to the corresponding The signals of the second register 16, the outputs of which are the outputs of the controlled generator 1 (4) of a random code, the polling input of which is connected to the synchronization input of the second register 16 and the inverse input of the BAN 11 element, the Q groups of the control inputs of the controlled generator 1 (A) random code is connected to the control inputs of the corresponding converters 13.1-13.N code - intensity.

Memory block 3.1 contains a group of memory blocks, the address inputs of which are combined and are the second group of address inputs of memory 3.1, and a group of multiplexers, whose codes are the outputs of memory 3.1, and the address inputs of multiplexers are combined The first group of address inputs of the memory 3.1. At the same time, the k-th (k 1, m) spacing of the j-ro (j l,) memory block of the group is connected to the j-M input of the k-ro multiplexer of the group.

The random process generator works as follows.

In order to form a random (semi-Markov-satisfying Markov property with no aftereffect only by half), the process, which is explicitly stated, prepares the generator. To do this, in blocks 2.1-2.N of the memory of group 2, the entry is a matrix and the |, I, J, k 1, N probabilities are

which are t-bit binary codes X-, lv 45

recorded with probabilities and: ratio, where j is the address of the line k-ro of block 2.k of memory 2. Similarly, in M (first in order) blocks 3.1-Z.M, the memories of group 3 record p-bit codes, 1 1, М, which are the corresponding mapping of the elements of the matrix of discrete laws ffi (t) j of the waiting time (pre) of the semi-Markov process in state j, provided that the next state that the generator goes into is k. Accordingly, in the (M + 1) -and block 3. (M + 1) of the memory of group 3 write the binary codes of the scale factors C

ary

: converter of random codes X е 0, М-1, into random ones.

5 25 30

40

50

intervals with average M (T1 v g. 1

  tS.,

where is the intensity of the stabilized source of a random stream of pulses that is part of the converter 5 code-time interval.

The initial state of the simulated semi-Markov process is set by setting the appropriate binary code X V in register 16 of the first controlled gene Q

0

five

five

Rattor 1 random code.

The impulse on the niisHS 7 Start is fed to the input of the synchronization input of register 8 and writes the code X Vj into it. From the output of register 8, this code arrives at the address inputs of the 2.1-2.N memory blocks of group 2 and the second address inputs of the 3.1 blocks. - 3R (M + 1) of the memory of group 3. From the outputs of 2.1-2 memory blocks. In the first group 2, binary codes of the elements of the matrix row and I transition probabilities are fed to the corresponding control inputs (probabilities setpoints) of the first controlled generator 1 5 of the random code. After the time of the last operation, specified by the first element 6 of the delay, the pulse from the bus 7, the generator starts to interrogate the first controlled generator 1 of a random code.

The operation of the first (second) controlled generator 1 (4) of a random code is as follows. The pulses of the generator 10 of a random stream of pulses through the permanently open (in the absence of polling pulses) element BANGE 11 is fed to the input of the switch 12. At the outputs of the latter, N independent random streams of pulses with the same intensities are formed. With the help of converters 13.1-13.N code - the intensity of group 13 of the intensity of N random streams of pulses is attenuated in proportion to the binary codes present at the corresponding control inputs (set the probabilities of the codes) of the first (second) controlled generator 1 (4) of the random code.

In register 14 of inconsistent random events, a Markov process with N states is modeled using an unconditional transition intensity matrix. (i.e., such a matrix in which the elements of one column are the same). AT

0

0

Accordingly, the final probabilities of register states 14 are equal to the relative weight of the corresponding binary codes in the aggregate of all binary codes present at the control inputs of the controlled oscillator 1 (4) of the random code. Using the encoder 15, the spatially distributed event, which is that at the current time, the k-th bit of register 14 is in the one state, is converted into positional binary code k, which is present at the input of register 16.

A polling impulse of the controlled generator 1 of a random code affects the synchronization input of register 16 and fixes in it the random binary code k of the next state to which the device enters. The code of the next state of the device is sent to the first groups of address inputs of blocks 3.1-3. (M + 1) of memory of group 3. From the outputs of blocks 3.1-Z. M of memory of group 3 to the corresponding group of control inputs (probability setting) The second controlled generator 4 of the random code receives the binary codes of the discrete law (t) of the distribution of the residence time of the process in state j, provided that the next state is k. The control input (scale) converter 5 code - time interval enters the binary code of the coefficient

After the time specified by the second delay element 9, in register 16 of the second controlled oscillator 4, the binary code X is fixed, using a code 5 converter — the time interval is transformed into a pulse of the device of the semi-Markov process kept on waiting for T on the second cycle of its operation.

The operation of the 5 code-time converter consists in making the transition from the discrete distribution of the waiting time to the continuous one.

At the beginning of the second cycle of operation, the code k of the next state, developed by the second controlled generator 4, is transferred to register 8, at the output of which it becomes a code.

five

0 5

0

five

Q

five

0

five

exponential distribution

j k current state. In the future, the operation of the semi-Markov process generator is repeated.

The generator forms a discrete Markov process if the waiting times in each state are random and having.

The generator models the Markov chain if the waiting times in each connection are the same and equal to one.

Claims (2)

1. A random process generator containing the first controlled random code generator, the outputs of which are connected to the corresponding information inputs of the register, the first group of memory blocks, the outputs of which are connected to the corresponding groups of control inputs of the first controlled random code generator, characterized in that the purpose of expanding the functional possibilities due to the formation of random processes that are explicitly specified, the second controlled random code generator, whose outputs are co are connected with the corresponding information inputs of the converter; the code is the time interval whose output is connected to the input of the first delay element, the output of which is connected to the input of the second delay element, the second group of memory blocks, outputs: which, besides the outputs of the last memory block groups are connected to the corresponding groups of control inputs of the second controlled random code generator, the polling input of which is connected to the synchronization input 1 of the converter code — time interval and with the output of the second ele The delay, whose input is connected to the polling input of the first controlled random code generator, whose outputs are connected to the first groups of address inputs of memory blocks of the second group, the second groups of address inputs of memory blocks of which are connected to the address inputs of memory blocks of the first group and with corresponding the outputs of the register, the synchronization input of which is connected to the input of the first delay element, the outputs of the last memory block of the second group are connected to their corresponding control inputs of the converter - time constant interval. . I 2. The generator according to claim 1, distinguished by the fact that the direct generator of the cny iHalHoro code contains a generator of a random stream of pulses, the output of which is connected to the direct input of the BANGE element, the output Q of which is connected to the input of the switch, the outputs of which are connected to the inputs of the corresponding converters code - the intensity of the group, the outputs of the converters code - the intensity of which is connected to the corresponding inputs of the first register, the outputs of which are connected to the corresponding inputs of the encoder, the outputs of which are connected to the corresponding The secondary inputs of the second register, whose outputs are the outputs # 1 of the controlled random code generator, the polling input of which is connected to the synchronization input of the second register and the inverse input of the BAN element, the control input group of the controlled random code generator are connected to the control inputs of the corresponding wook (their converters code - intensity.