JPS58220537A - Circuit supervising and controlling method of data terminal station - Google Patents

Abstract

PURPOSE:To reduce a mean delay time of data retransmission in case of generation of concurrence, by operating the retransmission probability by a result of supervision of a receiving circuit and the number of transmission waiting data in a data terminal station, in case when it is known that one-delivery of a data due to generation of concurrence or deterioration of a circuit occurs at the point of time when the operation is shifted to the transmitting operation. CONSTITUTION:A transmission request signal TREQ1 is generated in a section 1, SD1 is transmitted first in a section 2, and when it is received normally by the central station, a response signal RD1 is sent back from the central station. Subsequently, when another transmission request signal TREQ2 is generated in a section 3, and a transmission modulating data SD2 is transmitted in a section 4, if it cannot be received normally by the central station due to deterioration of the circuit, a response signal RD2 is no sent back, therefore, a retransmission controlling circuit operates, and retransmission of every section of a section 5 and thereafter is decided like probability. Also, when other transmission request signal TREQ3 is generated, two data are connected to the transmission queue, the number TQ of a transmission waiting data becomes 2, and a value of a circuit evaluating value output RXM becomes small due to generation of an error of RD of the sections 3, 4 and 5, therefore, the retransmission probability of a section 6 and thereafter becomes large.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention performs data transmission between one central station and a large number of data terminal stations that transmit a small amount of data each time using a contention method between the data terminal stations. The present invention relates to improvements in data line control methods.

Conventionally, when controlling a data line using a contention method, if a valid response cannot be obtained from the central station, this is judged based on the collision of transmitted data between multiple data terminal stations, and when retransmission is performed at each station, the transmission until retransmission is By setting the waiting time randomly each time, the probability of continuation of contention between multiple stations is reduced, or by setting a different time corresponding to the unique address number of the data terminal station, collision of retransmitted data is avoided. Methods such as these are used. In the former method, if there are many types of transmission waiting times that can be randomly selected, the probability of re-collision of transmitted data will decrease, but the average value of the data re-transmission delay will increase; In addition, in the latter method, when there are many data terminal stations, there is always a terminal station with a large retransmission delay, resulting in an imbalance in the timeliness of retransmitted data. Furthermore, if the line quality deteriorates in both cases, it is indistinguishable from the above-mentioned transmission data collision, and the same operation as when the contention occurs will simply be repeated, resulting in the transmission data probability (transmission f - It has the disadvantage that it is completely invulnerable to line disturbances, such as a decrease in the probability that the information will be correctly transmitted to the receiving side.

The present invention was made to eliminate these conventional drawbacks,
Data equivalent to conventional data □ “While using the evening line, we aim to reduce the average delay time of data retransmission when contention occurs, and also ensure that the data sent from the data terminal station to the central station is reliably transmitted even in the event of line disturbance. The feature is to prevent deterioration of the so-called remoteness, that is, the accuracy rate, and to reduce the loss of data generated at each terminal station at the central station. This has the advantage that it can be effectively applied to cases where the frequency of contention cannot be ignored, such as in wireless lines, or where the line quality is poor and varies significantly depending on location and time.

An overview of the present invention will be shown below, followed by a more detailed explanation based on Examples.

The outline of the present invention is as follows. The average value of the received input signal level and the synchronization pattern of received demodulated data from the central station within a certain amount of time in the past at multiple terminal stations that are connected to one central station by a common transmission line and are in a mutually competitive relationship. A circuit that constantly monitors the quality of the receiving line using a means to detect the number of data errors and the number of errors in the data, and a circuit that constantly monitors the quality of the receiving line using a means to detect the number of errors in the data. and a circuit that probabilistically determines retransmission or non-retransmission each time a predetermined transmission opportunity arrives when it is determined that there is no valid response, and the retransmission probability is calculated based on the monitoring results of the receiving line and the data. It operates according to the number of data waiting to be sent in the terminal station, and when the line deteriorates or the number of data waiting to be sent increases, the retransmission probability is increased to establish data reception, and when the line quality is good, it is in a competitive state. A station that identifies this and obtains a valid response from the central station for itself refrains from the next retransmission, and a station that identifies a response for another station executes the next retransmission, thereby reducing the duration of contention. That is the point.

Next, the above content will be explained in more detail.

FIG. 1 is a side view of the structure of a data transmission device of a data terminal station embodying the present invention. In the figure, 1 is a receiving section, 2 is a data transmission control section, and 8 is a transmitting section. Receiver l receives input signal RXI
If N is input, it is demodulated and demodulated data RD is output. On the other hand, the analog signal obtained by rectifying RXIN, R
XA is a reception level signal for using the level of the reception input signal RXI'N to monitor the reception line.

Next, the data transmission control unit 2 has the role of monitoring the data reception line, processing the transmitted and received data, verifying and queuing, and controlling the data transmission. It is used for frame synchronization, error checking, and received data line monitoring, and also outputs terminal received data RXD to the terminal within the data terminal station if there is validly received data addressed to the own station. Further, when a transmission request signal TREQ indicating that transmission data has been generated at the terminal is input from the terminal, the terminal transmission data TXD is taken into the transmission queue and data transmission control is performed. SD is transmission modulation data that is output by processing TXD during transmission, and is output to the transmitter together with the transmission activation signal TXG.

Next, the transmitter 3 has a data modulation function, activates the transmission output signal TXOUT by the transmission activation signal TXG, and transmits the data modulation signal to the other station (central station).

FIG. 2 is a time chart of the received data line monitoring operation of the data transmission control unit 2 in FIG. 1, and FIG. FIG. 2 is a diagram illustrating a configuration example of a line evaluation circuit (provided). The received data line monitoring function of the data transmission control section 2 shown in FIG. 1 will be explained below with reference to FIGS. 2 and 3. First, the horizontal axis of the time chart in FIG. 2 shows the passage of time, and the left end of the chart shows the signal names. It is assumed that the reception level signal RXA at the lowest stage in the figure changes between high and low levels as shown in the figure, and the dashed line EL in the figure represents a threshold value that causes a reception demodulation error. The large and small rectangular portions of the received demodulated data RD in the second stage schematically represent clusters of data, and F represents the synchronization pattern of the data frame.

The unmarked rectangular part indicates the information part of the 1 frame following the immediately preceding frame synchronization pattern F, and the shaded part indicates the section in which a demodulation error has occurred.

Each pulse of the FDPoFEP in the middle part of FIG.
The synchronous first-period detection pulse and the synchronous error pulse generated within the same period are respectively shown. In addition, RXM at the bottom schematically shows the line evaluation value output of the reception line monitoring result created in the transmission control unit 2, and n, n-1, n-2, etc. in the figure indicate evaluation values that change from moment to moment. It represents.

If the level of the reception level signal RXA is now below the threshold EL in the time period T□ and T2, a demodulation error will occur in the diagonal/lined area on RD, and the frame synchronization pattern F received during this period will not be detected. No synchronization detection pulse FDP is generated. At this time, by utilizing the periodicity of F, a frame synchronization error pulse FEP can be created as shown in the figure using a known method.

In the operating situation configured as described above, the value of the line evaluation value output RXM at the time point indicated by P in the figure, for example, is obtained by extracting the average level of the reception level signal RXA over a certain time period TM in the past from P. This can be expressed numerically, or it can be expressed as the difference between the number of FDP and FEP pulses within the TM time period. In the operation example shown in Figure 2, the initial RX
When the value of M is n, the influence of error occurrence on T0 is R
The decrease in the value of The increase from n-2 to n-1 is started. In this way, by quantifying the line quality in the vicinity of the current moment and directly correlating the size of the numerical value with the superiority or inferiority of the line quality,
The data transmission control unit 2 is provided with a reception line monitoring function.

FIG. 3 is a diagram showing an example of the configuration of a line evaluation circuit for obtaining the line evaluation value output RXM of FIG. 2. Of these, figure (A) shows the case where the received level signal RXA is used as an input, 4 is a low pass filter (LPF), and 5 is an analog-digital (LPF) that linearly or logarithmically converts the output of LPF 4 into a digital value. A/D) converter. When the cutoff frequency of the LPF 4 is set to approximately %TM, its output becomes approximately equal to the average value of the received level signal RXA within the time TM, and the operation shown in FIG. 2 can be realized. Note that LPF4 is omitted and RX
If you use an A/D converter that can sufficiently follow changes in A,
RXM can be similarly output by digital filtering of the digital output (which may be done by hardware or software).

Next, FIG. 3(B) shows a case where the frame synchronization detection pulse FDP and frame synchronization error pulse FEP shown in FIG. 2 are used, and 6 and 7 in the figure are delay circuits CD) and 8.
．． 9 is an OR gate, and 10 is an up/down counter. The delay circuit is for delaying the input signal by TM time. In Fig. 3(B), the pulse FDP is O
Through the R gate 8, the output (U,) causes the up/down counter 1° to be incremented by +1, but a pulse output also appears in the OR gate 9 after TM time through the delay circuit 6 and the OR gate 9, and the output (DOWN ), the counter 10 subtracts -1 and returns T. Cancel the addition before the time.

On the other hand, regarding the frame synchronization error pulse FEP, □F above
Contrary to the case of DP, first, a down output is generated through the OR gate 9 and -1 is subtracted from the up counter 10, but after the TM time, an up output is generated through the delay circuit 7 and the OR gate 8 to cancel this. Due to this operation, the output RXM of the amplifier down counter 10 becomes 1 of F'DP.
+1 is added by pulse, and by 1 pulse of FEP =
l is subtracted, but the effect is limited to the TM time, so the value is from the current time to the past TM time.
It is clear that the time is consistent with the pulse difference between FDP and FEP. Although pulses FDP and FEP were used as pass/fail pulses related to frame synchronization detection, since reception errors can be detected in error detection codes and error correction codes that are generally used for data transmission, FDP is used as a pulse for detecting frame synchronization. RXM
It is possible to reflect this in the value of

The above was an explanation of the receiving line monitoring function, but
Next, the data transmission control function of the data transmission control section 2 shown in FIG. 1 will be explained with reference to FIGS. 4 to 6. Figure 4 shows the configuration side of the retransmission control circuit in the data transmission control section in Figure 1, Figure 5 shows a time chart of an example of the retransmission control operation of the data terminal station when line quality deteriorates, and Figure 6 is a time chart of an example of retransmission control operation at the time of contention between two data terminal stations.

In Fig. 4, 11 indicates read-off η and memory (ROM).
Then, the stored data A is manually generated using the line evaluation value output RXM shown in FIG. Output. Reference numeral 12 denotes a random pattern generation circuit O8C that generates a random binary code pattern, which generally shifts the M-sequence generation shift register using a high-speed clock that is asynchronous with the transmission timing, and extracts pattern output B from specific multiple shift stages. This can be easily realized by known means such as. 13 is 2 of A and B
Comparator COMP compares the magnitude of base numbers, and its output (A
>B') becomes H (high) level only when A>B. 1
4 is a flip-flop FF which inputs the A>B signal from 18 and samples and holds it. The sampling time is given by a predetermined transmission start timing signal T. The output of FF14 is the retransmission command output RE
When TRY is at H level, the transmitted data is retransmitted. F'F14's manual SET is forced to output RE
Manual RESET is used when TRY is set to H level to execute retransmission, and manual RESET is used when output RETR'Y is forcibly set to L level to prohibit retransmission.

Next, the operation shown in FIG. 4 will be explained.

Suppose now that the data end station has not received an effective response from the central station to the data it has sent out. At this time, the output A>B of the comparator 1B is evaluated in FF'14 at the next transmission start timing signal T, and according to this, the retransmission command output RE
The level of TRY is determined. If the maximum value of the output B of the random pattern generator 12 is BMAx, B generates any value between 0 and BMAX with equal probability, so the minimum and maximum values of the data output of the ROM 11 are respectively AMIN and AMAX, 0<AMINGUA<AMAX<BMAXe...(1)
If it is satisfied, the output of COMP 18 A>B
The probability that the retransmission command signal RETRY becomes H level P (A>B)t The probability that the retransmission command signal RETRY becomes H level satisfies equation (2).

P(A>B)=A/(BMAX+1)...-(2) Therefore, the following equation (3) is obtained from (1) and (2).

<1...(3) Now, in ROMII, RXMlTQ is the address, and RX
Since the name address data specified by M and TQ is A, R
The following functional relationship holds between XM, TQ, and A.

A=f (RXM, TQ)...φ...(4) Therefore, set the data in ROM 11 so that equation (4) becomes a monotonically decreasing function for RXM and a monotonically increasing function for TQ. If you do so, the probability P that the retransmission command output will be at H level
Since (A>B) is proportional to A according to equation (2),
When the reception line evaluation value output RXM is large, that is, the line quality is good, or when the number TQ of data waiting to be transmitted in the transmission queue is small, the retransmission probability decreases, and conversely, when RXM is small and the line quality is low, the retransmission probability decreases. When it is deteriorating or there is a lot of TQ,
Probabilities can be manipulated to increase the retransmission probability.

Next, the operation of the circuit (-) with the above effects will be explained in detail with reference to FIGS. 5 and 6. First, FIG. The format of the diagram is the same as that in Figure 2.The numbers from ■ to [phase] between the pulses at the transmission start timing T are merely identification numbers of the transmission timing section for convenience. TREQI to TREQ8 below each pulse of the transmission request signal TREQ (Fig. 1) is the transmission modulation data SD.
The transmission request signal and the transmission modulation data are shown in a one-to-one correspondence with each of the data SDI to SD8, respectively, and are caused by generation of mutually different data. Furthermore, each data RDI, RD2, and RD8 of the received demodulated signal RD is 5D11SD2,
This is response data from the central office to SD8.

To give an example, suppose that the transmission request signal TREQl is generated in the interval ■, and the SDI is first transmitted in the interval ■1.1:, and when this is normally received by the central station, the response signal RDI is sent to the central station. It will be returned by the station. Next, another transmission request signal TREQ2 was generated in section ■, and when transmit modulation data SD2 was transmitted in section ■, the line deteriorated as shown by diagonal lines in the figure, and normal reception could not be performed at the central station. In this case, since the response signal RD2 is not returned, the retransmission control circuit shown in FIG. 4 operates and stochastically determines the retransmission of each section after section (2). In this example, the retransmission command output RETRY in FIG. 4 happens to go to L level in section ■, and retransmission is not performed. However, if another transmission request signal THEQ8 is generated at this time, there are two in the transmission queue. Data (TRE
(corresponding to Q2 and TREQ3) are connected, and the number TQ of data waiting to be sent increases to 2. Further, since the value of the line evaluation value output RXM becomes smaller due to the occurrence of an error in the RD in the section ■■■, the retransmission probability after the section ■ increases, and for example, retransmission is repeated until the sections ■■■ and RD2 are normally received. In this way, when RD2 is received in section ■, the value of TQ decreases by l(1,000), but the value of RXM becomes even smaller due to the occurrence of an error in RD in section ■■, so the retransmission probability is large.
For example, the retransmission probability will not be lowered until the response signal Rp8 is obtained by performing retransmission in the interval [phase] following the interval (2), that is, until the error in RD is reduced.

Figure 6 shows two data terminal stations a and b when the line quality is good.
This is an example time chart of operation when there is a conflict between
It is similar to the figure. In order to distinguish the signals and data of stations A and B, add (a) and (b) to their respective names, and use signals that can be assumed to be the same for both stations A and B, such as RD and T. (a, b) is added to indicate that they are common. In the current section ■, when station a and station b happen to be close in time, transmission request signals TREQ1(a) and TREQ are sent.
If I(b) occurs, both stations transmit modulated data 5Dl(a) and 5DI(b) at the same time in interval (This is indicated by diagonal lines in SD).After this, both stations perform retransmission control from section ■, but in this example, unfortunately, the retransmission probability results of both stations coincide over section ■■■. In this case, it is obvious that there is no response from the central station during this period.If the next transmission request signal TREQ2(b) has already been generated at station b in section 5, then the transmission request signal TREQ2(b) from station b is waiting for transmission. The value of the number TQ of terminal transmission data is 1.
2 (-Reinforcement) The retransmission probability is greater than that of station a. In this example, in interval ■, station b has SDl
(b) Retransmission: Two successes and a response signal RD 1 (b) is obtained. This can also be detected at station a (Nio I/),
And the line quality is good, so there is no possibility of delivery failure due to competition.
Since this can be identified by both stations a and b, it is possible to shift to an alternate transmission mode (two modes that do not depend on the retransmission probability) using the SET and RESET inputs of FF14 in FIG. A station sends a valid response to its own station (the inquiring side refrains from the next transmission, and vice versa; the station that detects a valid response to the other station executes the next transmission (1゜next) The described line supervisory control method will be examined from the viewpoint of reducing the average time of retransmission operations (two or more) due to contention between two data terminal stations. The maximum number is N. In the conventional method, one of the N opportunities is randomly selected as the retransmission time. The non-delivery rate P□ is the number of cases in which the re-delivery times chosen by both parties happen to coincide, so P = N' XN = N ...0...-(5
) 1 ~. On the other hand, the non-delivery probability P2 according to the method of the present invention depends on the case where the probability trial results of retransmission commands in every interval are the same over <N-1 intervals except the last, so the retransmission probability P (
If A>B) is written as P□ for simplicity, it becomes -1 (2PB, -2PR+1) (6).

P that satisfies p, , , < p for various values of N
The range of FL is, for example, as N increases, the range of values of P□ that satisfies the condition that the non-delivery rate in the case of the method of the present invention is smaller than the non-delivery rate in the case of the conventional method increases.For example, P
□2%, when N-10, 1-3 P□=10 P2=(%)=1.95X10,
It can be seen that the method of the present invention provides an improvement of nearly two orders of magnitude. From this, it is clear that, conversely, the maximum number of retransmission opportunities N that gives the same non-delivery rate (P1=P2) is smaller when the present invention is used, and the average data delay time is reduced.

As explained in detail above, according to the line monitoring control method of the present invention, it is possible to roughly identify whether the cause of data loss is line disturbance or contention, and in any case, data loss can be caused by line disturbance or contention. This has the advantage of preventing deterioration of remoteness ratio and reducing the average data delay time. Furthermore, in order to realize the method of the present invention, no equipment other than the conventionally required equipment (central office and data terminal station) is required, and functions such as a simple line evaluation circuit and retransmission control circuit can be implemented at the data terminal station. This has the advantage of being an economically inexpensive addition since it can be easily implemented using software.

[Brief explanation of drawings]

Fig. 1 shows the structure of a data transmission device of a data terminal using the method of the present invention, Fig. 2 shows a time chart of an example of the received data line monitoring operation of the data transmission control section in Fig. 1, and Fig. 8 shows Figure 2 shows an example of the configuration of a line evaluation circuit that obtains the line evaluation value output, Figure 4 shows an example of the configuration of the retransmission control circuit of the data transmission control section in Figure 1, and Figure 5 shows data when line quality deteriorates. FIG. 6 is a time chart of an example of a retransmission control operation of a terminal station. FIG. 6 is a time chart of an example of a retransmission control operation when there is contention between two data terminal stations. l...Receiving section, 2...Data transmission control section, 8
...Transmitter, 4...LPF, 5...A/
D converter, 6, 7...Delay circuit, 8.9...
OR gate, 10... Reversible counter, 12...
Random pattern generation circuit, 13... comparator,
■4...Flip-flop, FDP@kai/conflict synchronization detection pulse, FEP...-frame synchronization error pulse, RXM...
・Line evaluation value output, RXA: Reception level signal, RD: Demodulated data output, RXIN: Reception data input, RXD: Reception data output to terminal, SD:
Transmission modulation data, TXD...Terminal transmission data, TXOUT...Transmission output signal, TREQ...Transmission request signal, TXG...Transmission start signal, Patent applicant Kokusai Denki Co., Ltd. Agent Dai Ball 1 off-campus person 1st flash 2 Figure 3 Flash (A) (8) Figure 4 ESET

Claims (1)

[Claims] Each of the multiple terminal stations connected to one central station via a common transmission line and in a mutually competitive relationship synchronizes the average value of the received input signal level and the received demodulation data from the central station within a certain period of time in the past. A circuit that constantly monitors the quality of the transmission line using a means to detect patterns and the number of data errors, and a circuit that constantly monitors the quality of the transmission line using a means to detect patterns and the number of data errors, and when a data error occurs due to contention or line deterioration when the terminal station shifts to transmission operation. The circuit is equipped with a circuit that probabilistically determines whether to retransmit or not to retransmit data each time a predetermined transmission opportunity arrives, and the retransmission probability is determined based on the monitoring results of the receiving line and the transmission waiting in the data terminal station. It operates based on the number of data, and when the transmission line deteriorates or the number of data waiting to be sent increases, the retransmission probability is increased to ensure data reception, and when the line quality is good, it identifies a contention state. A terminal station that obtains a valid response from the central station for itself shall refrain from the next retransmission, and a terminal station that has identified the central station's response for another station shall perform the next retransmission to reduce the duration of the contention. A data terminal station line monitoring and control method characterized by the following.