JP3755430B2 - Transmitter - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a transmission apparatus that transmits an optical signal in a communication system that performs half-duplex communication using an optical signal.
[0002]
[Prior art]
For example, as disclosed in Japanese Patent Application Laid-Open No. 8-56198 (Patent No. 3137167), it is attached to the ceiling of an office and connected to a network, for example, an information processing device on a desk, etc. There is known a communication system that performs light communication with a slave unit. Such a communication system is used, for example, for relaying between a 10 Mbps network such as IEEE802.3 and an information processing apparatus.
[0003]
In the invention disclosed in this publication, stable full-duplex communication is realized by canceling the reflection component of the optical signal from the parent device that is reflected around and returns to the parent device.
[0004]
Recently, since a 100 Mbps network such as IEEE802.3u has become widespread, it is desired to use a communication system using light for relaying such a network.
[0005]
By the way, in the invention disclosed in the above-mentioned publication, stable communication cannot be performed simply by improving the transmission rate between the parent device and the child device.
[0006]
In the conventional 10 Mbps communication system, the modulation frequency of the optical signal is 12.5 MHz, and the length of one wavelength of the optical signal (the length of the signal phase is 360 degrees different) is 24 m (= 300,000 km / 12. 5 MHz). For this reason, the phase of the signal light from the master unit and the reflected light reflected from the outside world and returning to the master unit was relatively small. For this reason, the processing load of the process of canceling the return light is not so large.
[0007]
On the other hand, in the communication system of 100 Mbps, the modulation frequency of the optical signal is 125 MHz. For this reason, the length of one wavelength of the optical signal is about 2.4 m (= 300,000 km / 125 MHz), and depending on the reflection position, the phase of the reflected light that is reflected by the signal light from the base unit and returned to the base unit Becomes relatively large, and the processing load for canceling the return light increases.
[0008]
For this reason, the applicant of the present application has previously proposed a communication system that performs half-duplex communication between a parent device and a child device (Japanese Patent Application No. 2000-237286).
[0009]
[Problems to be solved by the invention]
However, as described above, in the case of performing half-duplex communication, a slave unit that is to transmit data transmits a line request to the master unit, and has a transmission permission from the master unit in accordance with the request. Until then, data transmission cannot be started.
[0010]
For this reason, the slave unit must be provided with a buffer and store packet data supplied from the information processing apparatus or the like.
[0011]
However, if the time from when the slave unit sends a line request to the master unit and the transmission of packet data starts in response to the transmission permission for this becomes longer than the packet interval, the next time before transmission of the packet held in the buffer is completed. Packets may be supplied. In such a case, it is necessary to generate a collision signal in a pseudo manner or to request retransmission of the packet to the information processing apparatus side. However, these processes reduce throughput. End up.
[0012]
Considering the lock time of a PLL (Phase Locked Loop) circuit used for generating a synchronization signal on the optical signal receiving side, the transmission of the data of the packet from the start of reception of the packet from the information processing apparatus or the like It is difficult to reduce the time to start to below the minimum packet interval.
[0013]
The present invention has been made in view of the above-described overwork, and an object thereof is to provide a transmission apparatus that can contribute to an improvement in throughput.
[0014]
[Means for Solving the Problems]
  In order to solve the above-described problem, a transmission device according to the present invention provides:It is configured in units of packets between the terminal device and at least one or more slave units, and between the slave unit and the master unit.Data to sendCan be transmitted at substantially the same transmission rate, and the data is transmitted between the slave unit and the master unit.A transmission device that transmits via a half-duplex optical communication transmission line,
  In order to temporarily store the data from the terminal device in the slave unit, the minimum packet interval of the data is provided corresponding to a case where the time until the transmission of the data to the master unit is shorterA plurality of storage means;
  A write control means for sequentially writing the data from the terminal device by switching the plurality of storage means for each packet;
  Read control means for switching the plurality of storage means and sequentially reading the data stored in the plurality of storage means for each packet when the slave unit has acquired a transmission right to the master unit;
  SaidRead from multiple storage meansSaidDataThe abovePer packetTo the aboveVia a half-duplex optical transmission lineIn the base unitWith transmission means to send,
WithTheIt is characterized by that.
[0015]
  Further, the transmission device according to the present invention is:SaidA plurality of storage means each for the transmission right;ofTime required for acquisitionAnd saidMinimum packet interval for dataAnd obtained based onHave storage capacityRukoIt is characterized by.
[0016]
DETAILED DESCRIPTION OF THE INVENTION
The present invention can be applied to a communication system in which a terminal device such as a personal computer is connected to a network such as a LAN (local area network) by optical communication.
[0017]
For example, as shown in FIG. 1, this communication system can perform relay between a backbone network (backbone) 1 and a terminal device 2 including an information processing device such as a personal computer. A master unit 10 connected to the backbone 1 and a slave unit 20 to which the terminal device 2 is connected are provided. The subunit | mobile_unit 20 is installed on the desk etc. of an office, for example, and the main | base station 10 is installed, for example on the ceiling of an office. In addition, a plurality of slave units 20 can communicate with one master unit 10.
[0018]
The backbone 1 is composed of a network using, for example, IEEE 802.3u known as 100BASE-T as a lower layer protocol. Further, the terminal device 2 and the slave unit 20 are also connected by the same network. The transmission rate of IEEE802.3u is 100 Mbps.
[0019]
Therefore, in order to relay between the backbone 1 and the terminal device 2, a bandwidth of 100 Mbps or more is necessary. However, in communication using light, there is a slight overhead due to the addition of an optical preamble or the like compared to communication in IEEE 802.3u. Therefore, in this communication system, an optical signal in communication between the parent device 10 and the child device 20 is used. The modulation frequency is set to 125 MHz.
[0020]
Further, each of the parent device 10 and the child device 20 can transmit so-called back pressure for suppressing transmission of data to the backbone 1 and the terminal device 2 by IEEE802.3x.
[0021]
The base unit 10 includes an uplink system 100 that transmits data destined for the backbone 1 to the backbone 1, a downlink system 110 that transmits data destined for the slave unit 20 to the slave unit 20, and half of the base unit 10. A half-duplex control system 120 that controls the duplex communication and a control unit 150 that controls the operation of the entire base unit 10 are provided.
[0022]
  Further, the slave unit 20 transmits uplink data 2 addressed to the master unit 10 to the master unit 10.10 and downlink system 2 for transmitting data addressed to terminal device 2 to terminal device 200 and a half-duplex control system 220 that controls half-duplex communication between the master unit 10 and a control unit 250 that controls the operation of the entire slave unit 20.
[0023]
For example, as shown in FIG. 2, the uplink system 100 of the parent device 10 includes a light receiving unit 101 that receives an optical signal from the child device 20 and a receiving unit 102 that extracts reception data from the optical signal received by the light receiving unit 101. A switching unit 103 that switches data to be transmitted to the backbone 1, a line request unit 104 that requests a line, a transmission data detection unit 105 that detects data addressed to the backbone 1, and a transmission that transmits data to the backbone 1 Part 106.
[0024]
Further, the downlink system 110 of the parent device 10 transmits the data from the receiving unit 111 that receives data from the backbone 1, the switching unit 113 that switches data to be transmitted to the child device 20, and the data from the switching unit 113. And a light emitting unit 115 that transmits transmission data from the transmission unit 114 as an optical signal.
[0025]
The light receiving unit 101 can receive light signals from a wide angle (for example, about 60 to 150 degrees), and the light emitting unit 115 can emit light signals at a wide angle. ing. Thereby, communication with the some subunit | mobile_unit 20 installed as mentioned above can be performed now.
[0026]
The half-duplex control system 120 of the base unit 10 detects a line request from the slave unit 20 and controls half-duplex communication according to the line request detected by the line request detection unit 121. A half-duplex processing unit 122 that performs the detection, a use permission processing unit that executes detection of an empty optical line, generation of a line use permission signal for the slave unit 20 in accordance with instructions from the half-duplex processing unit 122 and the control unit 150 123.
[0027]
Further, base unit 10 receives received data detection unit 131 that detects data addressed to handset 20 and a pseudo signal (retransmission request, collision signal, back pressure, for example) that can retransmit the data in accordance with control from control unit 150. And the like.
[0028]
Further, the downlink system 200 of the slave unit 20 includes, for example, a light receiving unit 201 that receives an optical signal from the base unit 10 and a reception that extracts received data from the optical signal received by the light receiving unit 201, as illustrated in FIG. Unit 202, switching unit 203 that switches data to be transmitted to terminal device 2, line request unit 204 that requests transmission permission to base unit 10, transmission data detection unit 205 that detects data addressed to terminal device 2, And a transmission unit 206 for transmitting data to the terminal device 2.
[0029]
  The uplink system 210 of the slave unit 20 includes a receiving unit 211 that receives data from the terminal device 2, a buffer 212a and a buffer 212b that hold data from the terminal device 2,Base unit 10A switching unit 213 for switching data to be transmitted to the transmission unit, a transmission unit 214 for transmitting data from the switching unit 213, and a light emitting unit 215 for transmitting transmission data from the transmission unit 214 as an optical signal.
[0030]
Since the slave unit 20 only needs to be able to communicate with one master unit 10, the optical signal received by the light receiving unit 201 is limited to a narrow angle (for example, about 5 to 60 degrees) as compared to the light receiving unit 101. The range in which the light emitting unit 215 emits an optical signal is also limited to a narrow angle as compared with the light emitting unit 115.
[0031]
The half-duplex control system 220 of the slave unit 20 includes a use permission detection unit 221 that detects use permission of the optical line from the base unit 10, permission to transmit data, and the use permission detected by the use permission detection unit 221. Accordingly, a half-duplex processing unit 222 that controls half-duplex communication and a line request processing unit 223 that executes generation of a line request signal for the base unit 10 and the like are provided.
[0032]
Further, the slave unit 20 receives, for example, a received data detection unit 231 that detects data addressed to the master unit 10 and a pseudo signal (retransmission request, collision signal, back pressure) that can retransmit the data in accordance with control from the control unit 250. And the like, a line request unit 241 for generating a line request for the base unit 10, a write control unit 242 for controlling writing to and reading from the buffer 212a and the buffer 212b, and a read control unit 243. I have.
[0033]
By the way, as described above, communication is performed between the slave unit 20 and the terminal device 2 using, for example, IEEE 802.3u. In IEEE802.3u, data is transmitted and received in units of packets of a predetermined length.
[0034]
For example, as shown in FIG. 4, the packet includes a header that stores a preamble, a destination address, a transmission source address, and the like, and a payload that stores transmission data itself. An interval (interframe gap) of at least 12 octets (12 bytes) is provided between adjacent packets. In the case of IEEE 802.3u, since one bit is 0.01 μSec, the minimum interval (Tmin: minimum packet interval) is 0.96 (= 0.01 × 12 × 8) μSec.
[0035]
Further, optical signals are transmitted and received between the parent device 10 and the child device 20 in units of packets having a predetermined length. As shown in FIG. 5, this packet has a 50 at the beginning of the packet in order to lock the phase of the PLL built in the receiving unit 102 of the parent device 10 that receives this packet and the receiving unit 202 of the child device 20. An optical preamble of about symbol (50 bits) is provided.
[0036]
In a communication system that performs communication using an optical signal, the reflected optical signal may be incident on the light receiving unit of the receiving apparatus depending on the surrounding environment or the like. For this reason, when trying to improve the quality of communication, for example, as disclosed in Japanese Patent Application Laid-Open No. 8-56198 (Patent No. 3137167), the influence of reflected light on the receiving side device (particularly the master unit) is reduced. It is necessary to perform a process (returning light cancellation process) to prevent the reception.
[0037]
By the way, in a conventional communication system using an optical signal, for example, IEEE 803.2 or the like having a transmission rate of 10 Mbps is intended to be relayed. Therefore, the modulation frequency of the optical signal is, for example, about 12.5 MHz. In this case, one wavelength of the optical signal (the length in which the phase of the signal is different by 360 degrees) is about 24 m (= 300,000 km / 12.5 MHz). For this reason, in a normal use environment such as an office, the phase difference between the direct light that is directly incident on the light receiving unit from the light emitting unit and the reflected light that is reflected from the outside and enters the light receiving unit from the light emitting unit is the wavelength of the optical signal. Was relatively small. Under such conditions, even if the return light cancellation processing as described above is executed in the receiving device, the processing load is not so high, and full-duplex communication can be realized.
[0038]
On the other hand, in this communication system, since the modulation frequency of the optical signal is 125 MHz as described above, one wavelength is 2.4 m (= 30 km / 12.5 MHz). For this reason, since one wavelength of the optical signal is small, the phase difference between the direct light and the reflected light becomes large even under the same installation conditions. Therefore, similarly to the above, if the receiving side apparatus tries to execute the return light cancellation process, the processing load increases and it becomes difficult to realize full-duplex communication.
[0039]
For this reason, in this communication system, communication between the parent device 10 and the child device 20 is performed by half-duplex communication. In this half-duplex communication, the side that intends to transmit data requests the other party to transmit and starts transmission after obtaining transmission permission.
[0040]
In addition, by performing half-duplex communication in this way, the timing of the optical signal from the own device and the timing of the optical signal from the communication partner do not overlap in time as in full-duplex communication. Thus, it is not necessary to perform the complicated return light cancellation process as described above.
[0041]
By the way, in order to realize half-duplex communication, the master unit 10 can be configured to transmit a line request when transmitting data to the slave unit 20. However, when the configuration is such that a plurality of slave units 20 can communicate with one master unit 10 as described above, transmission control between the slave units 20 is performed when communication control is performed on the master unit 10 side. It is considered that adjustment and the like are facilitated. Therefore, hereinafter, a case where the slave unit 20 transmits a line request when transmitting data to the master unit 10 will be described.
[0042]
In the half-duplex communication, when the slave unit 20 transmits data, a line request is transmitted to the master unit 10. This line request is configured according to the format shown in FIG. When such a line request is received, the parent device 10 transmits a transmission permission to the child device 20 if it is in a receivable state. Having received such a transmission permission, the slave unit 20 transmits the data to be transmitted to the master unit 10 as an optical signal.
[0043]
When the handset 20 receives a packet from the terminal device 2, the handset 20 tries to send the received packet to the base unit 10. However, as described above, the handset 20 sends a line request and receives a transmission permission for this request. Cannot transmit the data of the packet In such a case, it is common to provide a first-in first-out (FIFO) type buffer for holding the packet data from the terminal device 2 in the slave unit 20. When receiving the packet from the terminal device 2, the handset 20 stores the received packet data in a buffer and transmits a line request to the base unit 10.
[0044]
If the interval between the packets is longer than the time until the line request, transmission permission, and data transmission start, the packet data stored in the buffer is transmitted before the next packet data is supplied from the terminal device 2. In order to achieve this, one buffer is sufficient.
[0045]
However, as described above, in half-duplex communication, it takes time to start transmission of a line request, transmission permission, and data, so the time from the start of packet reception to the start of transmission is In some cases, the minimum packet interval (Tmin) may be exceeded. In such a case, if only one FIFO type buffer is provided, if the next packet is simply written in the buffer, the boundary of the packet in the buffer becomes unknown. For this reason, a packet cannot be transmitted correctly.
[0046]
For this reason, in this communication system, as shown in FIG. 3 described above, two FIFO type buffers 212a (first storage means) and 212b (second storage means) are provided on the handset 20 side. These buffers 212 a and 212 b may be configured as, for example, a single FIFO memory, or may be configured by a built-in memory of a processor constituting the control unit 250.
[0047]
The write control unit 242 controls which buffer 212 a and 212 b stores the received packet data in accordance with an instruction from the control unit 250. The write control by the write control unit 242 is executed according to a flowchart shown in FIG. 6, for example.
[0048]
When the power of the slave unit 20 is turned on, the processing from step S1 in the figure is executed, and the control unit 250 waits for reception of a packet. When a packet from the terminal device 2 is supplied to the reception unit 211, the reception data detection unit 231 detects reception of the packet and notifies the control unit 250 of the reception. In response to this, the control unit 250 determines whether or not the received packet is an odd number (step S2). If it is an odd-numbered packet, the control unit 250 executes the processing from step S3 to step S5, and holds the data of the packet supplied from the reception unit 211 in the buffer 212a via the write control unit 242. On the other hand, if the packet is not an odd-numbered packet, the control unit 250 executes the processing from step S6 to step S8, and holds the data of the packet supplied from the reception unit 211 in the buffer 212b via the write control unit 242. Let
[0049]
Through the above processing, the packet data from the terminal device 2 is alternately stored in the buffer 212a and the buffer 212b.
[0050]
In addition, reading of the packet data stored in the buffers 212 a and 212 b is controlled by the read control unit 243 in accordance with control from the control unit 250. The read control by the read control unit 243 is executed according to, for example, the flowchart shown in FIG.
[0051]
When the power of the slave unit 20 is turned on, the processing from step S11 in FIG. 7 is executed in parallel with the above-described write control, and the control unit 250 waits for a request from the line request unit 241. When the packet from the terminal device 2 is supplied to the receiving unit 211, the line request unit 241 requests the control unit 250 to transmit a line request.
[0052]
When transmission of the line request is requested, the control unit 250 detects whether or not the optical line is free according to the use permission of the optical line from the use permission detection unit 221. If the optical line is free, the control unit 250 proceeds to step S15 and transmits a line request via the line request processing unit 223. In parallel with this, the control unit 250 instructs the switching unit 213 to select data from the line request processing unit 223. As a result, the line request generated by the line request processing unit 223 is supplied to the light receiving unit 101 via the transmission unit 214 and the light emitting unit 215.
[0053]
Such a line request is supplied to the use permission processing unit 123 via the light receiving unit 101, the receiving unit 102, and the line request detecting unit 121. In response to the supplied line request, the use permission processing unit 123 instructs the generation of the use permission in accordance with the control from the control unit 150 if the optical line is free. When the received data detection unit 131 does not detect a packet from the backbone 1, the control unit 150 controls the switching unit 113 to supply the output of the use permission processing unit 123 to the transmission unit 114. . For this reason, the use permission from the use permission processing unit 123 is supplied to the light receiving unit 201 via the switching unit 113, the transmission unit 114, and the light emitting unit 115.
[0054]
On the other hand, when the optical line is not available, waiting processing (steps S13 and S14) is executed. However, when the waiting state exceeds a predetermined time, the terminal device 2 is requested to retransmit the packet as a timeout (steps S19 and S20).
[0055]
After transmitting the line request, the control unit 250 waits for the use permission detection unit 221 to detect transmission permission from the parent device 10.
[0056]
At this time, waiting processing (steps S17 and S18) is executed within a predetermined time. However, when the waiting state exceeds a predetermined time, the terminal device 2 is requested to retransmit the packet as a timeout (steps S19 and S20).
[0057]
When the use permission detection unit 221 detects transmission permission from the parent device 10, the control unit 250 determines whether or not the packet to be sent is an odd number (step S21). If the packet is an odd-numbered packet, the control unit 250 executes the processing from step S22 to step S24, and causes the packet data held in the buffer 212a to be read via the read control unit 243. On the other hand, if the packet is not an odd-numbered packet, the control unit 250 executes the processing from step S25 to step S27, and causes the packet data held in the buffer 212b to be read via the read control unit 243.
[0058]
When the reading of data from the selected buffers 212a and 212b is completed, the control unit 250 ends the transmission process (step S28).
[0059]
Through the above processing, the data held in the buffer 212a and the buffer 212b are alternately read and supplied to the switching unit 213. The switching unit 213 supplies the packet data supplied from the buffers 212 a and 212 b to the transmission unit 214 in accordance with an instruction from the control unit 250. The transmission unit 214 drives the light emitting unit 215 according to the supplied data to transmit an optical signal. Thus, an optical signal corresponding to the packet data from the terminal device 2 is supplied to the parent device 10 side. This optical signal is received by the light receiving unit 101, and packet data is reproduced by the receiving unit 102. The reproduced data is supplied to the backbone 1 via the switching unit 103 and the transmission unit 106.
[0060]
By the way, the sizes of the buffers 212a and 212b described above are actually read from the buffers 212a and 212b after the packet request is transmitted to the buffer 212a or 212b after the packet is supplied from the terminal device 2. It is determined according to the time obtained by subtracting the above-mentioned minimum inter-packet interval (Tmin) from twice the time (ΔT) until transmission of the received data starts.
[0061]
That is, for example, as shown in FIG. 8, when receiving the packet from the terminal device 2, the slave unit 20 transmits the line request to the master unit 10 while storing the received packet data in the buffer. In response to this, when transmission permission is supplied from the parent device 10, the child device 20 starts transmitting packet data held in the buffer.
[0062]
Therefore, the terminal device 2 supplies the buffer (for example, the buffer 212a) that holds the first packet at least from the start of packet reception to the start of transmission of the received packet (ΔT in FIG. 8). A capacity capable of holding the data to be stored is required.
[0063]
In addition, when the next packet is received, the handset 20 stores the data of the received packet in the other buffer (for example, the buffer 212b). If transmission of the previous packet is not completed when reception of this packet is started, the line request is started after transmission of the previous packet is completed.
[0064]
When the transmission rate (r1) of the packet supplied from the terminal device 2 is the same as the transmission rate (r2) of the packet transmitted to the parent device 10, the transmission end of the previous packet is terminated. After Δt.
[0065]
In addition, since there is at least the above-mentioned minimum packet interval (Tmin) between the previous packet and the subsequent packet, a line request for the subsequent packet is transmitted after Δt−Tmin from the start of reception of the subsequent packet. become. In response to this, the slave unit 20 starts transmitting a subsequent packet in response to the transmission permission supplied from the master unit 10 after Δt.
[0066]
Therefore, the buffer (for example, the buffer 212b) that holds the next packet needs to have a capacity capable of holding the data supplied from the terminal device 2 during 2 × Δt−Tmin that is the sum of these delay times. .
[0067]
By the way, when packets from the terminal device 2 are continuously supplied, it is not clear which buffer 212a, 212b of the first packet supplied at the first packet interval is stored.
[0068]
Therefore, if at least the capacity of each of the buffers 212a and 212b is set to be equal to or larger than the capacity capable of holding the data supplied from the terminal device 2 during 2 × Δt−Tmin, it is possible to receive two packets at the minimum packet interval. Even if it exists, a packet can be transmitted reliably. As described above, if the packet can be transmitted reliably, the frequency of retransmission of the packet, the collision signal generated in a pseudo manner, and the like are reduced, which can contribute to an improvement in throughput.
[0069]
Further, the third packet is held in a buffer (for example, the buffer 212a) that holds the first packet. The line request for this packet is received after the transmission of the second packet (that is, reception of the second packet). Only after 2 × Δt−Tmin) from the end can it be transmitted.
[0070]
There is at least a minimum packet interval Tmin from the end of reception of the second packet to the start of reception of the third packet. Therefore, it takes 2 × Δt−2 × Tmin from the start of receiving the third packet to the start of transmission of the line request for the packet, and the transmission of the packet is started after Δt from the start of transmission of the line request. The
[0071]
Therefore, in order to reliably transmit the third packet supplied at the minimum packet interval to the base unit 10, the terminal device 2 supplies the buffers 212a and 212b to at least 3 × Δt−2 × Tmin. A capacity capable of holding the data to be stored is required.
[0072]
By the way, under normal conditions, the number of consecutive packets supplied at the minimum packet interval is not so large. In addition, when the number of consecutive packets supplied at the minimum packet interval increases, other problems such as frequent packet collisions and traffic congestion often occur. It is necessary to plan. For this reason, if two to three packets supplied at the minimum packet interval can be reliably transmitted, a packet loss rate sufficient for practical use can often be realized.
[0073]
Therefore, by configuring the capacities of the buffers 212a and 212b as described above, the packet loss rate can be reduced to a practically sufficient level while minimizing an increase in cost due to an increase in the buffer capacity.
[0074]
By the way, in IEEE 802.3u, retransmission of a packet can be realized by hardware, but a retransmission request for requesting retransmission is a packet transmission source within a predetermined time (T5 = 5.12 μSec). For example, it is necessary to supply to the slave unit 20 when viewed from the backbone 1.
[0075]
Also, the retransmission request supplied from the backbone 1 to the terminal device 2 via the parent device 10 and the child device 20 is transmitted and received by half-duplex communication as described above. For this reason, T6 is the time required for transmission of a line request, transmission permission, etc. from the parent device 10 to the child device 20.
[0076]
In such an environment, a packet whose transmission is not started even after the time T5-T6 has elapsed since the slave unit 20 started reception does not reach the backbone 1, and a retransmission request is sent from the backbone 1 side. Is not delivered to the terminal device 2 within a predetermined time T5.
[0077]
For this reason, even if the buffers 212a and 212b have a capacity larger than the data supplied from the terminal device 2 at a time equal to or greater than T5, it is not possible to respond to the retransmission request and is meaningless. Therefore, it is sufficient that the buffers 212a and 212b have a capacity of data supplied from the terminal device 2 during T5-T6.
[0078]
Therefore, in this communication system, the upper limit of the capacity of the buffers 212a and 212b is the capacity of data supplied from the terminal device 2 during T5-T6.
[0079]
In this communication system, by setting the upper limit of the capacities of the buffers 212a and 212b as described above, it is possible to prevent the buffer capacity from being increased more than necessary and the cost of the apparatus from being increased unnecessarily.
[0080]
By the way, when only a single buffer for holding packet data is provided in the uplink system of the slave unit 20, a packet loss occurs when the interval between packets is wide as shown in FIG. 9, for example. However, for example, as shown in FIG. 10, a problem may occur when packets with narrow intervals between packets are continuously supplied.
[0081]
When a FIFO type buffer is used as described above, if the next packet is supplied to the buffer before the transmission of the previous packet is completed, the uplink transmitter detects the packet boundary. It is conceivable that the data of two packets may be transmitted continuously as shown in FIG.
[0082]
In order to avoid such a situation, it is conceivable to prepare a buffer capable of randomly inputting / outputting individual data instead of a FIFO buffer. However, when such a buffer is used, it is necessary to individually manage the write address, read address, packet size, and the like of the area holding each packet, which increases the control load of the control unit and the like. . In addition, since a buffer having a large capacity compared to the packet size is required, the cost of the apparatus is unnecessarily increased.
[0083]
On the other hand, in the above-described embodiment, two systems of the buffer 212a and the buffer 212b are provided in the uplink system of the handset 20, and the packet data from the terminal device 2 is alternately held in these buffers 212a and 212b. I am doing so. Further, since the capacities of these buffers 212a and 212b are set as described above, even if two to three packets are supplied to the slave unit 20 at the minimum packet interval, each packet is surely confirmed. Can be transmitted to the base unit 10.
[0084]
Further, since each of the buffers 212a and 212b is a FIFO type buffer, the increase in control load is about the selection of the buffers 212a and 212b for storing packets.
[0085]
Therefore, this communication system can contribute to improvement in reliability while suppressing an increase in control load. Also, by improving the reliability, the frequency of retransmission of packets, the frequency of collision signals that are generated in a pseudo manner, and the like are reduced, which can contribute to an improvement in throughput.
[0086]
By the way, in the above description, it is assumed that the transmission rate (r1) of packets supplied from the terminal device 2 to the child device 20 is the same as the transmission rate (r2) of packets transmitted from the child device 20 to the parent device 10. Was. However, in the communication system described above, the transmission rate with the terminal device 2 is 100 Mbps, but since the modulation frequency of the optical signal is 125 MHz, a simple transmission rate is a packet transmitted from the slave unit 20 to the master unit 10. Is faster. However, since the above-mentioned optical preamble is added to the optical signal, it differs depending on the data length of the packet.
[0087]
Therefore, when the ratio of the transmission rates is specified due to circumstances such as limitation of the packet data length, the capacities of the buffers 212a and 212b may be adjusted accordingly.
[0088]
In the above description, an example in which two buffers 212a and 212b are provided in the uplink system 210 of the slave unit 20 has been described. However, the number of buffers is not limited to two, and an appropriate number of buffers should be provided. Can do.
[0089]
In this case, the capacity of the n buffers is equal to the time (Δt) required for acquiring the packet transmission right to the parent device 10 and the minimum interval between packets (minimum packet interval), as in FIG. Tmin).
[0090]
For example, when three buffers are provided, if the transmission rate of packets supplied from the terminal device 2 to the slave unit 20 and the transmission rate of packets transmitted from the slave unit 20 to the master unit 10 are the same, The lower limit of the capacity is 3 × Δt−2 × Tmin. When n buffers are provided under the same conditions, the lower limit of the capacity of each buffer is n × Δt− (n−1) × Tmin.
[0091]
Further, when the transmission rate of the packet supplied from the terminal device 2 to the child device 20 and the transmission rate of the packet transmitted from the child device 20 to the parent device 10 are different, the capacity of the buffer is appropriately set according to the difference in the transmission rate. change.
[0092]
As described above, since the period during which the retransmission request is to be transmitted is limited, the upper limit of the capacity of each buffer is obtained by subtracting the time (Δt) required to acquire the packet transmission right from this limit time. It is sufficient if it is sufficient for the capacity of data to be received within the specified time.
[0093]
【The invention's effect】
In the transmission apparatus of the present invention, the plurality of storage units read the data to be transmitted configured in units of packets for each packet, and the already stored data is sequentially read for each packet. The data read from the means is transmitted for each packet after acquiring the transmission right of the half-duplex optical communication transmission line.
[0094]
As a result, data to be transmitted is read into each storage means for each packet, and transmitted for each packet. For this reason, packet loss can be reduced compared with the case where only one buffer is provided, and the reliability of packet transmission can be improved. If the certainty of packet transmission is improved, the frequency of retransmission, the collision signal that is generated in a pseudo manner, and the like are also reduced, which can contribute to an improvement in throughput.
[0095]
  Also, a plurality of storage meansIs, Each transmission rightofTime required for acquisitionWhen,Minimum packet interval for dataAnd obtained based onHave storage capacityRukoAs a result, the throughput can be improved with the minimum necessary storage capacity. Thereby, it can contribute to the reduction of the cost of an apparatus.
[Brief description of the drawings]
FIG. 1 is a diagram showing a configuration of a communication system according to an embodiment of the present invention.
FIG. 2 is a block diagram showing a configuration of a base unit constituting the communication system.
FIG. 3 is a block diagram showing a configuration of a slave unit constituting the communication system.
FIG. 4 is a diagram showing a format of data transmitted / received between a slave unit and a terminal device constituting the communication system.
FIG. 5 is a diagram showing a format of data transmitted / received between a slave unit and a terminal device constituting the communication system.
FIG. 6 is a flowchart showing write control for a buffer in a slave unit constituting the communication system.
FIG. 7 is a flowchart showing a read control from a buffer in a slave unit configuring the communication system.
FIG. 8 is a diagram illustrating waveforms of main parts in communication between a parent device and a child device constituting the communication system.
FIG. 9 is a diagram illustrating waveforms of respective units in communication between a parent device and a child device when simply performing half-duplex communication when the interval between packets is wide.
FIG. 10 is a diagram illustrating waveforms of respective units in communication between a parent device and a child device when simply performing half-duplex communication when the interval between packets is narrow.
[Explanation of symbols]
1 ... Backbone
2 ... terminal device,
10 ... Master unit,
100 ... uplink system,
110 ... Downlink system,
120 ... half-duplex control system,
150 ... control system,
20 ... handset
200 ... Downlink system,
210 ... Uplink system,
212a, 212b, buffer,
220 ... half-duplex control system,
250 ... Control system

Claims (2)

The terminal device and at least one or more slave units, and between the slave unit and the master unit, configured to transmit data to be transmitted configured in units of packets at substantially the same transmission rate, and A transmitting device for transmitting the data between the slave unit and the master unit via a half-duplex optical communication transmission line,
In order to temporarily store the data from the terminal device in the slave unit, a plurality of data provided corresponding to a case where the minimum packet interval of the data is shorter than the time until transmission of the data to the master unit is shorter Storage means,
A write control means for sequentially writing the data from the terminal device by switching the plurality of storage means for each packet;
Read control means for switching the plurality of storage means and sequentially reading the data stored in the plurality of storage means for each packet when the slave unit has acquired a transmission right to the master unit;
And transmitting means for transmitting the data read out from said plurality of storage means, to said master unit via a pre-Symbol half duplex optical communication transmission path for each of the packet,
Transmitting apparatus characterized by comprising a. Said plurality of storage means, the transmission of claim 1, wherein the benzalkonium that Yusuke time and each required the acquisition of the transmission right, the are that the storage capacity obtained based on the minimum packet interval of the data apparatus.

Images