A long-distance ultra-low latency relay method for a half-duplex network

Abstract

The present invention discloses a long-distance ultra-low latency relay method for a half-duplex network. The method includes: the relay node uses a dual-channel hot redundant optical fiber as the signal transmission medium; performs multi-phase overclocking sampling on the received half-duplex network data, arranges the sampled data into parallel data in combination with k-codes, frame sequences, and CRC checks, performs serial-to-parallel conversion on the parallel data in accordance with the 8b10b coding format, and transmits the data signal after serial-to-parallel conversion through the dual-channel hot redundant optical fiber; the receiving end decodes the received optical signal, performs serial-to-parallel conversion, and obtains the required data after redundant reception processing of the obtained parallel data. The present invention can meet the transmission requirements of long distance and low latency for half-duplex networks.

Description

Technical Field

[0001] The present invention relates to the field of data transmission, and particularly to a relay implementation method for a half-duplex network, providing a long-distance, low-latency, and highly reliable relay solution for the half-duplex network. Background Art

[0002] A half-duplex network is a network in which data can be transmitted in two directions. Only one terminal is allowed to send data at the same time, and other terminals receive data. By switching the sending switches of each terminal time-divisionally, an effective connection is established for the entire network. For example, in using a half-duplex local area network, a workstation can send data on the line and then immediately receive data on the line. Sending and receiving data are achieved on the same line. Like a full-duplex network, a half-duplex network contains a time-division multiplexing two-way line. Compared with a full-duplex network, a half-duplex network reduces one cable transmission channel and has the characteristic of occupying fewer hardware lines. Especially in the fields of aviation and aerospace, the requirement for equipment weight reduction is getting higher and higher. Adopting a half-duplex network can effectively reduce the weight of the aircraft.

[0003] With the continuous development of communication network technology, the requirement for the number of network nodes is increasing. The load of the sending driver of a half-duplex network is limited. When the number of terminals plus the length of the cable exceeds the load capacity, effective communication cannot be carried out. To solve this problem, a relay processing method is adopted to perform a relay process on the network.

[0004] Traditional relays have certain limitations in communication distance, transmission latency, and communication stability.

[0005] Disadvantage 1: Traditional relays perform a series of decoding and encoding operations when processing data, and the transmission latency is in the order of microseconds.

[0006] Disadvantage 2: Due to factors such as interference and signal attenuation of communication cables, the transmission distance of a single relay is about 100 meters.

[0007] Disadvantage 3: During the relay process of traditional relays, data verification operations are not performed. If an error code occurs due to hardware, the error code cannot be repaired. Summary of the Invention

[0008] In view of the above deficiencies of the prior art, the present invention proposes a dual-redundancy relay method based on optical fiber. The method includes:

[0009] The relay node uses dual-channel hot-redundant optical fibers as the signal transmission medium; it performs multi-phase overclocking sampling on the received half-duplex network data, combines the sampled data with k-codes, frame sequences, and CRC checksums to arrange them into parallel data, performs parallel-to-serial conversion on the parallel data in accordance with the 8b10b encoding format, and transmits the data signal after parallel-to-serial conversion through the dual-channel hot-redundant optical fibers; the receiving end decodes the received optical signal and performs serial-to-parallel conversion, and obtains the required data after redundant reception processing on the obtained parallel data.

[0010] According to another aspect of the present invention, the method further includes: the relay node determines whether there is data to be sent based on the half-duplex network data obtained at its receiving end. When there is data to be sent, it enables transmission, disables reception, and starts transmission. When there is no data to be sent, it disables its transmission enable and enables reception.

[0011] According to still another aspect of the present invention, the method further includes: the parallel data is 32-bit parallel data, and its specific structure is: 8 bits are used for k-codes, frame sequences, and CRC checksums, and the remaining 24 bits are used to transmit 3 channels of half-duplex data.

[0012] According to yet another aspect of the present invention, the redundant reception processing specifically includes: when two identical frame numbers arrive in the two channels of the dual-channel, the frame that arrives first is adopted, and the frame with a CRC error is discarded; the polling method is used to process the write cache requests of the two channels. If there is an unfinished frame in the cache, and there is no frame number of the data to be written in the cache and the frame number can be continued with the previous frame, then the frame data is written into the cache and the frame number identifier is updated; if there is no unfinished frame in the cache, then while writing the data frame into the cache, it notifies the output cache end to start the mechanism of storing half-frame data and then outputting the data; when there is data in both channels at the same time, the output cache retains half-frame data to handle the difference in transmission time between the two channels. If the data frame of the channel with a faster transmission time has an error or the line is interrupted, it can be left for the channel with a slower transmission time for data handover operations.

[0013] The present invention directly samples the data of the half-duplex network by using the overclocking sampling method at the relay receiving end of the half-duplex network. The relay medium uses optical fiber transmission, which greatly shortens the delay. In addition, the encoding format uses 8b10b encoding, and the receiving end can effectively recover the clock of the sending end. At the same time, optical fiber transmission can effectively avoid crosstalk and signal attenuation in electrical signal transmission, and the maximum transmission distance exceeds 1000m. In addition, to improve the stability of the signal during the relay process and reduce the bit error rate, the dual-channel real-time redundancy method is used for the transmission of relay signals. Thus, it well solves the limitations existing in the traditional relay method in terms of communication distance, transmission delay, and communication stability. Brief Description of the Drawings

[0014] Figure 1 A schematic diagram of the relay method of the present invention is shown.

[0015] Figure 2 A schematic diagram of overclocking sampling according to the present invention is shown.

[0016] Figure 3 A timing diagram showing the processing of the relay sending end of the present invention is shown.

[0017] Figure 4 The frame format of the relay transmission of the present invention is shown.

[0018] Figure 5 The redundant receiving flow chart of the present invention is shown. Implementation

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0020] The implementation process of the relay method of the present invention is described in detail below with reference to the accompanying drawings:

[0021] Figure 1 The overall architecture of the relay method of the present invention is shown. The relay node samples the received half-duplex network data, arranges the sampled data into data frames according to a certain format, and transmits the signal through a dual-channel hot redundant optical fiber. The receiving party obtains the data sent by the relay node by de-redundanting the received optical signal through redundant receiving processing and decoding. The transmission and reception of optical signals can be implemented by using an optical module + Xilinx GTX.

[0022] 1) Sampling of half-duplex network data

[0023] Overclocking sampling is performed on the electrical signal on the half-duplex network, and the maximum rate supported for the half-duplex network is 200Mbps. The 200Mkz clock is divided into 8 different phases for overclocking sampling of the half-duplex network, which is equivalent to a sampling clock of 1600Mkz. The overclocking sampling method is shown in Figure 2 .

[0024] 2) Relay transmission of half-duplex network data

[0025] A single channel of a half-duplex network sends and receives data in a time-sharing manner. When there is no data to send, the current network sending channel should be closed, and when there is data to send, the current network receiving channel should be closed. When there is no data transmission in the entire half-duplex network, the receiving end level is 0, so by judging whether there is high-level data at the relay receiving end, it is judged whether there is bus data.Figure 3 As shown, the transmission of the half-duplex network relay port 2 comes from the reception of relay port 1, and the transmission of relay port 1 also comes from the reception of relay port 2.

[0026] 3) Orchestration of electro-optical conversion messages

[0027] Fiber optic relay transmission uses 8b10b encoding with a rate of 8 Gbps. After conversion processing, when the parallel-to-serial conversion width is 32 bits, the clock is 200 MHz. Take 8 bits out of 32 bits for k-code, frame sequence, and CRC check. The remaining 24 bits can transmit 3 channels of half-duplex data. Each frame is 8 32-bit long. Figure 4 Shown is a frame structure designed in the above manner.

[0028] 4) Redundant reception processing

[0029] The redundant reception design realizes real-time redundant reception of two channels and uses frame sequence numbers to perform redundant frame elimination processing. The basic processing method of redundant reception is as follows: when two frames with the same sequence number arrive, the frame that arrives first is adopted, and frames with CRC errors are discarded. Ensure that during the operation of the bus range extender device, any problems with one of the fiber optic lines will not affect the system's function, effectively increasing the system's reliability. The redundant reception flow chart is as shown in Figure 5.

[0030] First, perform CRC check on the data of channel 1 and channel 2, and directly discard frames with check errors. The polling method is used to process the write buffer requests of the two channels. If there is an unfinished frame in the buffer, and there is no frame sequence number of the data to be written in the buffer and it can continue the frame sequence number of the previous frame, then write the frame data into the buffer and update the frame sequence number identifier. If there is no unfinished frame in the transmit buffer, then while writing the data frame into the buffer, notify the output buffer end to start the mechanism of storing half-frame data and then outputting the data.

[0031] When there is data in both channels simultaneously, the output buffer retains half-frame data to handle the difference in transmission time between the two channels. If there is an error or a line interruption in the data frame of the channel with a faster transmission time, it can be left for the channel with a slower transmission time to perform data handover operations. The time elapsed by the two channels in the hardware circuit board and logical processing is infinitely close to being equal, and the time difference is mainly reflected in the external fiber optic length. Therefore, the fiber optic length difference between the two channels should meet the requirements of half a frame. The half-frame time is the time of 4 200M clocks, that is, 20 ns, and the fiber optic length difference between the two channels should not be less than 4 meters. If we want to increase the compatibility with the fiber optic length difference, we need to increase the data length retained in the output buffer, which will result in an increase in the communication delay time.

[0032] The relay design of the half-duplex network of the present invention has been introduced in detail above. Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or FPGA logic products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a relay scheme implemented on one or more FPGAs containing hardware design languages. Those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these changes and modifications.

Claims

1. A long-distance ultra-low latency relay method for a half-duplex network, characterized in that: The method comprises: The relay node uses dual-channel hot redundant optical fiber as the signal transmission medium; Perform multi-phase overclocking sampling on the received half-duplex network data, combine the sampled data with k code, frame sequence, and CRC check to arrange them into parallel data, perform parallel-to-serial conversion on the parallel data according to the 8b10b encoding format, and send the data signal after parallel-to-serial conversion through the dual-channel hot redundant optical fiber; The receiving end decodes the received optical signal, converts the serial to parallel signal, and obtains the required data after redundant reception processing of the obtained parallel data; The redundant receiving process specifically includes: when two identical frame numbers arrive in dual channels, the first arriving frame is adopted, and the frame with CRC error is discarded; the write cache request of the two channels is processed by polling method, if there are frames that have not been sent completely in the cache, and there is no frame number of data to be written in the cache and the frame number can be continued from the previous frame, the frame data is written into the cache and the frame number identifier is updated; if there are no frames that have not been sent completely in the cache, the data frame is written into the cache while the output cache is notified to start the mechanism of storing half a frame of data before outputting the data; when the two channels have data at the same time, the output cache retains half a frame of data to process the difference in transmission time of the two channels, if the data frame of the channel with faster transmission time has an error or the line is interrupted, the channel with slower transmission time performs data handover operation.

2. The method according to claim 1, characterized in that The method further includes: the relay node determines whether there is data to be sent, thereby controlling whether the transmission enable and reception enable are switched on or off.

3. The method according to claim 2, characterized in that The switch control method of the transmission enable and reception enable is specifically as follows: The relay node determines whether there is data to be sent by the half-duplex network data obtained from its receiving end. When there is data to be sent, it turns on the sending enable and turns off the receiving enable to start sending. When there is no data to be sent, it turns off the sending enable and turns on the receiving enable.

4. The method according to claim 1, characterized in that The parallel data is 32-bit parallel data, and its specific structure is: 8 bits are used for k code, frame sequence, and CRC check, and the remaining 24 bits are used to transmit 3-way half-duplex data.

Images