A high-bandwidth fixed-frame transmission system and method based on pipeline parallel marking
By adopting flow parallel marking technology and cache space module in the HDLC transmission system, the problems of high delay and low bandwidth in high-speed data transmission are solved, and high bandwidth and high precision data transmission is realized, simplifying the structure and reducing line delay .
Patent Information
- Application Number
- CN202211733598.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-12-30
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2042-12-30
AI Technical Summary
The existing HDLC transmission systems have problems in high-speed data transmission that chip performance is difficult to achieve when high latency, low bandwidth and high frequency.
Adopt high-bandwidth fixed-frame transmission system and method based on flow parallel marking, including head and tail correlation module, flow parallel marking module, cache space module and multi-frame separation and reorganization module. HDLC disframing and data reorganization is carried out through flow parallel processing technology, and the process is used to use The cache space absorbs ultra-short frame bursts, achieving high bandwidth and high precision data transmission.
It realizes the rapid transmission of one frame of data with a simple structure, small line delay, and a minimum fixed frame recognition of 1 byte length, solving the problem of difficult chip performance at high frequencies, and has the advantages of high bandwidth and high precision.
Smart Images

Figure CN116318539B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of digital communication, and in particular to a high-bandwidth fixed-frame transmission system and method based on pipeline parallel marking. Background Art
[0002] In the HDLC transmission system, the receiver generally has to perform HDLC framing and deframing operations. At the same time, in modern information transmission systems, the demand for high-speed data transmission is increasing, so high-bandwidth processing of HDLC deframing is involved. Currently, many technologies directly perform frame judgment on the input parallel data and generate judgment instructions. However, as the bit width of the input data increases (N bytes), each clock cycle needs to judge many situations (N 2 ), this method often brings a large judgment delay; at the same time, some devices increase the overall bandwidth by increasing the frequency of the system clock, but in high-speed data transmission systems, the clock frequency required is often greater than 1Ghz, which places extremely high requirements on chip performance and is difficult to achieve. Summary of the invention
[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a high-bandwidth fixed-frame transmission system and method based on pipeline parallel marking, which solves the problem that the chip performance is difficult to achieve at high latency, low bandwidth and high frequency in the prior art.
[0004] In order to achieve the above-mentioned purpose of the invention, the technical solution adopted by the present invention is: a high-bandwidth fixed-frame transmission system based on pipeline parallel marking, the system includes a head-tail association module, a pipeline parallel marking module, a cache space module and a multi-frame separation and reorganization module, the head-tail association module is connected to the pipeline parallel marking module, the pipeline parallel marking module is connected to the cache space module, the cache space module is connected to the multi-frame separation and reorganization module, the head-tail association module is used to associate the currently input parallel data with the data input before and after it, and expand the head and tail of the currently input data to facilitate the subsequent framing judgment, the pipeline parallel marking module is used to obtain data valid signals and marking indication signals, the cache space module is used to absorb ultra-short frame bursts to ensure that data is not lost, and the multi-frame separation and reorganization module is used to perform multi-frame separation and reorganization operations on valid data.
[0005] The beneficial effect of the above scheme is that the parallel pipeline structure is adopted, so that the scheme has the advantages of simple structure and small line delay.
[0006] In addition, the present invention also adopts a technical solution: a high-bandwidth fixed-frame transmission method based on pipeline parallel marking, the method comprising the following steps:
[0007] S1: Use the head-tail association technology to associate the currently input N-byte parallel data with the data input before and after it, and expand the head and tail of the currently input data;
[0008] S2: Using the pipeline parallel marking technology, according to the current input data after the head and tail extension, the Nth byte of the parallel N-byte data is processed at the first stage, and the HDLC de-framing operation is performed on it to generate the first-stage data valid signal and mark indication signal;
[0009] S3: using pipeline parallel marking technology, processing the N-1th byte of the parallel N-byte data at the second stage, performing HDLC deframing operation on it, and generating a second-stage data valid signal and a mark indication signal according to the data valid signal and the mark indication signal generated at the first stage;
[0010] S4: Using the same method as S2 and S3, after N stages of pipeline marking, a data valid signal and a marking indication signal of the Nth stage are generated;
[0011] S5: Utilizing a cache module, the data valid signal and the mark indication signal generated by the pipeline parallel marking are cached to absorb short frame bursts;
[0012] S6: Utilize multi-frame separation and reassembly technology to perform multi-frame separation and reassembly operations on valid data according to cached data valid signals and mark indication signals, and output parallel N-byte data in frames to achieve high-bandwidth fixed-frame transmission of data based on pipeline parallel marking.
[0013] The beneficial effects of the above scheme are: through the above technical scheme, pipeline parallel processing is adopted, which has the advantages of high bandwidth and high precision, and the minimum frame can identify a frame of data with a length of 1 byte, and data can be transmitted accurately and quickly, while solving the problem of difficult to achieve chip performance at high frequency.
[0014] Furthermore, S1 utilizes the head-tail association technique to shift the input N-byte parallel data to the left and store it in the N*3-byte shift register A, then takes out the parallel data B associated with the head and tail of the current data from the shift register A, and stores the parallel data B to obtain B1.
[0015] The beneficial effect of the above further solution is: through the above technical solution, the input data is associated and expanded using the head-tail association technology for use in subsequent operations.
[0016] Furthermore, S2 includes the following sub-steps:
[0017] S2-1: Using the pipeline parallel marking technology, HDLC de-frame the Nth byte of the current parallel data B1 to generate a data valid signal C1 and a valid byte number mark signal Hmod1. The specific method for generating the data valid signal C1 and the valid byte number mark signal Hmod1 is as follows:
[0018] (1) If the current byte is 0x7e or 0x7d, it is invalid data, and C1=0 and Hmod1=0 are generated;
[0019] (2) If the byte before the current byte is 0x7d, store the current byte D1 in C1 after XORing it with 0x20, and generate C1=D1^0x20, Hmod1=1, where ^ is the XOR operator;
[0020] (3) If the current byte is any other case, put the current byte D1 into the lower 8 bits of C1, generating C1=D1, Hmod1=1;
[0021] S2-2: Using the pipeline parallel marking technology, the frame start judgment is performed on the Nth byte of the current parallel data B1, and a frame interval marking signal F1 is generated. The frame interval marking is performed in groups of 2 bits. The specific method for generating the frame interval marking signal F1 is as follows:
[0022] (1) If the current valid byte is at the start of a frame but not at the end of a frame, F1 = 0x2 is generated;
[0023] (2) If the current valid byte is at the end of the frame but not the start of the frame, F1 = 0x1 is generated;
[0024] (3) If the current valid byte is at the end of the frame and also at the start of the frame, F1 = 0x3 is generated;
[0025] (4) If the current valid byte is neither the start of a frame nor the end of a frame, F1 = 0x0 is generated;
[0026] S2-3: Use pipeline parallel marking technology to count the frame byte of the Nth byte of the current parallel data B1, generate a frame byte count mark signal Scnt1, and perform frame count marking with log2N+1 bits as a group. The method for generating the frame byte count mark signal Scnt1 is specifically as follows:
[0027] (1) If the current byte is valid data, Scnt1 = 0x1 is generated;
[0028] (2) If the current byte is not valid data, Scnt1 = 0x0 is generated.
[0029] The beneficial effect of the above further solution is: through the above technical solution, the pipeline parallel marking technology is used to generate the first-level data valid signal C1 and the mark indication signals Hmod1, F1 and Scnt1.
[0030] Furthermore, S3 includes the following sub-steps:
[0031] S3-1: parallel data B1 is stored to obtain parallel data B2, and HDLC de-framing is performed on the N-1th byte of the current parallel data B2 using pipeline parallel marking technology to generate a data valid signal C2 and a valid byte number marking signal Hmod2. The specific method for generating the data valid signal C2 and the valid byte number marking signal Hmod2 is as follows:
[0032] (1) If the current byte is 0x7e or 0x7d, it is invalid data, and C2=C1 and Hmod2=Hmod1 are generated;
[0033] (2) If the byte before the current byte is 0x7d, the current byte D2 is XORed with 0x20 and then left-shifted to C1 to obtain C2, resulting in C2=C1<<8+D2^0x20, Hmod2=Hmod1+1, where << is a left shift operator;
[0034] (3) If the current byte is any other case, the current byte D2 is directly shifted left to C1 to obtain C2, resulting in C2 = C1 << 8 + D2, Hmod2 = Hmod1 + 1;
[0035] S3-2: Using the pipeline parallel marking technology, the frame start judgment is performed on the N-1th byte of the current parallel data B2, and a frame interval marking signal F2 is generated. The frame interval marking is performed in groups of 2 bits. The specific method for generating the frame interval marking signal F2 is as follows:
[0036] (1) If the current valid byte is at the start of a frame but not at the end of a frame, F2 = F1 << 2 + 0x2;
[0037] (2) If the current valid byte is at the end of the frame but not the start of the frame, F2 = F1 << 2 + 0x1;
[0038] (3) If the current valid byte is at the end of the frame and also at the start of the frame, F2 = F1 << 2 + 0x3;
[0039] (4) If the current valid byte is neither the start of a frame nor the end of a frame, F2 = F1 is generated;
[0040] S3-3: Using pipeline parallel marking technology, the frame byte count is performed on the N-1th byte of the current parallel data B2 to generate a frame byte count mark signal Scnt2. The frame count mark is performed with log2N+1 bits as a group. The method for generating the frame byte count mark signal Scnt2 is specifically as follows:
[0041] (1) If the current byte is valid data and the current data is the start of a frame, Scnt2 = Scnt1 < <log2N+1+1;
[0042] (2) If the current byte is not valid data and the current data is not the start of a frame, Scnt2=Scnt1+1 is generated.
[0043] The beneficial effect of the above further scheme is: through the above technical scheme, using the pipeline parallel marking technology, according to the valid signal C1 and mark indication signals Hmod1, F1 and Scnt1 obtained in the first stage, the valid signal C2 and mark indication signals Hmod2, F2 and Scnt2 of the second stage are exited.
[0044] Further, in S4, the data valid signal C is obtained after N-2 processings are performed in sequence according to the same method as steps S3-1, S3-2 and S3-3. N And mark indication signal Hmod N Scnt N and F N .
[0045] The beneficial effect of the above further solution is: through the above technical solution, by analogy, after N-2 times of pipeline processing, the data valid signal C is obtained. N And mark indication signal Hmod N Scnt N and F N .
[0046] Furthermore, S6 includes the following sub-steps:
[0047] S6-1: According to Hmod N The number of invalid bytes marked in C N Extract valid data from
[0048] S6-2: Using Scnt N Take log2N+1 bits as a group, judge the value of each log2N+1 bit from high to low, and separate the valid data into frames according to the value;
[0049] S6-3: F N Every 2 bits and Scnt N Each log2N+1 bit in the table has a one-to-one correspondence, using FN After the frames are separated, the data in the same frame are spliced again to form parallel data with a bit width of N bytes, and frame start and frame end flags are generated at the same time.
[0050] The beneficial effect of the above further scheme is: through the above technical scheme, using the multi-frame separation and recombination technology, the multi-frame separation operation of the data is realized, and the data is output in units of frames. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] Figure 1 This is a structural diagram of a high-bandwidth fixed-frame transmission system based on pipeline parallel marking.
[0052] Figure 2 The flowchart is a high-bandwidth frame transmission method based on pipeline parallel marking. DETAILED DESCRIPTION
[0053] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.
[0054] Embodiment 1, as Figure 1 As shown, a high-bandwidth fixed-frame transmission system based on pipeline parallel marking, the system includes a head-tail association module, a pipeline parallel marking module, a cache space module and a multi-frame separation and reorganization module, the head-tail association module is connected to the pipeline parallel marking module, the pipeline parallel marking module is connected to the cache space module, the cache space module is connected to the multi-frame separation and reorganization module, the head-tail association module is used to associate the currently input parallel data with the data input before and after it, and expand the head and tail of the currently input data to facilitate the subsequent framing judgment, the pipeline parallel marking module is used to obtain data valid signals and marking indication signals, the cache space module is used to absorb ultra-short frame bursts to ensure that data is not lost, and the multi-frame separation and reorganization module is used to perform multi-frame separation and reorganization operations on valid data.
[0055] Embodiment 2, as Figure 2 As shown, a high-bandwidth fixed-frame transmission method based on pipeline parallel marking, the transmission method comprises the following steps:
[0056] S1: Use the head-tail association technology to associate the currently input N-byte parallel data with the data input before and after it, and expand the head and tail of the currently input data;
[0057] S2: Using the pipeline parallel marking technology, according to the current input data after the head and tail extension, the Nth byte of the parallel N-byte data is processed at the first stage, and the HDLC de-framing operation is performed on it to generate the first-stage data valid signal and mark indication signal;
[0058] S3: using pipeline parallel marking technology, processing the N-1th byte of the parallel N-byte data at the second stage, performing HDLC deframing operation on it, and generating a second-stage data valid signal and a mark indication signal according to the data valid signal and the mark indication signal generated at the first stage;
[0059] S4: Using the same method as S2 and S3, after N stages of pipeline marking, a data valid signal and a marking indication signal of the Nth stage are generated;
[0060] S5: Utilizing a cache module, the data valid signal and the mark indication signal generated by the pipeline parallel marking are cached to absorb short frame bursts;
[0061] S6: Utilize multi-frame separation and reassembly technology to perform multi-frame separation and reassembly operations on valid data according to cached data valid signals and mark indication signals, and output parallel N-byte data in frames to achieve high-bandwidth fixed-frame transmission of data based on pipeline parallel marking.
[0062] In S1, the head-tail association technology is used to shift the input N-byte parallel data to the left and store it in the N*3-byte shift register A, and then the parallel data B associated with the head and tail of the current data is taken out from the shift register A, and the parallel data B is stored to obtain B1.
[0063] S2 includes the following sub-steps:
[0064] S2-1: Using the pipeline parallel marking technology, HDLC de-frame the Nth byte of the current parallel data B1 to generate a data valid signal C1 and a valid byte number mark signal Hmod1. The specific method for generating the data valid signal C1 and the valid byte number mark signal Hmod1 is as follows:
[0065] (1) If the current byte is 0x7e or 0x7d, it is invalid data, and C1=0 and Hmod1=0 are generated;
[0066] (2) If the byte before the current byte is 0x7d, store the current byte D1 in C1 after XORing it with 0x20, and generate C1=D1^0x20, Hmod1=1, where ^ is the XOR operator;
[0067] (3) If the current byte is any other case, put the current byte D1 into the lower 8 bits of C1, generating C1=D1, Hmod1=1;
[0068] S2-2: Using the pipeline parallel marking technology, the frame start judgment is performed on the Nth byte of the current parallel data B1, and a frame interval marking signal F1 is generated. The frame interval marking is performed in groups of 2 bits. The specific method for generating the frame interval marking signal F1 is as follows:
[0069] (1) If the current valid byte is at the start of a frame but not at the end of a frame, F1 = 0x2 is generated;
[0070] (2) If the current valid byte is at the end of the frame but not the start of the frame, F1 = 0x1 is generated;
[0071] (3) If the current valid byte is at the end of the frame and also at the start of the frame, F1 = 0x3 is generated;
[0072] (4) If the current valid byte is neither the start of a frame nor the end of a frame, F1 = 0x0 is generated;
[0073] S2-3: Use pipeline parallel marking technology to count the frame byte of the Nth byte of the current parallel data B1, generate a frame byte count mark signal Scnt1, and perform frame count marking with log2N+1 bits as a group. The method for generating the frame byte count mark signal Scnt1 is specifically as follows:
[0074] (1) If the current byte is valid data, Scnt1 = 0x1 is generated;
[0075] (2) If the current byte is not valid data, Scnt1 = 0x0 is generated.
[0076] S3 includes the following steps:
[0077] S3-1: parallel data B1 is stored to obtain parallel data B2, and HDLC de-framing is performed on the N-1th byte of the current parallel data B2 using pipeline parallel marking technology to generate a data valid signal C2 and a valid byte number marking signal Hmod2. The specific method for generating the data valid signal C2 and the valid byte number marking signal Hmod2 is as follows:
[0078] (1) If the current byte is 0x7e or 0x7d, it is invalid data, and C2=C1 and Hmod2=Hmod1 are generated;
[0079] (2) If the byte before the current byte is 0x7d, the current byte D2 is XORed with 0x20 and then left-shifted to C1 to obtain C2, resulting in C2=C1<<8+D2^0x20, Hmod2=Hmod1+1, where << is a left shift operator;
[0080] (3) If the current byte is any other case, the current byte D2 is directly shifted left to C1 to obtain C2, resulting in C2 = C1 << 8 + D2, Hmod2 = Hmod1 + 1;
[0081] S3-2: Using the pipeline parallel marking technology, the frame start judgment is performed on the N-1th byte of the current parallel data B2, and a frame interval marking signal F2 is generated. The frame interval marking is performed in groups of 2 bits. The specific method for generating the frame interval marking signal F2 is as follows:
[0082] (1) If the current valid byte is at the start of a frame but not at the end of a frame, F2 = F1 << 2 + 0x2;
[0083] (2) If the current valid byte is at the end of the frame but not the start of the frame, F2 = F1 << 2 + 0x1;
[0084] (3) If the current valid byte is at the end of the frame and also at the start of the frame, F2 = F1 << 2 + 0x3;
[0085] (4) If the current valid byte is neither the start of a frame nor the end of a frame, F2 = F1 is generated;
[0086] S3-3: Using pipeline parallel marking technology, the frame byte count is performed on the N-1th byte of the current parallel data B2 to generate a frame byte count mark signal Scnt2. The frame count mark is performed with log2N+1 bits as a group. The method for generating the frame byte count mark signal Scnt2 is specifically as follows:
[0087] (1) If the current byte is valid data and the current data is the start of a frame, Scnt2 = Scnt1 < <log2N+1+1;
[0088] (2) If the current byte is not valid data and the current data is not the start of a frame, Scnt2=Scnt1+1 is generated.
[0089] In S4, the same method as steps S3-1, S3-2 and S3-3 is used to perform N-2 processing in sequence, and the data valid signal C is obtained. N And mark indication signal Hmod N Scnt N and F N .
[0090] S6 includes the following sub-steps:
[0091] S6-1: According to Hmod N The number of invalid bytes marked in C N Extract valid data from
[0092] S6-2: Using Scnt N Take log2N+1 bits as a group, judge the value of each log2N+1 bit from high to low, and separate the valid data into frames according to the value;
[0093] S6-3: F N Every 2 bits and Scnt N Each log2N+1 bit in the same order corresponds to each other, using F N After the frames are separated, the data in the same frame are spliced again to form parallel data with a bit width of N bytes, and frame start and frame end flags are generated at the same time.
[0094] This solution adopts pipeline parallel marking technology to improve the bandwidth of frame decomposition, and needs to adopt parallel processing structure to transmit multiple bytes of data simultaneously in one clock cycle. In order to reduce wiring delay and improve system processing bandwidth, the present invention adopts pipeline parallel marking technology to process data with pipeline parallel structure and mark the processed data at the same time so that the subsequent stage can reassemble the data frame. A cache space is added before the multi-frame separation and reorganization module to store the valid signal C output by the pipeline parallel marking module. N And mark indication signal Hmod N Scnt N and F N Caching can effectively absorb ultra-short frame bursts and enhance system reliability. For example, when the parallel N-byte data contains multiple ultra-short frames, the short frame burst is large. The post-stage multi-frame separation and reassembly module transmits data in frames, and it takes multiple clock cycles to process the current N-byte data. Increasing the cache space can cache the data valid signal and marking information output by the pipeline parallel marking module during the waiting cycle, effectively absorbing the short frame burst.
[0095] In one embodiment of the present invention, the head-tail association technology is used to associate the currently input parallel data (N bytes) with the data input before and after it, and the head and tail of the currently input data are extended to facilitate the subsequent frame determination; the pipeline parallel marking technology is used to perform N levels of pipeline parallel marking within N processing clock cycles, remove the inter-frame filler, and perform HDLC deframing operations at the same time, and finally obtain valid data and marking indications; the cache space is used to cache the valid data and marking indications, and the cache space can effectively absorb ultra-short frame bursts to ensure that data is not lost; the multi-frame separation and reorganization technology is used to perform multi-frame separation and reorganization operations on the valid data according to the marking indication read out from the cache space, and the frame data is output in units of frames. At the same time, according to the number of frames in the currently input parallel data, the rate of reading data from the cache space is controlled.
[0096] The present invention adopts pipeline parallel processing, has no bandwidth waste, has the advantages of high bandwidth and high precision, and can identify a frame of data with a minimum length of 1 byte; has strong compatibility, and can be compatible with the framing processing of data with different parallel bit widths by only increasing / decreasing the number of processing clock cycles; adopts a parallel pipeline structure, has the advantages of simple structure and small line delay; has a cache space inside, and can absorb ultra-short frame bursts in parallel data.
[0097] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the invention.
Claims
1. A high-bandwidth fixed-frame transmission method based on pipeline parallel marking, characterized in that: The method comprises the following steps: S1: Use the head-tail association technology to associate the currently input N-byte parallel data with the data input before and after it, and expand the head and tail of the currently input data; S2: Using the pipeline parallel marking technology, according to the current input data after the head and tail extension, the Nth byte of the parallel N-byte data is processed at the first stage, and the HDLC de-framing operation is performed on it to generate the first-stage data valid signal and mark indication signal; S3: using pipeline parallel marking technology, processing the N-1th byte of the parallel N-byte data at the second stage, performing HDLC deframing operation on it, and generating a second-stage data valid signal and a mark indication signal according to the data valid signal and the mark indication signal generated at the first stage; S4: Using the same method as S2 and S3, after N stages of pipeline marking, a data valid signal and a marking indication signal of the Nth stage are generated; S5: Utilizing a cache module, caches the data valid signal and the mark indication signal generated by pipeline parallel marking, so as to absorb the ultra-short frame burst; S6: Utilize multi-frame separation and reassembly technology to perform multi-frame separation and reassembly operations on valid data according to cached data valid signals and mark indication signals, and output parallel N-byte data in frames to achieve high-bandwidth fixed-frame transmission of data based on pipeline parallel marking.
2. The high-bandwidth fixed-frame transmission method based on pipeline parallel marking according to claim 1 is characterized in that: In S1, the head-tail association technology is used to shift the input N-byte parallel data to the left and store it in the N*3-byte shift register A, and then the parallel data B associated with the head and tail of the current data is taken out from the shift register A, and the parallel data B is stored to obtain B1.
3. The high-bandwidth fixed-frame transmission method based on pipeline parallel marking according to claim 2 is characterized in that: The S2 includes the following sub-steps: S2-1: Using the pipeline parallel marking technology, HDLC de-frame the Nth byte of the current parallel data B1 to generate a data valid signal C1 and a valid byte number mark signal Hmod1. The specific method for generating the data valid signal C1 and the valid byte number mark signal Hmod1 is as follows: (1) If the current byte is 0x7e or 0x7d, it is invalid data, and C1=0 and Hmod1=0 are generated; (2) If the byte before the current byte is 0x7d, store the current byte D1 in C1 after XORing it with 0x20, and generate C1=D1^0x20, Hmod1=1, where ^ is the XOR operator; (3) If the current byte is any other case, put the current byte D1 into the lower 8 bits of C1, generating C1=D1, Hmod1=1; S2-2: Using the pipeline parallel marking technology, the frame start judgment is performed on the Nth byte of the current parallel data B1, and a frame interval marking signal F1 is generated. The frame interval marking is performed in groups of 2 bits. The specific method for generating the frame interval marking signal F1 is as follows: (1) If the current valid byte is at the start of a frame but not at the end of a frame, F1 = 0x2 is generated; (2) If the current valid byte is at the end of the frame but not the start of the frame, F1 = 0x1 is generated; (3) If the current valid byte is at the end of the frame and also at the start of the frame, F1 = 0x3 is generated; (4) If the current valid byte is neither the start of a frame nor the end of a frame, F1 = 0x0 is generated; S2-3: Use pipeline parallel marking technology to count the frame byte of the Nth byte of the current parallel data B1, generate a frame byte count mark signal Scnt1, and perform frame count marking with log2N+1 bits as a group. The method for generating the frame byte count mark signal Scnt1 is specifically as follows: (1) If the current byte is valid data, Scnt1 = 0x1 is generated; (2) If the current byte is not valid data, Scnt1 = 0x0 is generated.
4. The high-bandwidth fixed-frame transmission method based on pipeline parallel marking according to claim 3 is characterized in that: The S3 includes the following sub-steps: S3-1: parallel data B1 is stored to obtain parallel data B2, and HDLC de-framing is performed on the N-1th byte of the current parallel data B2 using pipeline parallel marking technology to generate a data valid signal C2 and a valid byte number marking signal Hmod2. The specific method for generating the data valid signal C2 and the valid byte number marking signal Hmod2 is as follows: (1) If the current byte is 0x7e or 0x7d, it is invalid data, and C2=C1 and Hmod2=Hmod1 are generated; (2) If the byte before the current byte is 0x7d, the current byte D2 is XORed with 0x20 and then left-shifted to C1 to obtain C2, resulting in C2=C1<<8+D2^0x20, Hmod2=Hmod1+1, where << is a left shift operator; (3) If the current byte is any other case, the current byte D2 is directly shifted left to C1 to obtain C2, resulting in C2 = C1 << 8 + D2, Hmod2 = Hmod1 + 1; S3-2: Using the pipeline parallel marking technology, the frame start judgment is performed on the N-1th byte of the current parallel data B2, and a frame interval marking signal F2 is generated. The frame interval marking is performed in groups of 2 bits. The specific method for generating the frame interval marking signal F2 is as follows: (1) If the current valid byte is at the start of a frame but not at the end of a frame, F2 = F1 << 2 + 0x2; (2) If the current valid byte is at the end of the frame but not the start of the frame, F2 = F1 << 2 + 0x1; (3) If the current valid byte is at the end of the frame and also at the start of the frame, F2 = F1 << 2 + 0x3; (4) If the current valid byte is neither the start of a frame nor the end of a frame, F2 = F1 is generated; S3-3: Using pipeline parallel marking technology, the frame byte count is performed on the N-1th byte of the current parallel data B2 to generate a frame byte count mark signal Scnt2. The frame count mark is performed with log2N+1 bits as a group. The method for generating the frame byte count mark signal Scnt2 is specifically as follows: (1) If the current byte is valid data and the current data is the start of a frame, Scnt2 = Scnt1 < <log2N+1+1; (2) If the current byte is not valid data and the current data is not the start of a frame, Scnt2=Scnt1+1 is generated.
5. The high-bandwidth fixed-frame transmission method based on pipeline parallel marking according to claim 4 is characterized in that: In the step S4, the data valid signal C is obtained after N-2 processings are performed in sequence according to the same method as steps S3-1, S3-2 and S3-3. N And mark indication signal Hmod N Scnt N and F N .
6. The high-bandwidth fixed-frame transmission method based on pipeline parallel marking according to claim 5 is characterized in that: The S6 comprises the following sub-steps: S6-1: According to Hmod N The number of invalid bytes marked in C N Extract valid data from S6-2: Using Scnt N Take log2N+1 bits as a group, judge the value of each log2N+1 bit from high to low, and separate the valid data into frames according to the value; S6-3: F N Every 2 bits and Scnt N Each log2N+1 bit in the table has a one-to-one correspondence, using F N After the frames are separated, the data in the same frame are spliced again to form parallel data with a bit width of N bytes, and frame start and frame end flags are generated at the same time.
7. A transmission system for a high-bandwidth fixed-frame transmission method based on pipeline parallel marking according to any one of claims 1 to 6, characterized in that: The transmission system includes a head-tail association module, a pipeline parallel marking module, a cache space module and a multi-frame separation and reorganization module. The head-tail association module is connected to the pipeline parallel marking module, the pipeline parallel marking module is connected to the cache space module, and the cache space module is connected to the multi-frame separation and reorganization module. The head-tail association module is used to associate the currently input parallel data with the data input before and after it, and to expand the head and tail of the currently input data to facilitate the subsequent framing judgment. The pipeline parallel marking module is used to obtain data valid signals and marking indication signals. The cache space module is used to absorb ultra-short frame bursts to ensure that data is not lost. The multi-frame separation and reorganization module is used to perform multi-frame separation and reorganization operations on valid data.
