WO2017012517A1 - Hybrid physical coding sub-layer and method for transmitting and receiving data, and storage medium - Google Patents

Abstract

Disclosed are a hybrid physical coding sub-layer and a method for transmitting and receiving data. The hybrid physical coding sub-layer comprises a sending module and a receiving module; the sending module comprises four basic physical coding sub-layer (PCS) sending units, a code block replacing and selecting unit and an unbalanced value calculation processing unit; the receiving module comprises an aligned code detection processing unit, a code block restoring and selecting unit, four basic PCS receiving units and an unbalanced value calculation processing unit. The code block replacing and selecting unit, the unbalanced value calculation processing unit, the aligned code detection processing unit and the code block restoring and selecting unit combine a corresponding four-channel unit and a single-channel unit, and different units are selected for corresponding processing.

Description

Mixed physical coding sublayer and data transmission and reception method, storage medium Technical field

The present invention relates to an electronic communication Ethernet technology, and in particular, to a hybrid coding sublayer, a data transmission and reception method, and a storage medium.

Background technique

In technologies that allow computers and other network devices to form local area networks, Ethernet has become a major networking technology and has been standardized in the IEEE 802.3 family of standards. Ethernet standards evolve over time, allowing different variants of existing Ethernet protocols to support higher bandwidth, improved media access control, different physical media channels, or other functions. For example, IEEE 802.3 now has processing standards covering speeds from 10 Mbit/s, 100 Mbit/s, 1 Gbit/s to 10 Gbit/s and beyond, and has managed physical channels (such as coaxial cable, fiber optic and unshielded). / Shielded twisted pair cable) deformation.

At present, the development of Gigabit Ethernet interface technology is relatively mature. Gigabit Ethernet is characterized by high efficiency, high speed and high performance, and has been widely used in industries such as finance, commerce, education, and government agencies. In order to meet the requirements of high-speed data transmission for Ethernet communication, and at the same time, it has the advantages of simple wiring and good signal integrity in circuit design. Cisco (CISCO) has proposed serial Gigabit Media Independent Interface (SGMII, Serial Gigabit Media Independent). The interface specification of the Interface and the Quad Serial Gigabit Media Independent Interface (QSGMII) is actually the physical coding of the original fiber-optic Gigabit Ethernet (1000BASE-X) PCS (Physical Coding Sublayer). An extension of the sublayer specification. At the same time, SGMII/QSGMII has also been widely used in Gigabit Ethernet communication transmission. Therefore, in the current gigabit interface physical coding sublayer design, 1000BASE-X/SGMII/QSGMII needs to be externally supported to meet the application needs of different scenarios and with different manufacturers. The interface is docked. Therefore, the low-cost Ethernet interface design that can support multiple Gigabit port morphological modes is a technical problem that needs to be solved to construct a high-compatibility Ethernet communication system.

Summary of the invention

To solve the above technical problem, an embodiment of the present invention provides a mixture coding sublayer, a data transmission and reception method thereof, and a storage medium.

The hybrid coding sublayer provided by the embodiment of the present invention includes: a sending module, where the sending module includes:

The basic transmission unit of the four physical coding sublayers (PCS) is configured to transmit four sets of Gigabit Media Independent Interfaces (GMII, Gigabit Media Independent Interface) to the four medium access control (MAC, Media Access Control) layers. The issued data is simultaneously ordered to replace the insertion and encoding processing;

The code group replacement selection unit is configured to replace the K28.5 code group of the first channel with the K28.1 code group when the mixed coding sublayer works in the 4-channel scene after the encoding process; When working in a single channel scenario, no replacement processing is performed;

The unbalanced value calculation processing unit is configured to calculate an unbalanced value required for the encoding process by using the 4-channel unbalanced value when the mixed-coded sub-layer operates in the 4-channel scene; configured to operate the single-channel when the mixed-coded sub-layer operates In the scene, the unbalanced value required for the encoding process is calculated using the single channel unbalanced value.

In the embodiment of the present invention, each basic sending unit of the PCS includes:

The standard transmission state machine module is configured to perform an ordered set replacement insertion process on the MAC layer transmitting side of one channel by the GMII, and encapsulate the data into a code group format suitable for 8B/10B encoding;

The 8B/10B encoding module is configured to perform 8B/10B encoding processing on the code stream processed by the standard transmitting state machine module, wherein the 8B/10B encoding process requires an unbalanced value as an input; after the encoding process, the data is sent to the SERDES transmitting side. Do parallel data to serial data conversion;

The auto-negotiation processing module is configured to perform capability coordination operations at both ends of the communication through auto-negotiation.

In the embodiment of the present invention, the code group replacement selecting unit is further configured to: according to the control signal, select whether to replace the K28.1 code group of the first channel with the K28.5 code group;

The unbalanced value calculation processing unit is further configured to: according to the control signal, select an unbalanced value required to calculate the encoding process by using the 4-channel unbalanced value, or calculate an unbalanced value required for the encoding process by using the single-channel unbalanced value. .

A hybrid coding sublayer provided by another embodiment of the present invention includes: a receiving module, where the receiving module includes:

The alignment code detection processing unit is configured to use the 4-channel alignment code to detect data alignment from the SERDES when the mixed-coded sub-layer operates in a 4-channel scene; when the mixture coding sub-layer operates in a single-channel scene, use the single Channel alignment code detection aligns the data sent from the SERDES;

The code group restoration selection unit is configured to restore the K28.1 code group of the first channel to the K28.5 code group when the mixture coding sublayer works in the 4-channel scene after data alignment; when the mixture encodes the sub-layer When working in a single channel scenario, no restoration is performed;

4 PCS basic receiving units, after data alignment, decode 4 channels of data and remove the ordered set code group, and restore the data stream to the GMII data stream and send it to the 4 MAC layer receiving sides;

The unbalanced value calculation processing unit is configured to calculate an unbalanced value required for the decoding process by using the 4-channel unbalanced value when the mixed-coded sub-layer operates in the 4-channel scene; configured to operate the single-channel when the mixed-coded sub-layer operates In the case of the scene, the unbalanced value required for the decoding process is calculated using the single channel unbalanced value.

In the embodiment of the present invention, each PCS basic receiving unit includes:

The 8B/10B decoding module is configured to perform 8B/10B decoding operation on the aligned data;

Standard Receive State Machine module configured to remove ordered sets of 8B/10B decoded data The code group transmits the data stream to the GMII data stream and sends it to the MAC layer receiving side.

In the embodiment of the present invention, the code group restoration selecting unit is further configured to: according to the control signal, select to use 4-channel alignment code detection, or single channel alignment code detection to align data sent from the SERDES;

The unbalanced value calculation processing unit is further configured to: according to the control signal, select an unbalanced value required to calculate a decoding process by using a 4-channel unbalanced value, or calculate an unbalanced value required for the decoding process by using a single-channel unbalanced value; .

A hybrid coding sublayer provided by another embodiment of the present invention includes the foregoing transmission module, and a receiving module.

The data transmission method of the mixture coding sublayer provided by the embodiment of the present invention includes:

Receiving data sent by the MAC layer transmitting side through the GMII;

And performing an ordered set replacement insertion and an encoding process on the data, wherein, when performing the encoding process, when the mixed coding sublayer operates in a 4-channel scene, the unbalanced value required for the encoding process is calculated by using the 4-channel unbalanced value, When the mixture coding sublayer works in a single channel scenario, the unbalanced value required for the encoding process is calculated using the single channel imbalance value;

When the mixture coding sublayer works in a 4-channel scene, the K28.5 code group of the first channel is replaced with the K28.1 code group, and sent to the SERDES transmission side for parallel data to serial data conversion; When the coding sublayer works in a single channel scenario, it does not perform replacement processing, and directly sends data to the SERDES transmitting side to convert parallel data to serial data.

In the embodiment of the present invention, the performing the ordered set replacement insertion and the encoding processing on the data includes:

The data sent by the GMII of the MAC layer of one channel is subjected to the ordered set replacement insertion process, and the data is encapsulated into a code group format suitable for 8B/10B coding;

The processed code stream is subjected to 8B/10B encoding processing, wherein the 8B/10B encoding process requires an unbalanced value as an input.

In the embodiment of the present invention, the method further includes:

The capability coordination operation at both ends of the communication is completed through auto-negotiation.

The data receiving method of the mixture coding sublayer provided by the embodiment of the present invention includes:

When the mixture coding sublayer works in a 4-channel scene, the 4-channel alignment code is used to detect the alignment of the data sent from the SERDES; when the mixture coding sub-layer works in a single-channel scene, the single-channel alignment code detection will be issued from the SERDES. Data alignment;

After the data is aligned, when the mixture coding sublayer works in a 4-channel scene, the K28.1 code group of the first channel is restored to the K28.5 code group; when the mixture coding sub-layer works in a single-channel scene, Reduction process

After the data is aligned, the 4-channel data is decoded and the ordered set code group is removed, and the data stream is restored to the GMII data stream and sent to the four MAC layer receiving sides; wherein, when the mixture coding sub-layer works In the 4-channel scene, the unbalanced value required for the decoding process is calculated by using the 4-channel unbalanced value; when the mixed-coded sub-layer operates in a single-channel scene, the imbalance required for the decoding process is calculated by using the single-channel unbalanced value. value.

In the embodiment of the present invention, after the data is aligned, the 4-channel data is decoded and the ordered set code group is removed, and the data stream is restored to the GMII data stream and sent to the four MAC layer receiving sides, including:

The 8B/10B decoding operation is performed on the aligned data; the ordered set code group is removed from the 8B/10B decoded data, and the data stream is restored to the GMII data stream and sent to the MAC layer receiving side.

A storage medium is provided in the embodiment of the present invention. The storage medium stores a computer program, and the computer program is configured to execute the data transmission method of the hybrid coding sublayer.

A storage medium is provided in the embodiment of the present invention. The storage medium stores a computer program configured to execute the data receiving method of the hybrid coding sublayer.

The technical solution of the embodiment of the present invention provides an almost no increase on the basis of the QSGMII design. Adding any additional resource cost can support multiple interface modes including the Gigabit Ethernet physical coding sublayer of QSGMII/SGMII/1000BASE-X. By carefully analyzing the similarities and differences between the QSGMII and SGMII/1000BASE-X physical coding sublayer processing structures, the original QSGMII reusable logic module can be extracted, and a small amount of SGMII/1000BASE-X logic module resources can be added to obtain multiple thousand simultaneous support. The mega interface mode and the resource cost has almost no added physical coding sublayer. Compared to the direct use of SGMII/1000BASE-X plus QSGMII physical coding sub-layer design can save at least 1/4 of the resource consumption. The hybrid coding sublayer of the embodiment of the present invention includes a sending module, and the sending module includes: four identical PCS basic transmitting units. Each PCS basic sending unit includes a standard transmitting state machine module, an 8B/10B encoding module, and a self-negotiation processing module. The code group replacement selection unit is configured to replace the K28.5 code group of the first channel with the K28.1 code group when the mixed coding sublayer works in the 4-channel scene after the encoding process; When working in a single-channel scene, no replacement processing is performed. The unbalanced value calculation processing unit is configured to calculate an unbalanced value required for the encoding process by using the 4-channel unbalanced value when the mixed-coded sub-layer operates in the 4-channel scene; configured to operate the single-channel when the mixed-coded sub-layer operates In the scene, the unbalanced value required for the encoding process is calculated using the single channel unbalanced value.

The hybrid coding sublayer of the embodiment of the present invention further includes a receiving module, and the receiving module includes: four identical PCS basic receiving units. Each PCS basic receiving unit includes a standard receiving state machine module and an 8B/10B decoding module. The alignment code detection processing unit is configured to use the 4-channel alignment code to detect data alignment from the SERDES when the mixed-coded sub-layer operates in a 4-channel scene; when the mixture coding sub-layer operates in a single-channel scene, use the single Channel alignment code detection aligns the data sent from the SERDES. The code group restoration selection unit is configured to restore the K28.1 code group of the first channel to the K28.5 code group when the mixture coding sublayer works in the 4-channel scene after data alignment; when the mixture encodes the sub-layer When working in a single-channel scene, no restoration is performed. The unbalanced value calculation processing unit is configured to calculate the unevenness required for the decoding process by using the 4-channel unbalanced value when the mixed-coded sub-layer operates in the 4-channel scene The value is configured to calculate the imbalance value required for the decoding process using the single channel imbalance value when the mixture coding sublayer operates in a single channel scenario. In the embodiment of the present invention, a single QSGMII can support an additional multiple transmission modes (SGMII, 1000BASE-X) with little additional resource consumption, and the application cost is low, and the communication device can be flexibly applied to diversification. Scene.

DRAWINGS

1 is a structural diagram of a SGMII/1000BASE-X physical coding sublayer;

2 is a structural diagram of a QSGMII physical coding sublayer;

3 is a schematic structural view of a mixture coding sub-layer of an embodiment of the present invention;

Figure 4 is a schematic diagram of single channel unbalanced value processing;

Figure 5 is a schematic diagram of 4-channel unbalanced value processing;

6 is a flow chart of data transmission of a mixture coding sublayer according to an embodiment of the present invention;

7 is a flow chart of data reception of a mixture coding sublayer of an embodiment of the present invention.

detailed description

The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

As shown in Figure 1, the black code in the figure shows the physical coding sublayer structure of SGMII/1000BASE-X.

The SGMII/1000BASE-X physical coding sublayer transmitting side performs the ordered set replacement insertion and encoding processing on the MAC layer transmitting side by the data sent by the GMII, and the bit width is expanded from 8 bits (bit) to 10 bits after encoding. The physical coding sublayer outputs a 10-bit data stream at 125Mhz to the SERializer serializer/DESerializer deserializer (SERDES) transmit side 1, and the SERDES transmit side 1 performs parallel-to-serial conversion on the parallel data into a serial bit stream output.

At this time, the actual rate is 10bit×125Mhz=1.25Gbps, and the effective rate is 1.25Gpbs*8/10=1Gbps for removing 8B/10B encoding. For the physical coding sublayer of the SGMII/1000BASE-X alone, the corresponding SERDES transmission side 1 and the SERDES reception side 1 need only be fixed in a processing mode of 10 bit width. Note that the bit width of the SERDES transmitting side 1 and the SERDES receiving side 1 is actually 40 bits, but the fixed operation is only in the mode where only the lower 10 bits are valid.

The receiving side of the SGMII/1000BASE-X physical coding sublayer receives 10 bits of parallel data from the SERDES receiving side 1 after serial-to-parallel conversion recovery, and performs re-alignment to perform decoding operation, and removes the ordered set code group to restore the data stream to GMII. The data stream form is sent back to the MAC layer receiving side.

The SGMII/1000BASE-X physical coding sublayer transmission side mainly includes a PCS basic transmission unit and a single channel unbalanced value calculation unit. The PCS basic transmission unit includes a standard transmission state machine module, an 8B/10B coding module, and a self-negotiation processing module.

Standard Transmit State Machine Module: It is mainly based on IEEE 802.3-2008_section3 Figure 36-5, Figure 36-6, 1000BASE-X transmission state machine structure diagram, which realizes the ordered set conversion and insertion and abnormality of Gigabit Ethernet PCS. The encapsulation process, such as a jump of the situation, adds a start code /S/, an end code/T/, an idle code/I/code to the data, and the data stream is encapsulated into a code group form suitable for 8B/10B encoding.

The 8B/10B encoding module performs 8B/10B encoding operation according to the inserted code stream, and the 8B/10B encoding processing requires an unbalanced value as an input, which is provided by the single channel unbalanced value calculating unit. After the encoding is completed, the data can be directly sent to the SERDES transmitting side 1 to convert the parallel data to the serial data.

The auto-negotiation processing module: mainly implements the 1000BASE-X auto-negotiation processing state machine in IEEE standard 802.3-2008_section3 Figure 37-6, and can complete the capability coordination operation at both ends of the communication through auto-negotiation. SGMII and 1000BASE-X are different in the negotiation page table. In this module, the capability extraction function is treated differently to ensure that it can be distinguished by mode. The capabilities supported by SGMII/1000BASE-X are common to other processing and state machines.

The single channel unbalanced value processing unit (transmitting side) is responsible for providing unbalanced value processing of a single 8B/10B encoding module. The specific principle is that the input of the next encoding is the output of the previous encoding unbalanced value. Single channel processing directly sends the unbalanced value of the previous input to the next encoding, as shown in Figure 4. From the above principle analysis, it can be concluded that the processing only requires 1 bit of registers, and the logical resources are almost negligible.

The receiving side of the SGMII/1000BASE-X physical coding sublayer mainly includes a single channel alignment code detecting unit, a single channel unbalanced value calculating unit, and a PCS basic receiving unit. The PCS basic receiving unit includes an 8B/10B decoding module and a standard receiving state machine module.

Single channel alignment code detection unit: mainly finds the alignment code from the SERDES receiving side 1 data stream to align the data to the state before the parallel conversion of the transmitting side. Since the serial bit stream has been disturbed by the serial value stream of the serial bit stream recovered by the SERDES receiving side 1 processing, it is indispensable that the processing needs to regain the boundary of the 10-bit data. From the above principle analysis, it can be concluded that the processing requires about 40 bits of registers.

Single channel unbalanced value calculation unit (receiving side): responsible for providing unbalanced value processing of the single channel 8B/10B decoding module. The specific principle is that the unbalanced value input of the next decoding operation is the output of the previous decoding operation unbalanced value. Single channel processing directly sends the unbalanced value of the previous input to the next decoding. If the data on the transmitting side is incorrect or the link is unstable, the data stream is abnormal. The receiving side can easily detect the error according to the unbalanced value check.

8B/10B decoding module: mainly performs 8B/10B decoding operation on the aligned data, and also needs to detect the correctness of the unbalanced value of the decoded data, the correctness of the encoding code group, the validity of the control code, and finally the 10-bit data. Revert back to an 8-bit ordered collection. The 8bit data stream is re-decapsulated into the GMII mode data after being processed by the receiving state machine module. At the same time, it is necessary to detect whether the various control characters conform to the protocol, and the unqualified ordered set code group needs to be marked with an error flag.

Standard Receive State Machine Module: Implements IEEE Standard 802.3-2008_section3Figure 36-7a, Figure 36-7b. Perform decapsulation on the 8B/10B decoded data, including removing the start code /S/, the end code/T/, restoring the idle code/I/, and detecting the correctness of the code group encapsulation for the illegal package. The data stream is flagged with an error and sent back to the MAC layer via GMII. At the same time, it is necessary to give a link synchronization status indication according to the code group status in conjunction with the synchronization state machine of the standard IEEE standard 802.3-2008_section3 Figure 36-9.

As shown in FIG. 2, the processing structure diagram of the physical coding sublayer of QSGMII is shown.

The sending side of the QSGMII physical coding sublayer sends the data sent by the MAC layer transmitting side 0, the MAC layer transmitting side 1, the MAC layer transmitting side 2, and the MAC layer transmitting side 3 through 4 sets of GMIIs, and performs ordered set replacement insertion and encoding processing, wherein The first channel needs to perform special processing on the ordered set code group (replace the K28.5 code group with the K28.1 code group), and the bit width after encoding is changed from 4×8 bit to 4×10 bits. The physical coding sublayer outputs a 4×10 bit data stream to the SERDES transmitting side 2 at 125 Mhz, and the SERDES transmitting side 2 performs parallel-to-serial conversion on the parallel data into a serial bit stream output.

At this time, the actual rate is 4×10 bit×125Mhz=5 Gbps, and the effective rate is 5 Gpbs×8/10=4 Gbps except for the 8B/10B encoding. For the physical coding sublayer of the individual QSGMII, the corresponding SERDES transmitting side 2 and the SERDES receiving side 2 need only be fixed in a processing mode of 40 bit width. Note that the bit width of the transmitting side and the receiving side of the SERDES2 is actually 40 bits, but the fixed operation is in the 40 bit all effective mode at this time.

The receiving side of the QSGMII physical coding sublayer receives 40-bit parallel data from the SERDES receiving side 2 after serial-to-parallel conversion recovery, and performs decoding operation after channel realignment, wherein the first channel needs to restore K28.1 to K28.5 code. The group is used for de-encapsulation processing of the ordered set, and the data stream is restored to 4 sets of GMII data streams after being removed from the ordered set code group, and sent back to the MAC layer receiving side 0, the MAC layer receiving side 1, the MAC layer receiving side 2, and the MAC. The layer receives side 3.

The transmitting side of the QSGMII physical coding sublayer mainly includes four PCS basic transmission units, a 4-channel unbalanced value calculation unit, and a K28.1 code group replacement unit.

The K28.1 code group replacement unit is mainly used to replace the K28.5 code group of the first channel, so as to facilitate The receiving side detects the K28.1 code group to align the data boundary of the channel, which is a unique processing method of QSGMII.

The PCS basic transmission unit structure is as described above. The QSGMII transmission side needs to use four PCS basic transmission units, including the PCS basic transmission side 0, the PCS basic transmission side 1, the PCS basic transmission side 2, and the PCS basic transmission side 3.

4-channel unbalanced value calculation unit (transmission side): Since the four channel code groups in the link propagation are transmitted using the same SERDES, the imbalance value actually needs to indicate the correlation between the four channels, then Equilibrium value processing should not be propagated back and forth in only one channel. In the specific implementation, the code group encoded by the previous channel needs to be transmitted to the code group of the next channel code, so that the correlation of coding between different channels is realized. The structure of the specific implementation principle is shown in Figure 5.

The receiving side of the QSGMII physical coding sublayer mainly includes four PCS basic receiving units, a 4-channel unbalanced value calculating unit, and a 4-channel alignment code detecting unit.

The PCS basic receiving unit structure is as described above. The QSGMII receiving side needs to use four PCS basic receiving units including the PCS basic receiving side 0, the PCS basic receiving side 1, the PCS basic receiving side 2, and the PCS basic receiving side 3.

The K28.1 code group restoration unit is configured to wait for the data boundary of the four channels to be aligned, and then restore the K28.1 code group of the first channel to the K28.5 code group, so that the QSGMII state machine processing module on the receiving side is The original SGMII/1000BASE-X can share a set of processing mechanisms.

4-channel unbalanced value calculation unit (receiving side): Since the code groups of the four channels in the link propagation are propagated using the same SERDES, the unbalanced value actually needs to represent the association between the four channels. In this way, as long as one channel does not work correctly, the unbalanced value propagation of the other four channels will be incorrect, resulting in decoding errors of all channels. The specific implementation structure can be seen in Figure 5.

The 4-channel alignment code detection unit mainly finds the alignment code from the SERDES data stream to align the data to the state before the parallel conversion of the transmitting side. Compared to single-channel processing, alignment in 4 channels not only represents the boundary of parallel data for each channel, but also the number of channels per channel. According to the segmentation boundary point. When the position is determined, the data for all channels is already aligned. On the contrary, if the position is judged incorrectly, it will cause confusion of all channel data.

It can be seen that QSGMII is actually slightly adjusted on the structure of 4 sets of SGMII. From the previous analysis, SGMII can actually use one of the PCS basic transmission units and one PCS basic reception unit of QSGMII, plus a single-channel unbalanced value calculation unit (transmission side) and a single-channel unbalanced value calculation unit ( Receive side), single channel alignment code detection unit to get. Therefore, the QSGMII structure can be modified as shown in Figure 3, which is the QSGMII/SGMII/1000BASE-X hybrid PCS processing structure. In FIG. 3, the SERDES transmitting side 3 and the SERDES receiving side 3 cannot be fixed in the 10-bit effective or 40-bit effective mode at this time, and it is necessary to determine the working bit width of the transmitting side and the receiving side of the SERDES3 according to the control signal (pcs_mod) at this time. In this way, the multi-mode hybrid PCS implementation can be realized by the module change of the PCS layer.

As shown in FIG. 3, the hybrid coding sublayer of the embodiment of the present invention includes: a sending module 31, where the sending module 31 includes:

The four physical coding sub-layer PCS basic transmission unit 311 is configured to perform orderly set replacement insertion and coding processing on the data sent by the four media access control MAC layer transmission sides via the four sets of Gigabit media independent interfaces GMII;

The code group replacement selecting unit 312 is configured to replace the K28.5 code group of the first channel with the K28.1 code group when the mixed texture sublayer works in the 4-channel scene after the encoding process; When the layer works in a single channel scenario, no replacement processing is performed;

The unbalanced value calculation processing unit 313 is configured to calculate an unbalanced value required for the encoding process by using the 4-channel unbalanced value when the mixed-coded sub-layer operates in the 4-channel scene; configured to operate the encoded sub-layer in the hybrid In the channel scenario, the unbalanced value required for the encoding process is calculated using the single channel unbalanced value.

Each PCS basic sending unit 311 includes:

Standard transmit state machine module (not shown) configured to send the MAC layer of a channel The data sent by the sending side via the GMII is subjected to an ordered set replacement insertion process, and the data is encapsulated into a code group form suitable for 8B/10B encoding;

The 8B/10B encoding module (not shown) is configured to perform 8B/10B encoding processing on the code stream processed by the standard transmitting state machine module, wherein the 8B/10B encoding process requires an unbalanced value as an input; after the encoding process Send data to the SERDES transmitting side for parallel data to serial data conversion;

The auto-negotiation processing module (not shown) is configured to complete the capability coordination operation at both ends of the communication by auto-negotiation.

The code group replacement selecting unit 312 is further configured to: select, according to the control signal, whether to replace the K28.1 code group of the first channel with the K28.5 code group;

The unbalanced value calculation processing unit 313 is further configured to select, according to the control signal, an unbalanced value required to calculate the encoding process by using the 4-channel unbalanced value, or calculate the imbalance required for the encoding process by using the single-channel unbalanced value. value.

As can be seen in Figure 3, compared to the original QSGMII implementation, the mode selection is added to several modules that need to distinguish between single-channel and 4-channel processing. The pcs_mod signal is used to indicate that the current PCS needs to work at 4. In a channel scene or in a single channel scene. In this way, by analyzing the structural similarities and differences between QSGMII and SGMII/1000BASE-X in detail, the processing methods that can work in three arbitrary modes are obtained.

As shown in FIG. 3, the hybrid coding sublayer of another embodiment of the present invention includes: a receiving module 32, and the receiving module 32 includes:

The alignment code detection processing unit 321 is configured to use the 4-channel alignment code to detect data aligned from the SERDES when the mixed-coded sub-layer operates in a 4-channel scene; and when the mixed-coded sub-layer operates in a single-channel scene, Single channel alignment code detection aligns the data sent from the SERDES;

The code group restoration selection unit 322 is configured to process the sub-layers when the data is aligned When working in a 4-channel scene, the K28.1 code group of the first channel is restored to the K28.5 code group; when the mixture coding sub-layer works in a single-channel scene, no restoration process is performed;

4 PCS basic receiving units 323, after data alignment, decode the 4 channels of data and remove the ordered set code groups, and restore the data streams to the GMII data stream and send them to the 4 MAC layer receiving sides;

The unbalanced value calculation processing unit 324 is configured to calculate an unbalanced value required for the decoding process by using the 4-channel unbalanced value when the mixed-coded sub-layer operates in the 4-channel scene; configured to work when the mixed-coded sub-layer operates in the single In the channel scenario, the unbalanced value required for the decoding process is calculated using the single channel imbalance value.

Each PCS basic receiving unit 323 includes:

8B/10B decoding module (not shown) configured to perform 8B/10B decoding operation on the aligned data;

The standard receiving state machine module (not shown) is configured to remove the ordered set code group from the 8B/10B decoded data, and restore the data stream to the GMII data stream and send it to the MAC layer receiving side.

The code group restoration selecting unit 322 is further configured to select, according to the control signal, to align the data sent from the SERDES by using 4-channel alignment code detection or single channel alignment code detection;

The unbalanced value calculation processing unit 324 is further configured to: according to the control signal, select an unbalanced value required to calculate a decoding process by using a 4-channel unbalanced value, or calculate an imbalance required for the decoding process by using a single-channel unbalanced value. value.

As shown in FIG. 3, the hybrid coding sublayer of another embodiment of the present invention includes the above-described transmitting module 31 and receiving module 32. The mixture coding sublayer of FIG. 3 mainly performs selection processing on the code group replacement and reduction selection unit, the unbalance value calculation processing unit, and the alignment code detection processing unit of the original QSGMII processing structure.

The alignment code detection processing unit combines a 4-channel alignment code detection unit with a single-channel alignment code check The measurement unit determines which unit to use for processing according to pcs_mod.

The unbalanced value calculation processing unit combines a 4-channel unbalanced value calculation unit and a single-channel unbalanced value calculation unit, and determines which unit to use for processing according to pcs_mod.

According to pcs_mod to decide whether to do the original K28.1 code group replacement and restore operation. When working in QSGMII mode (ie 4-channel working scene), the K28.1 code group operates as described above. When working in SGMII/1000BASE-X mode (ie single-channel working scene), the direct pipeline delays one beat output.

FIG. 6 is a schematic flowchart of a data transmission method of a mixture coding sublayer according to an embodiment of the present invention. As shown in FIG. 6, the data transmission method of the mixture coding sublayer includes the following steps:

Step 601: Receive data sent by the MAC layer transmitting side through the GMII.

Step 602: Perform an ordered set replacement insertion and an encoding process on the data, wherein, when performing the encoding process, when the mixed coding sublayer works in a 4-channel scene, the 4-channel unbalanced value is used to calculate the coding process. The equalization value, when the mixture coding sublayer works in a single channel scenario, uses the single channel imbalance value to calculate the imbalance value required for the encoding process.

Specifically, the data sent by the GMII of the MAC layer of one channel is subjected to the ordered set replacement insertion process, and the data is encapsulated into a code group format suitable for 8B/10B coding;

The processed code stream is subjected to 8B/10B encoding processing, wherein the 8B/10B encoding process requires an unbalanced value as an input.

After that, the capability coordination operation at both ends of the communication is completed through auto-negotiation.

Step 603: When the mixture coding sublayer works in the 4-channel scene, replace the K28.5 code group of the first channel with the K28.1 code group, and send it to the SERDES transmission side to perform parallel data to serial data conversion; When the mixture coding sublayer works in a single channel scenario, no replacement processing is performed, and data is directly sent to the SERDES transmitting side for parallel data to serial data conversion.

FIG. 7 is a schematic flowchart of a data receiving method of a mixture coding sublayer according to an embodiment of the present invention. As shown in FIG. 7, the data receiving method of the mixture coding sublayer includes the following steps:

Step 701: When the mixture coding sublayer works in a 4-channel scene, the 4-channel alignment code is used to detect the alignment of the data sent from the SERDES; when the mixture coding sub-layer works in a single-channel scene, the single-channel alignment code detection is used. Data alignment from SERDES.

Step 702: After the data is aligned, when the mixture coding sublayer works in the 4-channel scene, the K28.1 code group of the first channel is restored to the K28.5 code group; when the mixture coding sublayer works in the single channel scene When it is not restored.

Step 703: After the data is aligned, the 4-channel data is decoded and the ordered set code group is removed, and the data stream is restored to the GMII data stream and sent to the 4 MAC layer receiving side; wherein, when the mixture is encoded When the layer works in a 4-channel scene, the 4-channel unbalanced value is used to calculate the unbalanced value required for the decoding process; when the mixed-coded sub-layer operates in a single-channel scene, the single-channel unbalanced value is used to calculate the decoding process. Unbalanced value.

Specifically, the 8B/10B decoding operation is performed on the aligned data; the ordered set code group is removed from the 8B/10B decoded data, and the data stream is restored to the GMII data stream and sent to the MAC layer receiving side.

Compared to the original QSGMII implementation logic resources increased by about 62bit registers and three 2-to-1 selector resources, compared to the original overall logic consumption, the added resources are minimal, almost no additional resource consumption. Compared with the 4-channel processing, the functions of the single-channel unbalanced value calculation unit (sending side), the single-channel unbalanced value calculation unit (receiving side), and the single-channel alignment code detection unit are relatively simple, and the corresponding QSGMII logic basis is corresponding. Adding these modules adds about 1% of the resources, which is almost negligible. Compared to the direct use of SGMII/1000BASE-X plus QSGMII physical coding sub-layer design can save at least 1/4 of the resource consumption.

At present, there are other related technologies that have the implementation of multi-mode Ethernet architecture. However, the focus is on complex design ideas and new architecture processing, but it does not take into account the similarity between QSGMII and SGMII/1000BASE-X structures and the reuse of resources. Sexually. For a variety of etheric modules The processing can be summarized into several processing methods:

The physical coding sublayers of different rates are used to select before entering the SERDES, so as to achieve the purpose of working at multiple rates. This kind of processing method is a waste of resources and a high cost.

Using a non-standard process, the large bandwidth division is reused for a small rate. For example, a 1Gbps rate interface can actually be used to transmit 10 100Mbps interfaces. This is actually similar to the QSGMII processing method, that is, 10 100M interfaces are all used as 1G, and a single 100M interface is used to achieve 100M. The disadvantage of this type of processing is that it is not a standard interface and cannot be interfaced with the 100M standard interface. Although the QSGMII works in the single channel mode and achieves a rate of 1 Gbps, the separate QSGMII cannot be connected to the SGMII (although the rate is consistent), so the present application implements the functions of the QSGMII/SGMII through resource multiplexing.

Compared with the rest of the prior art, the present application analyzes the similarities and differences of several mode structures in detail, and carries out the multiplexing of logical resources with ingenious and high efficiency, and achieves the realization of realizing the low-cost multi-mode physical coding sub-layer port. Externally support 1000BASE-X/SGMII/QSGMII to meet the application needs of different scenarios and interface with different manufacturers' interfaces.

The embodiment of the invention further describes a storage medium in which a computer program is stored, the computer program being configured to execute the data transmission method of the mixture coding sublayer described in the foregoing embodiment.

The embodiment of the invention further describes a storage medium in which a computer program is stored, the computer program being configured to execute the data receiving method of the mixture coding sublayer described in the foregoing embodiment.

It is to be understood that while the specification has been described in detail, the embodiments of the invention may The above detailed description is only for the specific description of the possible embodiments of the present invention, and is not intended to limit the scope of the present invention. Equivalent embodiments or modifications are within the scope of the invention.

Industrial applicability

By carefully analyzing the similarities and differences of the physical coding sublayer processing structures of QSGMII and SGMII/1000BASE-X, the original QSGMII reusable logic module is extracted, and a small amount of SGMII/1000BASE-X logic module resources are added to obtain simultaneous support. A physical coding sublayer with a Gigabit interface mode and little resource cost. Compared to the direct use of SGMII/1000BASE-X plus QSGMII physical coding sub-layer design can save at least 1/4 of the resource consumption.

Claims (14)

A hybrid coding sublayer, the hybrid coding sublayer comprising: a transmitting module, the transmitting module comprising: 4 physical coding sub-layer PCS basic transmission units are configured to perform orderly set replacement insertion and coding processing on data sent by four sets of Gigabit media independent interfaces GMII on the transmission side of the four medium access control MAC layers; The code group replacement selection unit is configured to replace the K28.5 code group of the first channel with the K28.1 code group when the mixed coding sublayer works in the 4-channel scene after the encoding process; When working in a single channel scenario, no replacement processing is performed; The unbalanced value calculation processing unit is configured to calculate an unbalanced value required for the encoding process by using the 4-channel unbalanced value when the mixed-coded sub-layer operates in the 4-channel scene; configured to operate the single-channel when the mixed-coded sub-layer operates In the scene, the unbalanced value required for the encoding process is calculated using the single channel unbalanced value. The mixture coding sublayer of claim 1 wherein each PCS basic transmission unit comprises: The standard transmission state machine module is configured to perform an ordered set replacement insertion process on the MAC layer transmitting side of one channel by the GMII, and encapsulate the data into a code group format suitable for 8B/10B encoding; The 8B/10B encoding module is configured to perform 8B/10B encoding processing on the code stream processed by the standard transmitting state machine module, wherein the 8B/10B encoding process requires an unbalanced value as an input; after the encoding process, the data is sent to the SERDES transmitting side. Do parallel data to serial data conversion; The auto-negotiation processing module is configured to perform capability coordination operations at both ends of the communication through auto-negotiation. The mixture encoding sublayer according to claim 1 or 2, wherein The code group replacement selecting unit is further configured to: according to the control signal, select whether to replace the K28.1 code group of the first channel with the K28.5 code group; The unbalanced value calculation processing unit is further configured to: according to the control signal, select an unbalanced value required to calculate the encoding process by using the 4-channel unbalanced value, or calculate an unbalanced value required for the encoding process by using the single-channel unbalanced value. . A hybrid coding sublayer, the hybrid coding sublayer comprising: a receiving module, the receiving module comprising: The alignment code detection processing unit is configured to use the 4-channel alignment code to detect data alignment from the SERDES when the mixed-coded sub-layer operates in a 4-channel scene; when the mixture coding sub-layer operates in a single-channel scene, use the single Channel alignment code detection aligns the data sent from the SERDES; The code group restoration selection unit is configured to restore the K28.1 code group of the first channel to the K28.5 code group when the mixture coding sublayer works in the 4-channel scene after data alignment; when the mixture encodes the sub-layer When working in a single channel scenario, no restoration is performed; 4 PCS basic receiving units, after data alignment, decode 4 channels of data and remove the ordered set code group, and restore the data stream to the GMII data stream and send it to the 4 MAC layer receiving sides; The unbalanced value calculation processing unit is configured to calculate an unbalanced value required for the decoding process by using the 4-channel unbalanced value when the mixed-coded sub-layer operates in the 4-channel scene; configured to operate the single-channel when the mixed-coded sub-layer operates In the case of the scene, the unbalanced value required for the decoding process is calculated using the single channel unbalanced value. The mixture coding sublayer of claim 4, wherein each of the PCS basic receiving units comprises: The 8B/10B decoding module is configured to perform 8B/10B decoding operation on the aligned data; The standard receiving state machine module is configured to remove the ordered set code group from the 8B/10B decoded data, and restore the data stream to the GMII data stream and send the data stream to the MAC layer receiving side. The mixture encoding sublayer according to claim 4 or 5, wherein The code group restoration selection unit is further configured to select, according to the control signal, to align the data sent from the SERDES by using 4-channel alignment code detection or single channel alignment code detection; The unbalanced value calculation processing unit is further configured to: according to the control signal, select an unbalanced value required to calculate a decoding process by using a 4-channel unbalanced value, or calculate an unbalanced value required for the decoding process by using a single-channel unbalanced value; . A mixture processing code sublayer comprising the transmission module according to any one of claims 1 to 3, and the reception module according to any one of claims 4 to 6. A data transmission method for a mixture coding sublayer, the method comprising: Receiving data sent by the MAC layer transmitting side through the GMII; And performing an ordered set replacement insertion and an encoding process on the data, wherein, when performing the encoding process, when the mixed coding sublayer operates in a 4-channel scene, the unbalanced value required for the encoding process is calculated by using the 4-channel unbalanced value, When the mixture coding sublayer works in a single channel scenario, the unbalanced value required for the encoding process is calculated using the single channel imbalance value; When the mixture coding sublayer works in a 4-channel scene, the K28.5 code group of the first channel is replaced with the K28.1 code group, and sent to the SERDES transmission side for parallel data to serial data conversion; When the coding sublayer works in a single channel scenario, it does not perform replacement processing, and directly sends data to the SERDES transmitting side to convert parallel data to serial data. The data transmission method of the mixture coding sub-layer according to claim 8, wherein the performing the ordered set replacement insertion and the coding process on the data comprises: The data sent by the GMII of the MAC layer of one channel is subjected to the ordered set replacement insertion process, and the data is encapsulated into a code group format suitable for 8B/10B coding; The processed code stream is subjected to 8B/10B encoding processing, wherein the 8B/10B encoding process requires an unbalanced value as an input. The data transmission method of the mixture coding sub-layer according to claim 8, wherein the method further comprises: The capability coordination operation at both ends of the communication is completed through auto-negotiation. A data receiving method for a mixture coding sublayer, the method comprising: When the mixture coding sublayer works in a 4-channel scene, the 4-channel alignment code is used to detect the alignment of the data sent from the SERDES; when the mixture coding sub-layer works in a single-channel scene, the single-channel alignment code detection will be issued from the SERDES. Data alignment; After the data is aligned, when the mixture coding sublayer works in a 4-channel scene, the K28.1 code group of the first channel is restored to the K28.5 code group; when the mixture coding sub-layer works in a single-channel scene, Reduction process After the data is aligned, the 4-channel data is decoded and the ordered set code group is removed, and the data stream is restored to the GMII data stream and sent to the four MAC layer receiving sides; wherein, when the mixture coding sub-layer works In the 4-channel scene, the unbalanced value required for the decoding process is calculated by using the 4-channel unbalanced value; when the mixed-coded sub-layer operates in a single-channel scene, the imbalance required for the decoding process is calculated by using the single-channel unbalanced value. value. The data receiving method of the mixed coding sublayer according to claim 11, wherein after the data is aligned, the 4-channel data is decoded and the ordered set code group is removed, and the data stream is restored to the GMII data. The stream is sent to the receiving side of the four MAC layers, including: The 8B/10B decoding operation is performed on the aligned data; the ordered set code group is removed from the 8B/10B decoded data, and the data stream is restored to the GMII data stream and sent to the MAC layer receiving side. A storage medium storing a computer program configured to execute the data transmission method of the mixture coding sublayer according to any one of claims 8 to 10. A storage medium storing a computer program configured to execute the data receiving method of the mixture encoding sublayer of claim 11 or 12.

Images