阅读说明：本技术 一种pmd子层链路训练的方法及系统 (Method and system for PMD sublayer link training ) 是由 段有楠 戚晨希 邱建峰 于 2020-04-01 设计创作，主要内容包括：本发明揭示了一种PMD子层链路训练的方法及系统,所述方法包括：调整发送端均衡器的参数C(–1)和C(+1),均调整到最大眼高的位置,接收端均衡器根据最大眼高对应的一组信道参数模板进行参数调整。本发明在当前链路状况不确定的情况下,可以将对端设备的发送和本端设备的接收均衡器参数动态调整到比较理想的水平,正常建立连接,且能保证较长时间的传输不出现错包。(The invention discloses a method and a system for PMD sublayer link training, wherein the method comprises the following steps: and adjusting parameters C (-1) and C (+1) of the equalizer at the transmitting end to the position of the maximum eye height, and adjusting the parameters of the equalizer at the receiving end according to a group of channel parameter templates corresponding to the maximum eye height. Under the condition that the current link condition is uncertain, the invention can dynamically adjust the parameters of the sending equalizer of the opposite terminal equipment and the receiving equalizer of the local terminal equipment to relatively ideal levels, normally establish connection and ensure that error packets do not occur in long-time transmission.)

1. A method of PMD sublayer link training, the method comprising:

s1, adjusting the parameter C (-1) of the equalizer at the sending end to the position of the maximum eye height;

s2, adjusting the parameter C (+1) of the equalizer at the transmitting end to the position of the maximum eye height;

and S3, the receiving end equalizer adjusts parameters according to a group of channel parameter templates corresponding to the maximum eye height, wherein the parameters comprise a parameter C (-1) and a parameter C (+ 1).

2. The method of claim 1, wherein before the step of S1, the parameters C (-1) and C (+1) are adjusted to initial states.

3. The method of claim 1, wherein in the S1 and S2, the parameter adjustment is performed by invoking a single step training message interaction process, and the invoking single step training message interaction process is a common process of the steps S1 and S2.

4. The method according to claim 1, wherein the S1 includes:

s11, placing the cursor at the position of the current parameter C (-1) at the middle position of the graduated scale;

s12, judging whether the cursor at the current parameter C (-1) position is larger than 0 and whether the state replied by the receiving end is maximum or minimum, if so, entering the step S13;

s13, gradually decreasing the cursor at the current parameter C (-1) position from the middle position to 0;

s14, gradually increasing the cursor at the current parameter C (-1) position from 0 to the tail end of the graduated scale, recording the eye height while increasing progressively each time, and finding out the graduated position with the maximum eye height;

and S15, finally, gradually decreasing the cursor at the current parameter C (-1) position from the tail end of the scale back to the scale position with the maximum eye height.

5. The method according to claim 1, wherein the S2 adjustment process is the same as the S1 adjustment process, and the S2 specifically includes:

s21, placing the cursor at the position of the current parameter C (+1) at the middle position of the graduated scale;

s22, judging whether the cursor at the position of the current parameter C (+1) is larger than 0 and whether the state replied by the receiving end is the maximum or the minimum, if so, entering the step S13;

s23, gradually decreasing the cursor at the current parameter C (+1) position from the middle position to 0;

s24, gradually increasing the cursor at the current parameter C (+1) position from 0 to the tail end of the graduated scale, recording the eye height while increasing progressively each time, and finding out the graduated position with the maximum eye height;

and S25, finally, gradually decreasing the cursor at the current parameter C (+1) position from the tail end of the scale back to the scale position with the maximum eye height.

6. The method according to claim 1, wherein the S3 includes: according to a pre-determined channel parameter template Rx _ preset [ n ], respectively configuring parameters and measuring the eye height at the moment, finding a group of channel parameter templates Rx _ preset [ max _ n ] corresponding to the maximum eye height, then finely adjusting parameters of a receiving end, and finally performing parameter configuration according to a receiving end parameter fine adjustment value corresponding to the maximum eye height, wherein n is a channel parameter template number which is an integer greater than 0, and max _ n is the channel parameter template number of the maximum eye height.

7. The method of claim 3, wherein the flow of the single-step training message interaction comprises:

the transmitting end assembles the training frame and transmits to the receiving end;

the sending end judges whether the normal range, the maximum state and the minimum state of the parameter adjustment fed back by the receiving end are received or not, and if the normal range, the maximum state and the minimum state are received, the states are stored into the state replied by the receiving end;

then setting all parameters of the sending end to be unchanged, and assembling training frames to a receiving end;

the sending end judges whether the state that the parameter fed back by the receiving end is not adjusted is received, and if the state is received, the flow is ended.

8. A system for PMD sublayer link training, the system comprising:

the parameter adjusting device of the sending end equalizer comprises a parameter C (-1) adjusting module and a parameter C (+1) adjusting module, wherein the parameter C (-1) adjusting module is used for adjusting the parameter C (-1) of the sending end equalizer to the position with the maximum eye height, and the parameter C (+1) adjusting module is used for adjusting the parameter C (+1) of the sending end equalizer to the position with the maximum eye height;

and the receiving end equalizer parameter adjusting device is used for adjusting parameters according to a group of channel parameter templates corresponding to the maximum eye height, wherein the parameters comprise a parameter C (-1) and a parameter C (+ 1).

9. The system of claim 8, further comprising: and the parameter adjusting device of the equalizer at the sending end adjusts parameters by calling the single-step training message interaction module, and the parameter C (-1) adjusting module and the parameter C (+1) adjusting module share the single-step training message interaction module.

10. The system of claim 8, wherein the parameter C (-1) adjustment module and the parameter C (+1) adjustment module each comprise:

the vernier setting unit is used for placing the vernier at the current parameter position in the middle position of the graduated scale;

the vernier judgment unit is used for judging whether the vernier at the current parameter position is larger than 0 and whether the state returned by the receiving end is not the maximum or the minimum, and if the state is met, the vernier position adjustment unit is started;

the vernier position adjusting unit is used for gradually decreasing the vernier at the current parameter position from the middle position to 0, gradually increasing the vernier at the current parameter position from 0 to the tail end of the graduated scale, recording the eye height while increasing the eye height each time, finding the graduated position with the maximum eye height, and finally gradually decreasing the vernier at the current parameter position from the tail end of the graduated scale back to the graduated position with the maximum eye height.

Technical Field

The invention belongs to the technical field of PMD sublayer link training, and particularly relates to a method and a system for PMD sublayer link training.

Background

After the connection is established between the Ethernet ports, under the scene of determining the link condition, the signal quality can be adjusted by manually configuring the equalizer parameters of the sending end and the receiving end. However, if the current link conditions are uncertain, it is difficult to manually tune the equalizer parameters of the port. The IEEE802.3 standard describes link training of a PMD (Physical media dependent interface) sublayer, so that a receiving end of a port can automatically adjust and optimize equalizer parameters of a transmitting end through a training frame which is continuously transmitted, thereby achieving the purposes of improving link performance and reducing error codes.

The IEEE802.3 standard takes a 10G BASE-KR (ten-gigabit Ethernet physical coding sublayer (Pcs) standard defined by IEEE 802.3) port as an example, and details a data frame format used by PMD sublayer control link training and an interactive flow of a home-peer single link training step. However, in a practical application scenario, a complete link quality training should include several training steps, and the initiating device determines which FFE (Feed forward equalizer) parameter of the peer device is adjusted at each step, whether to increase or decrease, how many steps the whole link training is performed, when to end, and the like. These specific implementations, not specified by the IEEE802.3 standard, are designed by the manufacturers themselves.

Currently, a set of universal parameter adjustment steps is defined for all 10G BASE-KR ports, and adjustment is performed according to fixed steps after training is started, however, the application range of the universal parameter adjustment steps is limited, good effect can be obtained only by docking with equipment of the same manufacturer, and the universal parameter adjustment steps are difficult to adapt to implementation modes of different manufacturers.

Therefore, in view of the above technical problems, there is a need to provide a PMD sublayer link quality training scheme that can adapt to different manufacturers.

Disclosure of Invention

In view of the above, the present invention provides a method and system for PMD sublayer link training.

In order to achieve the above object, an embodiment of the present invention provides the following technical solutions:

a method of PMD sublayer link training, the method comprising:

s1, adjusting the parameter C (-1) of the equalizer at the sending end to the position of the maximum eye height;

s2, adjusting the parameter C (+1) of the equalizer at the transmitting end to the position of the maximum eye height;

and S3, the receiving end equalizer adjusts parameters according to a group of channel parameter templates corresponding to the maximum eye height, wherein the parameters comprise a parameter C (-1) and a parameter C (+ 1).

In one embodiment, before the step S1, the parameter C (-1) and the parameter C (+1) are adjusted to the initial state.

In one embodiment, in S1 and S2, the parameter adjustment is performed by invoking a single step training message interaction, which is a common procedure in S1 and S2.

In one embodiment, the S1 includes:

s11, placing the cursor at the position of the current parameter C (-1) at the middle position of the graduated scale;

s12, judging whether the cursor at the current parameter C (-1) position is larger than 0 and whether the state replied by the receiving end is maximum or minimum, if so, entering the step S13;

s13, gradually decreasing the cursor at the current parameter C (-1) position from the middle position to 0;

s14, gradually increasing the cursor at the current parameter C (-1) position from 0 to the tail end of the graduated scale, recording the eye height while increasing progressively each time, and finding out the graduated position with the maximum eye height;

and S15, finally, gradually decreasing the cursor at the current parameter C (-1) position from the tail end of the scale back to the scale position with the maximum eye height.

In an embodiment, the S2 adjustment process is the same as the S1 adjustment process, and the S2 specifically includes:

s21, placing the cursor at the position of the current parameter C (+1) at the middle position of the graduated scale;

s22, judging whether the cursor at the position of the current parameter C (+1) is larger than 0 and whether the state replied by the receiving end is the maximum or the minimum, if so, entering the step S13;

s23, gradually decreasing the cursor at the current parameter C (+1) position from the middle position to 0;

s24, gradually increasing the cursor at the current parameter C (+1) position from 0 to the tail end of the graduated scale, recording the eye height while increasing progressively each time, and finding out the graduated position with the maximum eye height;

and S25, finally, gradually decreasing the cursor at the current parameter C (+1) position from the tail end of the scale back to the scale position with the maximum eye height.

In one embodiment, the S3 includes: according to a pre-determined channel parameter template Rx _ preset [ n ], respectively configuring parameters and measuring the eye height at the moment, finding a group of channel parameter templates Rx _ preset [ max _ n ] corresponding to the maximum eye height, then finely adjusting parameters of a receiving end, and finally performing parameter configuration according to a receiving end parameter fine adjustment value corresponding to the maximum eye height, wherein n is a channel parameter template number which is an integer greater than 0, and max _ n is the channel parameter template number of the maximum eye height.

In one embodiment, the process of single-step training message interaction includes:

the transmitting end assembles the training frame and transmits to the receiving end;

the sending end judges whether the normal range, the maximum state and the minimum state of the parameter adjustment fed back by the receiving end are received or not, and if the normal range, the maximum state and the minimum state are received, the states are stored into the state replied by the receiving end;

then setting all parameters of the sending end to be unchanged, and assembling training frames to a receiving end;

the sending end judges whether the state that the parameter fed back by the receiving end is not adjusted is received, and if the state is received, the flow is ended.

Correspondingly, another embodiment of the present invention provides the following technical solution:

a system for PMD sublayer link training, comprising:

the parameter adjusting device of the sending end equalizer comprises a parameter C (-1) adjusting module and a parameter C (+1) adjusting module, wherein the parameter C (-1) adjusting module is used for adjusting the parameter C (-1) of the sending end equalizer to the position with the maximum eye height, and the parameter C (+1) adjusting module is used for adjusting the parameter C (+1) of the sending end equalizer to the position with the maximum eye height;

and the receiving end equalizer parameter adjusting device is used for adjusting parameters according to a group of channel parameter templates corresponding to the maximum eye height, wherein the parameters comprise a parameter C (-1) and a parameter C (+ 1).

In one embodiment, the system further comprises: and the parameter adjusting device of the equalizer at the sending end adjusts parameters by calling the single-step training message interaction module, and the parameter C (-1) adjusting module and the parameter C (+1) adjusting module share the single-step training message interaction module.

In one embodiment, the parameter C (-1) adjusting module and the parameter C (+1) adjusting module each include:

the vernier setting unit is used for placing the vernier at the current parameter position in the middle position of the graduated scale;

the vernier judgment unit is used for judging whether the vernier at the current parameter position is larger than 0 and whether the state returned by the receiving end is not the maximum or the minimum, and if the state is met, the vernier position adjustment unit is started;

the vernier position adjusting unit is used for gradually decreasing the vernier at the current parameter position from the middle position to 0, gradually increasing the vernier at the current parameter position from 0 to the tail end of the graduated scale, recording the eye height while increasing the eye height each time, finding the graduated position with the maximum eye height, and finally gradually decreasing the vernier at the current parameter position from the tail end of the graduated scale back to the graduated position with the maximum eye height.

The invention has the following beneficial effects: the invention judges the signal quality based on the eye diagram level of the receiving end, designs a link quality training system based on the eye diagram level, dynamically adjusts the equalizer parameters of the opposite end/the local end under the condition that the current link condition is uncertain, thereby leading the signal transmission loss to be properly compensated, adjusting the parameters of the sending equalizer of the opposite end equipment and the receiving equalizer of the local end equipment to be at a more ideal level, normally establishing connection (link up), and ensuring that the transmission for a longer time does not generate error packets.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a schematic diagram of a Training frame format;

FIG. 2 is a diagram illustrating a format of a Coefficient update field;

FIG. 3 is a diagram illustrating the format of the Status report field;

FIG. 4 is a schematic flow diagram of the process of the present invention;

FIG. 5 is a schematic flow chart of steps 1 and 2 of the present invention;

FIG. 6 is a schematic flow chart of step 3 of the present invention;

FIG. 7 is a flow chart illustrating a single step training message interaction according to the present invention;

FIG. 8 is a schematic flow chart of step 4 of the present invention.

Detailed Description

In order to make those skilled in the art better understand the technical solution of the present invention, the technical solution in the embodiment of the present invention will be clearly and completely described below with reference to the drawings in the embodiment of the present invention, and it is obvious that the described embodiment is only a part of the embodiment of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The invention discloses a method and a system for PMD sublayer link training, which judge the signal quality based on the eye diagram level of a receiving end, dynamically adjust the parameters of an opposite end/local end equalizer under the condition that the current link condition is uncertain, can adjust the parameters of the receiving equalizer of local end equipment and the sending of opposite end equipment to a more ideal level, normally establish connection (linkup) and can ensure that error packets do not occur in long-time transmission.

Before describing the scheme, an a-side and a B-side are taken as examples to explain a specific mode of lower PMD sublayer link training. Suppose two 10G BASE-KR ports with A end and B end are interconnected through DAC (Direct Attach Cable). The link training is a bidirectional process, which includes a process (a- > B) initiated by the a end and instructing the B end to perform parameter adjustment, and a process (B- > a) initiated by the B end and instructing the a end to perform parameter adjustment. The processes of starting and ending A- > B and B- > A are independent respectively, can be carried out simultaneously or staggered, but the link training is considered to be finally finished only if the two-way training is finished.

As shown in fig. 1, in the format of a Training frame, the Training frame includes 4 parts, 2 of which are used for parameter adjustment, Coefficientupdate, and Status report.

Specific definitions of Coefficientupdate and Status report are shown in fig. 2 and fig. 3, respectively, where Coefficientupdate: the method is mainly used for instructing the B-end equipment to execute what parameter adjusting action. Bits 0 to 1, 2 to 3, and 4 to 5 respectively represent parameters C (-1), C (0), and C (+1) of the equalizer at the 3 transmitting ends, and whether the parameters are increased (01), decreased (10), or unchanged (00). The increase or decrease can only be adjusted by adding 1 or subtracting 1, if it is desired to add 10 to the opposite end, 10 times of increase action is required. Bit 12 indicates that the tuned end directly tunes 3 parameters to the initial state. Bit 13 indicates that the tuned terminal directly tunes 3 parameters to preset states. The other reserved bits are meaningless. Status report: mainly reflects the condition of parameter adjustment of the B terminal and feeds back the condition to the A terminal. Bits 0 to 1, 2 to 3, and 4 to 5 respectively represent parameters C (-1), C (0), and C (+1) of the equalizer at the 3 transmitting ends, and after one adjustment, the maximum (11), the minimum (10), the normal range (01), or the unadjusted range (00) is achieved. These are the bits that the B-side needs to use to reply to the a-side after its action. Specifically, bit 15 is set by the a-side, which indicates that the a-side considers that the link has been trained to an ideal state, and ends the training. The other reserved bits are meaningless.

As shown in fig. 4, the method for PMD sublayer link training disclosed in the present invention comprises the following steps:

step 1, adjusting a parameter C (-1) and a parameter C (+1) of a transmitting end to an initial state.

Referring to fig. 5, in a situation where a current link condition is uncertain, for a parameter C (-1) of the sending-end equalizer, the method of the present invention equates the parameter C (-1) to a scale, and adjusts sending-end parameters C (-1) and C (+1) from an initial state, that is, an initial state of Coefficientupdate.

And step 2, adjusting the parameter C (-1) of the equalizer at the transmitting end to the position of the maximum eye height.

As shown in fig. 5, step 2 mainly includes the following steps:

and S11, placing the cursor at the position of the current parameter C (-1) at the middle position of the graduated scale.

Specifically, the sender parameter C (-1) calls the flow of single step training message interaction (Updateflow) in the initial state, and the cursor is placed at the middle position of the scale, i.e. coeff _ pos is CN1_ L EN/2, where coeff _ pos is a local variable used for recording the cursor at the current parameter C (-1) position, and CN1_ L EN is the scale length of the parameter C (-1), i.e. the maximum adjustment range.

S12, determine whether the cursor at the current parameter C (-1) position is larger than 0 and whether the status returned by the receiving end is not the maximum or the minimum, if yes, go to step S13.

Specifically, it is determined that coeff _ pos is greater than 0 and L P _ stat is not max/min (maximum or minimum), if both are satisfied, step S13 is entered, wherein L P _ stat is a local variable used to save the status of the receiver reply, and if not, step S14 is advanced.

S13, gradually decreasing the cursor at the current parameter C (-1) position from the middle position to 0.

Specifically, setting the parameter C (-1) of Coefficientupdate to decrease (increment), calling Updateflow, and gradually decreasing coeff _ pos by 1: coeff _ pos, and then determining whether eye height eye _ h is 0, if yes, it indicates that the cursor at the position of parameter C (-1) has been decremented to 0, and S14 is entered, otherwise, it returns to S12. The eye diagram is a graph which is displayed by superimposing a signal at a receiving end on an oscilloscope, the overall characteristic of the signal is embodied, the quality of the signal can be estimated accordingly, the horizontal axis represents the period/time, and the vertical axis represents the amplitude.

And S14, gradually increasing the cursor at the current parameter C (-1) position from 0 to the tail end of the graduated scale, recording the eye height while increasing each time, and finding the graduated position with the maximum eye height.

Specifically, the maximum eye height is initially 0, i.e., max _ eye _ h is 0, where max _ eye _ h is a local variable used to store the maximum eye height, it is determined whether coeff _ pos is less than or equal to CN1_ L EN and L P _ stat is not max/min (maximum or minimum), i.e., coeff _ pos is less than or equal to CN1_ L EN and L P _ stat is not max/min, if both are satisfied, the current eye height eye _ h is read, and it is determined whether the current eye height eye _ h is greater than the maximum eye height max _ eye _ h, if so, the cursor position max _ coeff _ pos corresponding to the maximum eye height is set as the current cursor position, then the cursor is gradually incremented to the end of the scale, specifically, the parameter C (-1) of coeff _ anode is set as an increment (initial), the cursor position is called, the cursor position is gradually incremented by 1, i.e., the increment is set to the maximum eye height of CN1 n + apox _ P5, and the current eye height P _ pos is directly incremented to the loop, and the current eye height P _ pos is set as the increment.

And S15, finally, gradually decreasing the cursor at the current parameter C (-1) position from the tail end of the scale back to the scale position with the maximum eye height.

Specifically, in the above S14, if coeff _ pos is less than or equal to CN1_ L EN and L P _ stat is not satisfied with max/min, it is continuously determined whether the current cursor position is the cursor position max _ coeff _ pos corresponding to the maximum eye height, that is, coeff _ pos is not equal to max _ coeff _ pos.

And step 3, adjusting the parameter C (+1) of the equalizer at the transmitting end to the position of the maximum eye height.

As shown in fig. 6, the adjusting process of the parameter C (+1) is substantially the same as the adjusting process of the parameter C (-1), and specifically includes the following steps:

and S21, placing the cursor at the position of the current parameter C (+1) at the middle position of the graduated scale.

Specifically, the sender parameter C (+1) is placed at the middle position of the scale, i.e. coeff _ pos ═ CN1_ L EN/2, where coeff _ pos is a local variable used to record the cursor of the current parameter C (+1) position, CP1_ L EN is the scale length of the parameter C (+1), i.e. the maximum adjustment range.

S22, determine whether the cursor at the current parameter C (+1) position is greater than 0 and whether the status returned by the receiving end is the maximum or the minimum, if yes, go to step S23.

Specifically, it is determined that coeff _ pos is greater than 0 and L P _ stat is not max/min (maximum or minimum), if both are satisfied, step S23 is entered, wherein L P _ stat is a local variable used to save the status of the receiver reply, and if not, step S24 is advanced.

S23, gradually decreasing the cursor at the current parameter C (+1) position from the middle position to 0.

Specifically, setting the parameter C (+1) of Coefficientupdate to decrease (increment), calling Updateflow, and gradually decreasing coeff _ pos by 1: coeff _ pos, and then determining whether eye height eye _ h is 0, if yes, it indicates that the cursor at the position of parameter C (+1) has been decremented to 0, and S24 is entered, otherwise, it returns to S22.

And S24, gradually increasing the cursor at the current parameter C (+1) position from 0 to the tail end of the graduated scale, recording the eye height while increasing each time, and finding the graduated position with the maximum eye height.

Specifically, the maximum eye height is initially 0, i.e., max _ eye _ h is 0, where max _ eye _ h is a local variable used to store the maximum eye height, it is determined whether coeff _ pos is less than or equal to CP1_ L EN and L P _ stat is not max/min (maximum or minimum), i.e., coeff _ pos is less than or equal to CP1_ L EN and L P _ stat is not max/min, if both are satisfied, the current eye height eye _ h is read, and it is determined whether the current eye height eye _ h is greater than the maximum eye height max _ eye _ h, if so, the cursor position max _ coeff _ pos corresponding to the maximum eye height is set as the current cursor position, then the cursor is gradually incremented to the end of the scale, specifically, the parameter C (+1) of coeff is set to be incremented, the cursor position is called, the cursor position is gradually incremented by 1, i.e., the index C +1 of coeff _ pomin is gradually incremented to the end of the scale, and the current eye height ep _ po _ P3/P3 is directly incremented to the current eye height.

And S25, finally, gradually decreasing the cursor at the current parameter C (+1) position from the tail end of the scale back to the scale position with the maximum eye height.

Specifically, in the above S24, coeff _ pos ≦ CP1_ L EN, and L P _ stat is not satisfied with max/min, it is continuously determined whether the current cursor position is the cursor position max _ coeff _ pos corresponding to the maximum eye height, that is, coeff _ pos?, if yes, the step 4 is entered, if no, the cursor at the current parameter C (+1) position is gradually decreased from the end of the scale back to the scale position with the maximum eye height, specifically, the parameter C (+1) of the coeff Update is set to decrease (delete), the Update flow is called, the coeff _ pos is gradually decreased by 1, that is, coeff _ pos max — then, the determination is entered into coeff _ pos ═ coeff _ pos.

It should be noted that the above-mentioned single-step training message interaction flow is a common flow of the step 2 and the step 3, and as shown in fig. 7, it specifically includes:

the transmitting end assembles the training frame and transmits to the receiving end.

The sending end judges whether the normal range (updated), the MAXIMUM (MAXIMUM) and the MINIMUM (MINIMUM) of the parameter adjustment fed back by the receiving end is received, if so, the state is stored into the state (L P _ state) returned by the receiving end, wherein, MAXIMUM represents that if the parameter value adjusted by the receiving end equipment reaches the upper limit MAX _ L IMIT, the parameter reaches the MAXIMMUM state, the content of the update _ state field returned to the sending end is MAXIMUM, and MINIMUM represents that if the parameter value adjusted by the receiving end equipment reaches the lower limit MIN _ L IMIT, the parameter reaches the MINIMMUM state, and the content of the update _ state field returned to the sending end is MINIMUM.

And then setting all the parameters of the sending end to be unchanged, and assembling the training frame to the receiving end.

Namely, setting all three parameters of the Coefficient update at the sending end as (hold), wherein the hold represents: in a construction mode of the Training frame, after the receiving end completes parameter adjustment and replies update _ status to the sending end, the sending end constructs hold and sends the hold back to the receiving end, and the receiving end recovers the parameter state to NOT _ UPDATED after receiving the hold.

The sending end judges whether the state that the parameter fed back by the receiving end is not adjusted is received, and if the state is received, the flow is ended.

That is, the sending end determines whether to receive the NOT _ UPDATED state fed back by the receiving end, and if so, the flow is ended. If NOT, whether the NOT _ UPDATED state fed back by the receiving end is received or NOT is continuously judged until the NOT _ UPDATED state is received.

And step 4, adjusting parameters of the equalizer at the receiving end according to a group of channel parameter templates corresponding to the maximum eye height, wherein the parameters comprise a parameter C (-1) and a parameter C (+ 1).

Specifically, as shown in fig. 8, after both parameters C (-1) and C (+1) are adjusted, adjustment of the equalizer parameter at the receiving end is started. The receiver side adjustment does not involve message interaction with the opposite end (i.e., the sender side), but is performed internally at the home end. The adjusting process specifically comprises the following steps:

s41, finding a set of channel parameter templates Rx _ preset [ max _ n ] corresponding to the maximum eye height.

The receiving end equalizer configures parameters according to a pre-measured typical channel parameter template Rx _ preset n, wherein Rx _ preset [ n ] represents a typical channel receiving parameter template stored in an array form, n is a channel parameter template number, the eye height is an integer greater than 0, specifically, the maximum eye height max _ eye _ h is initially 0, the channel parameter template number n is initially 0, the receiving end equalizer configures parameters according to Rx _ preset [ n ], reads the current eye height eye _ h while configuring, judges whether the current eye height eye _ h is greater than the maximum eye height max _ eye _ h, if yes, the template number max _ n corresponding to the maximum eye height is set as the channel parameter template number n, if not, then the serial number n of the channel parameter template is gradually reduced by 1 and then circularly enters the equalizer of the receiving end to configure parameters according to Rx _ preset [ n ] until a group of channel parameter templates Rx _ preset [ max _ n ] corresponding to the maximum eye height is found. Where max _ n is the channel parameter template number of the maximum eye height.

And then, the equalizer at the receiving end configures parameters according to a channel parameter template Rx _ preset [ max _ n ] corresponding to the maximum eye height.

And S42, fine tuning the parameters of the receiving end.

Specifically, the maximum eye height max _ eye _ h is initially 0, and the value reg of the receiving-end parameter fine tuning is set to the channel parameter template Rx _ preset [ max _ n ] +3, i.e. each receiving-end parameter is subjected to positive 3 fine tuning. And then judging whether the value reg of the fine tuning of the receiving end parameters is more than or equal to the Rx _ preset [ max _ n ] -3 of the channel parameter template, if so, performing parameter configuration by the receiving end equalizer according to the value of the fine tuning of the receiving end parameters, simultaneously reading the current eye height eye _ h, and judging whether the current eye height eye _ h is more than the maximum eye height max _ eye _ h, if so, setting the value max _ reg of the fine tuning of the receiving end parameters corresponding to the maximum eye height as the value reg of the fine tuning of the current receiving end parameters, otherwise, decreasing the value reg of the fine tuning of the current receiving end parameters 1 by 1 and circularly entering the judgment whether the value reg of the fine tuning of the receiving end parameters is more than or equal to the Rx _ preset [ max _ n ] -3 of the channel parameter template, namely, performing fine tuning of minus 3 on each receiving end. If the value reg of the fine tuning of the receiving end parameter is smaller than the channel parameter template Rx _ preset [ max _ n ] -3, the process directly proceeds to step S43.

And S43, finally, performing parameter configuration according to the receiving end parameter fine adjustment value corresponding to the maximum eye height.

Specifically, the receiving end equalizer performs parameter configuration according to a receiving end parameter fine adjustment value max _ reg corresponding to the maximum eye height, and sets receiverready of the Status report field to 1, that is, it indicates that the link has been trained to an ideal state, and then assembles a training frame and sends the training frame to the sending end, and ends the training process.

It should be noted that the above-mentioned reading of the eye height is not realized by an oscilloscope, but by SerDes hardware logic, and the SerDes can calculate the eye height of the signal in real time in the process of receiving and recovering the signal, and can be obtained by reading the address of the corresponding register when in use.

In addition, the reason why the invention only adjusts the parameters C (-1) and C (+1) and does not adjust the parameter C (0) is that in practical application, the SerDes hardware implementation modes are different, and the change of the C (0) parameter of some SerDes can influence the high precision of a receiving end eye, thereby influencing the effect; on the other hand, considering that each training must start from initial state, the parameter C (0) can reach the requirement, so only two other parameter adjustments are made.

According to the technical scheme, the invention has the following advantages: the invention can be applied to all port forms (10G BASE-KR/25G BASE-KR/25G BASE-CR/40G BASE-KR4/40G BASE-CR4/100GBASE-KR4/100G BASE-CR4 and the like) supporting the control function of the PMD sublayer, and the ports can process the following application scenes: the local terminal and the opposite terminal are in butt joint, and because the link quality is uncertain, the parameters of sending and receiving equalizers applied at the two terminals are not suitable for the current link, and at the moment, a large number of error packets exist in data transmission, and port link up can not be established even seriously. After the link training is started, the parameters of the sending equalizer of the opposite terminal equipment and the receiving equalizer of the local terminal equipment can be adjusted to a more ideal level based on the method of the invention, link up is normally established, and the error packet can not occur in the long-time transmission.

Correspondingly, the invention discloses a system for PMD sublayer link training, which comprises the following steps:

the parameter adjusting device of the sending end equalizer comprises a parameter C (-1) adjusting module and a parameter C (+1) adjusting module, wherein the parameter C (-1) adjusting module is used for adjusting the parameter C (-1) of the sending end equalizer to the position with the maximum eye height, and the parameter C (+1) adjusting module is used for adjusting the parameter C (+1) of the sending end equalizer to the position with the maximum eye height;

and the receiving end equalizer parameter adjusting device is used for adjusting parameters according to a group of channel parameter templates corresponding to the maximum eye height, wherein the parameters comprise a parameter C (-1) and a parameter C (+ 1).

Preferably, the system of the present invention further comprises: and the parameter adjusting device of the equalizer at the sending end adjusts parameters by calling the single-step training message interaction module, and the parameter C (-1) adjusting module and the parameter C (+1) adjusting module share the single-step training message interaction module.

Preferably, the parameter C (-1) adjusting module and the parameter C (+1) adjusting module each include:

the vernier setting unit is used for placing the vernier at the current parameter position in the middle position of the graduated scale;

the vernier judgment unit is used for judging whether the vernier at the current parameter position is larger than 0 and whether the state returned by the receiving end is not the maximum or the minimum, and if the state is met, the vernier position adjustment unit is started;

the vernier position adjusting unit is used for gradually decreasing the vernier at the current parameter position from the middle position to 0, gradually increasing the vernier at the current parameter position from 0 to the tail end of the graduated scale, recording the eye height while increasing the eye height each time, finding the graduated position with the maximum eye height, and finally gradually decreasing the vernier at the current parameter position from the tail end of the graduated scale back to the graduated position with the maximum eye height.

The workflow and the principle of each module may be described with reference to the above method, which is not described herein again.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions.

For convenience of description, the above devices are described as being divided into various modules by functions, and are described separately. Of course, the functionality of the modules may be implemented in the same one or more software and/or hardware implementations in implementing one or more embodiments of the present description.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

As will be appreciated by one skilled in the art, embodiments of one or more embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, one or more embodiments of the present description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, one or more embodiments of the present description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

One or more embodiments of the present description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. One or more embodiments of the specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.

Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.