WO2024051511A1 - 一种频偏消除实现方法、设备以及系统 - Google Patents

一种频偏消除实现方法、设备以及系统 Download PDF

Info

Publication number
WO2024051511A1
WO2024051511A1 PCT/CN2023/115261 CN2023115261W WO2024051511A1 WO 2024051511 A1 WO2024051511 A1 WO 2024051511A1 CN 2023115261 W CN2023115261 W CN 2023115261W WO 2024051511 A1 WO2024051511 A1 WO 2024051511A1
Authority
WO
WIPO (PCT)
Prior art keywords
node
variable
length
additional information
information
Prior art date
Application number
PCT/CN2023/115261
Other languages
English (en)
French (fr)
Inventor
罗军平
潘伟
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2024051511A1 publication Critical patent/WO2024051511A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/0014—Carrier regulation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/0014—Carrier regulation
    • H04L2027/0024—Carrier regulation at the receiver end
    • H04L2027/0026—Correction of carrier offset

Definitions

  • the present application relates to the field of communication technology, and in particular, to a frequency offset elimination implementation method, device and system.
  • serial communication is widely used to realize communication between computers and between various parts within the computer.
  • the clocks of both communicating parties can be designed with reference to the standard clock. But in actual situations, due to production factors (for example, different crystal oscillators) or human factors (for example: spread spectrum settings), the clock frequency of the transmitter usually deviates from the clock frequency of the standard clock, and the clock frequency of the receiver There are also deviations from the clock frequency of the standard clock. If the above two deviations are inconsistent, the clock frequencies of both communicating parties will be inconsistent, resulting in data loss or resource waste.
  • This application discloses a frequency offset elimination implementation method, device and system, which can adjust the deviation of the clock frequencies of both communicating parties while effectively reducing the loss of link bandwidth.
  • the first aspect provides an implementation method for frequency offset elimination, including:
  • the sending end of the first node generates a variable-length padding message, and the sending end of the first node sends the variable-length padding message.
  • the receiving end of the second node receives the variable-length padding message.
  • the identification information of the variable-length padding message is the first identifier
  • the receiving end of the second node determines that the variable-length padding message does not include additional information.
  • the variable-length padding message includes additional information.
  • the variable-length stuffing message includes frequency offset compensation information.
  • the variable-length padding message further includes additional information; when additional information is not required, the variable-length padding message does not include additional information.
  • the additional information includes at least one of control information, status information, request information and communication information.
  • variable-length padding message including additional information or a variable-length padding message that does not include additional information. Only when additional information is needed, a relatively long variable-length padding message including additional information is generated. When no additional information is needed, a relatively short length variable-length padding message excluding additional information can be generated. Therefore, , which can not only perform frequency offset compensation, but also reduce the loss of link bandwidth.
  • variable-length padding message also includes identification information.
  • the identification information is the first identifier.
  • the length can be When the padding message requires additional information, the identification information is the second identifier.
  • the identifier encoding of the first identifier is different from the identifier encoding of the second identifier.
  • the message length of the variable-length padding message when additional information is required is greater than the message length of the variable-length padding message when no additional information is required.
  • the first clock of the first node and the second clock of the second node are not the same, or the first clock of the first node and the second clock of the second node are not the same.
  • the clocks are the same.
  • a frequency offset elimination implementation device including a generating module and a sending module.
  • the generating module is used to generate a variable-length padding message, wherein the variable-length padding message includes frequency offset compensation information.
  • the variable-length padding message also includes additional information.
  • the variable-length stuffing message does not include additional information, where the additional information includes at least one of control information, status information, request information and communication information; the sending module is used to send the length Variable padding message.
  • variable-length padding message also includes identification information.
  • the identification information is the first identifier.
  • the length can be When the padding message requires additional information, the identification information is the second identifier.
  • the identifier encoding of the first identifier is different from the identifier encoding of the second identifier.
  • the first clock of the first node and the second clock of the second node are not the same, or the first clock of the first node and the second clock of the second node are not the same.
  • the clocks are the same.
  • a frequency offset elimination implementation device including: a receiving module and a determining module.
  • the receiving module is configured to receive a variable-length padding message; the determining module is configured to determine the variable-length padding message when the identification information of the variable-length padding message is a first identifier. No additional information is included, and when the identification information of the variable-length padding message is the second identifier, it is determined that the variable-length padding message includes additional information.
  • the additional information includes at least one of control information, status information, request information and communication information.
  • the identifier encoding of the first identifier is different from the identifier encoding of the second identifier.
  • the message length of the variable-length padding message when additional information is required is greater than the message length of the variable-length padding message when no additional information is required.
  • a communication device in a fourth aspect, includes a processor and a memory; the processor is used to execute instructions stored in the memory, so that the communication device executes the first node as in the first aspect. Or the method described at any angle of the second node.
  • a fifth aspect provides a computer program product containing instructions, characterized in that when the instructions are executed by a communication device, the communication device causes the communication device to execute any angle of the first node or the second node as in the first aspect. the method described.
  • a readable storage medium including program instructions.
  • the program instructions When the program instructions are executed by a communication device, the communication device executes as described in the first aspect the first node or the second node from any angle. Methods.
  • a frequency offset elimination implementation system including a first node and a second node, wherein the first node and the second node can communicate with each other, and the first node is used to execute As in the method of the first node angle in the first aspect, the second node is used to perform the method as described in the second node angle in the first aspect.
  • FIG. 1 is a schematic diagram of the two clocks involved in this application controlling different modules respectively;
  • FIGS. 2A and 2B are schematic diagrams of padding messages sent by the sending end and padding messages received by the receiving end involved in this application;
  • Figure 3 is a schematic flow chart of a frequency offset elimination method provided by this application.
  • FIGS. 4A to 4B are schematic structural diagrams of some variable-length padding messages provided by this application.
  • Figure 5 is a schematic structural diagram of a communication system provided by this application.
  • Figure 6 is a schematic structural diagram of a communication device provided by this application.
  • Serial communication refers to a communication method that transmits one bit of data at a time on a computer bus or other data channel, and continues the above single process, for example, the Fast Peripheral Component Interconnect Standard (Peripheral Component Interconnect Standard) interconnect express (PCIe), compute express link (CXL), etc.
  • PCIe Fast Peripheral Component Interconnect Standard
  • CXL compute express link
  • Parallel communication refers to a communication method in which multi-bit data is transmitted simultaneously through parallel lines, such as the peripheral component interconnect (PCI) standard.
  • PCIe is a high-speed serial computer expansion bus standard and is a type of PCI.
  • PCI is a bus standard that connects computer motherboards and external devices. It is currently the most widely used interface in personal computers. Almost all motherboard products have this slot.
  • PCIe follows the programming concepts and communication standards of PCI, but PCIe has many advantages such as stronger anti-interference ability, faster data transmission rate and better compatibility.
  • CXL is an open interconnection standard.
  • CXL technology is based on the mature PCIe infrastructure and leverages the physical and electronic interfaces of PCIe 5.0 to provide advanced protocols in three key areas: input/output protocols, memory protocols and coherence interfaces. This enables CXL to provide Ultra-fast interconnects and better memory consistency, thereby reducing software stack complexity and overall system cost, further improving the performance of data-intensive applications.
  • serial communication Compared with parallel communication, serial communication has the advantages of simple structure, fewer signal lines, support for long-distance data transmission, and low communication cost. Therefore, many current computer products use serial communication to realize communication between computers. , communication between various parts within the computer.
  • the clock frequency of the standard clock is 100MHz
  • the clock frequency of the first clock is 99.9997MHz/100.0003MHz
  • the clock frequency of the second clock is 100.0003MHz/99.9997MHz.
  • the functional modules in the white area in the figure are driven by the first clock
  • the functional modules in the gray area are driven by the second clock.
  • the first clock drives the sending end 110 of the first port of the first node (hereinafter referred to as the sending end 110 of the first node) to send a data stream to the receiving end 120 of the second port (hereinafter referred to as the receiving end 120 of the second node).
  • the receiving end 120 of the second node uses the clock data recovery (clock data recovery, CDR) module 130 to recover the clock signal in the same phase as the sending end 110 of the first node (That is, the clock signal generated by the first clock), and then use this clock signal to collect data subsequently sent by the sending end 110 of the first node. Therefore, the receiving end 120 of the second node collects the data sent by the sending end 110 of the first node according to the clock frequency of the first clock.
  • clock data recovery clock data recovery
  • the receiving end 120 of the second node After receiving the data, the receiving end 120 of the second node first writes the data into the elastic buffer (elastic buffer) 140 of the second port of the second node. After that, the sending end 150 of the second port of the second node (hereinafter referred to as The sending end 150 of the second node reads data from the elastic buffer 140 and sends it to the upper-layer logic 160 of the second port of the second node for subsequent identification, processing, etc. of the data.
  • the sending end 150 of the second node reads data from the elastic buffer 140 and sends it to the upper-layer logic 160 of the second port of the second node for subsequent identification, processing, etc. of the data.
  • the receiving end 120 of the second node writes the read data into the elastic buffer 140 according to the clock frequency of the first clock.
  • the sending end 150 of the second node writes the read data from the elastic buffer 140 according to the clock frequency of the second clock. Read data in.
  • the clock frequency of the first clock is inconsistent with the clock frequency of the second clock, it means that the receiving end 120 of the second node writes data into the elastic buffer 140 at the same speed as the sending end 150 of the second node writes data from the elastic buffer.
  • the speed at which data is read in zone 140 is different.
  • the clock frequency of the first clock is greater than the clock frequency of the second clock, it means that the receiving end 120 of the second node writes data to the elastic buffer 140 faster than the sending end 150 of the second node reads data from the elastic buffer 140 As time goes by, the elastic buffer 140 may overflow, resulting in data loss.
  • the clock frequency of the first clock is lower than the clock frequency of the second clock, it means that the speed at which the receiving end 120 of the second node writes data to the elastic buffer 140 is slower than the speed at which the sending end 150 of the second node reads data from the elastic buffer 140 Speed, as time goes by, an idle state may occur in the elastic buffer 140, which may cause an empty state when reading data from the elastic buffer, causing underflow.
  • a padding message including a padding field may be sent between the sending end 110 of the first node and the receiving end 120 of the second node.
  • the padding message containing padding fields sent by the sending end 110 of the first node is approximately 40 bytes, and may include 3 padding fields (body), an end field (end) and an Control field (control).
  • the stuffing field usually includes 4 bytes and is used to carry the stuffing code.
  • the stuffing code can be data irrelevant to the communication between the first node and the second node; the end field usually includes 4 bytes and is used to carry the end.
  • Code; the control field is usually 24 bytes and is used to carry the control code.
  • the control code is used by the first node to control the second node.
  • the padding message sent by the sender of the first node can pass through two nodes and then reach the receiving end of the second node.
  • Each node can add or delete a padding field based on the padding message. Therefore, as shown in Figure 2B, the stuffing message received by the receiving end of the second node may be 32 bytes, 36 bytes, 40 bytes, 44 bytes or 48 bytes.
  • the padding message received by the receiving end of the second node may be 32 bytes; when one of the two nodes deletes a padding field, the second node The padding message received by the receiving end of the second node can be 36 bytes; when neither node deletes or adds padding fields, the padding message received by the receiving end of the second node can be 40 bytes; when the two nodes When one of the nodes adds a padding field, the padding message received by the receiving end of the second node can be 44 bytes; when both nodes add a padding field, the receiving end of the second node receives The padding message can be 48 bytes.
  • the padding message includes 40 bytes, including 3 padding fields, each padding field is 4 bytes, and passes through up to two intermediate nodes, and each intermediate node can be added or deleted.
  • a padding field is used as an example to illustrate. In actual applications, padding packets can include more or less bytes, can include more or less padding fields, and pass through more or fewer intermediate nodes. Each padding field may include more or less bytes, each intermediate node may add or delete more padding fields, the sending intervals of padding messages may be different, etc., which are not specifically limited here.
  • the speed at which the receiving end 120 of the second node writes data to the elastic buffer 140 is greater than the speed at which the sending end 150 of the second node reads data from the elastic buffer 140 , at this time, the padding field in the padding message can be increased, or the number of padding packets can be increased.
  • the data sent by the sending end 110 of the first node to the receiving end 120 of the second node can contain more data related to the first node.
  • the receiving end 120 of the second node can discard data that is irrelevant to the communication between one node and the second node, thereby reducing the data that needs to be buffered in the elastic buffer 140 to avoid an overflow state in the elastic buffer 140 .
  • the receiving end 120 of the second node writes data to the elastic buffer 140 at a slower speed than the sending end 150 of the second node reads data from the elastic buffer 140
  • the padding field in the padding message can be reduced, or the number of padding packets can be reduced.
  • the data sent by the sending end 110 of the first node to the receiving end 120 of the second node can contain less Data irrelevant to the communication between the first node and the second node, thereby increasing the data buffered in the elastic buffer 140 to avoid an idle state of the elastic buffer 140 .
  • This application is suitable for serial communication scenarios. Specifically, in the scenario of serial communication, both communicating parties (including the sending end of the first node and the receiving end of the second node) need to be driven by a clock to complete the transmission of data. However, when the frequency offset of the transmitting end of the first node is inconsistent with the frequency offset of the receiving end of the second node, data loss or resource waste may occur, so padding messages need to be sent for adjustment.
  • the frequency offset of the sending end of the first node refers to the deviation between the clock frequency of the sending end of the first node and the clock frequency of the standard clock.
  • the frequency offset of the receiving end of the second node refers to the difference between the clock frequency of the receiving end of the second node and the clock frequency of the standard clock. Frequency offset from the clock frequency of the standard clock.
  • This application provides a frequency offset elimination implementation method, which is suitable for serial communication systems, that is, the sending end of the first node and the receiving end of the second node communicate in a serial manner, and no additional information is required.
  • frequency offset elimination is performed by sending a variable-length padding message that does not include additional information.
  • frequency offset elimination and additional information are performed by sending a variable-length padding message that includes additional information.
  • Figure 3 is a schematic flow chart of a frequency offset elimination implementation method provided by this application. As shown in Figure 3, the frequency offset elimination implementation method provided by this application includes:
  • S101 The sending end of the second node sends a message to the receiving end of the first node.
  • the receiving end of the first node receives the message sent by the sending end of the second node.
  • the message is used to instruct the first node to send a filling message for frequency offset compensation.
  • the second node sends a message to the first node for instructing the first node to add a filling message.
  • the second node sends a message to the first node for instructing the first node to reduce the filling messages.
  • step S102 The first node determines whether an additional bearer is required. If not, proceed to step S103A to generate a variable-length padding message that does not include additional information. If necessary, proceed to step S103B to generate a variable-length padding message that includes additional information.
  • the length variable padding message includes frequency offset compensation information.
  • the length variable padding message also includes additional information.
  • the length variable padding message also includes additional information.
  • Variable padding messages do not include additional information. That is to say, the variable-length padding message includes two different forms.
  • the variable-length padding message includes frequency offset compensation information and additional information; when additional information is not required, the variable-length padding message Includes frequency offset compensation information and does not include additional information.
  • variable-length padding message also includes identification information.
  • the identification information is the first identifier.
  • the length can be When the padding message requires additional information, the identification information is the second identifier.
  • the additional information includes at least one of control information, request information, status information, and communication information.
  • the control information refers to information about the first node controlling the second node, for example, information about the first node instructing the second node to perform operation processing.
  • the request information refers to the information that the first node requests the second node to control the first node.
  • the status information is information about the status of the first node, such as the location, power consumption, etc. of the first node. Communication information may include other information required for communication, etc.
  • variable-length padding messages are sent in cycles, and in each cycle, the first node may choose to generate different variable-length padding messages according to whether it needs to carry additional information.
  • S103A The first node generates a variable-length padding message that does not include additional information.
  • variable-length padding message includes frequency offset compensation information, the first identifier, and does not include additional information.
  • the variable-length padding message in this form may include: a first padding field and a first identification field.
  • the first padding field is used to carry frequency offset compensation information, and the frequency offset compensation information may be information irrelevant to communication between the first node and the second node.
  • the first identification field is used to carry a first identifier, and the first identifier is used to identify this message as a variable padding message that does not include additional information.
  • the length of the first padding field may be 4 bytes, and the length of the first identification field may be 4 bytes. Of course, the length of the first padding field can also be more or less bytes, and the length of the first identification field can also be more or less bytes, which are not specifically limited here.
  • the number of generated variable-length padding messages that do not include additional information is increased, thereby increasing the number of inserted data streams.
  • the number of variable-length padding messages that do not include additional information is increased.
  • the data sent by the sending end of the first node to the receiving end of the second node can contain more information that is irrelevant to the communication between the second node and the first node. data, the receiving end of the second node can discard these data, thereby reducing the data that needs to be buffered in the elastic buffer to avoid an overflow state in the elastic buffer.
  • the message instructs the first node to reduce padding messages and does not require additional information reduce the number of variable-length padding messages that are generated that do not include additional information, thereby reducing the number of variable-length padding messages that do not include additional information that are inserted into the data stream. fill In this way, the data sent by the sending end of the first node to the receiving end of the second node can contain less data that is irrelevant to the communication between the second node and the first node, so as to avoid the elastic buffer becoming idle. state.
  • the number of padding fields in the generated variable-length padding message that does not include additional information is increased, so that, The data sent from the sending end of the first node to the receiving end of the second node may contain more data that is irrelevant to the communication between the second node and the first node, and the receiving end of the second node may discard these data, thus reducing the need for Buffer data in the elastic buffer to avoid an elastic buffer overflow condition.
  • the message instructs the first node to reduce the padding message and does not require additional information reduce the number of padding fields in the generated variable-length padding message that does not include additional information. In this way, the sending end of the first node sends to the first node
  • the data at the receiving end of the second node may contain less data irrelevant to the communication between the second node and the first node to avoid an idle state in the elastic buffer.
  • the period for sending a variable-length padding message that does not include additional information is shortened.
  • the first node The data sent by the sending end to the receiving end of the second node may contain more data that is irrelevant to the communication between the second node and the first node.
  • the receiving end of the second node can discard these data, thereby reducing the need for buffering in the elastic buffer. data in the buffer to avoid an overflow condition in the elastic buffer.
  • the message instructs the first node to reduce padding messages and does not need to perform additional functions increase the period for sending variable-length padding messages that do not include additional information.
  • the sending end of the first node sends to the second node
  • the data at the receiving end may contain less data irrelevant to the communication between the second node and the first node to avoid an idle state in the elastic buffer.
  • S104A The first node sends a variable-length padding message that does not include additional information to the second node.
  • the second node receives the variable-length padding message that does not include additional information and is sent by the first node.
  • S105A The second node identifies the variable-length stuffing message that does not include additional information, and performs frequency offset compensation.
  • the second node can identify this message as a variable-length padding message that does not include additional information for the first identifier based on the identification information in the variable-length padding message, and perform frequency offset. compensate.
  • S103B The first node generates a variable-length padding message including additional information.
  • variable-length padding message including additional information includes frequency offset compensation information, a second identifier, and additional information.
  • the variable-length padding message in this form may include: a second padding field, a second identification field, and an additional information field.
  • the second padding field is used to carry frequency offset compensation information, where the frequency offset compensation information is information irrelevant to communication between the first node and the second node.
  • the second identification field is used to carry a second identifier, and the second identifier is used to identify this message as a variable-length padding message including additional information.
  • the additional information field is used to carry additional information.
  • the additional information includes at least one of control function information, request information, status information, and communication information.
  • the length of the second padding field may be 4 bytes, the second identification field may be 4 bytes, and the additional information field may be 24 bytes.
  • the length of the second padding field can also be more or less bytes, the length of the second identification field can also be more or less bytes, and the additional information field can also be more or less words. section, there is no specific limit here.
  • the sending end of the first node increases the number of variable-length padding messages generated including additional information, thereby increasing The number of variable-length padding messages including additional information inserted into the data stream.
  • the data sent by the sending end of the first node to the receiving end of the second node can contain more information related to the second node and the first node.
  • the receiving end of the second node can discard these data, thereby reducing the data that needs to be buffered in the elastic buffer to avoid an overflow state in the elastic buffer.
  • the sending end of the first node reduces the number of variable-length padding packets including additional information generated, thereby reducing the number of padding packets including additional information inserted into the data stream.
  • the number of padding messages with variable length In this way, the data sent by the sending end of the first node to the receiving end of the second node can contain less data irrelevant to the communication between the second node and the first node, thereby increasing the need for buffering. Data in the elastic buffer to prevent the elastic buffer from becoming idle.
  • the sending end of the first node when the message instructs the first node to add a padding message and requires additional information, the sending end of the first node adds a padding field of the generated variable-length padding message including the additional information.
  • the data sent by the sending end of the first node to the receiving end of the second node can contain more data that is irrelevant to the communication between the second node and the first node, and the receiving end of the second node can discard these data. , thereby reducing the data that needs to be buffered in the elastic buffer to avoid an overflow state in the elastic buffer.
  • the sending end of the first node reduces the number of padding fields in the generated variable-length padding message including additional information.
  • the data sent by the sender to the receiver of the second node may contain less data that is irrelevant to the communication between the second node and the first node, thereby increasing the data that needs to be buffered in the elastic buffer to avoid the elastic buffer becoming idle. state.
  • the message instructs the first node to add a padding message and needs to perform additional functions reduce The period of sending variable-length padding messages including additional information.
  • the data sent by the sending end of the first node to the receiving end of the second node may contain more data that is irrelevant to the communication between the second node and the first node.
  • the receiving end of the second node can discard these data, thereby reducing the data that needs to be buffered in the elastic buffer to avoid an overflow state in the elastic buffer.
  • the message instructs the first node to reduce padding messages and requires additional information increase the period of sending variable-length padding messages including additional information.
  • the sending end of the first node sends data to the receiving end of the second node. can contain less data irrelevant to the communication between the second node and the first node, thereby increasing the data that needs to be buffered in the elastic buffer to avoid an idle state in the elastic buffer.
  • S104B The first node sends a variable-length padding message including additional information to the second node.
  • the second node receives the variable-length padding message including additional information sent by the first node.
  • the second node identifies the variable-length padding message including additional information, and performs functions corresponding to frequency offset compensation and additional information.
  • the second node can identify this message as a variable-length padding message including additional information for the second identifier based on the identification information in the variable-length padding message, and perform frequency offset compensation. Functions corresponding to additional information.
  • the identifier encoding of the first identifier is different from the identifier encoding of the second identifier, and the length of the variable-length padding message that does not include additional information is shorter than that of the variable-length padding message that includes additional information.
  • the first clock of the first node and the second clock of the second node are not the same.
  • the first clock of the first node and the second clock of the second node are not the same.
  • the two clocks can also be the same. In this case, the period in which the first node sends padding packets to the second node can be longer to reduce unnecessary bandwidth consumption and at the same time carry additional information when needed.
  • the sending end of the first node when the link does not require additional information, the sending end of the first node sends a variable-length padding message that does not include additional information to eliminate frequency offset.
  • the sending end of the first node sends a variable-length padding message including additional information.
  • Figure 5 shows a schematic structural diagram of a communication system provided by this application.
  • the communication system may include a device 310 and a device 320 .
  • the device 310 may be deployed in the above-mentioned first node, and the device 320 may be deployed in the above-mentioned second node, so as to be able to adjust the clock offset between the second node and the first node and send additional information.
  • the device 310 includes a generating unit 311 and a sending unit 312.
  • the generating unit 311 is configured to generate a variable-length padding message, where the variable-length padding message includes frequency offset compensation information.
  • the variable-length padding message also includes additional information
  • the variable-length stuffing message does not include additional information, where the additional information includes at least one of control information, status information, request information and communication information.
  • the sending unit 312 is configured to send the variable length padding message.
  • the device 320 includes a receiving unit 321 and a determining unit 322.
  • the receiving unit 321 is used at the receiving end to receive the variable-length padding message.
  • Determining unit 322 configured to determine that the variable-length padding message does not include additional information when the identification information of the variable-length padding message is the first identifier. When the information is the second identifier, it is determined that the variable-length stuffing message includes additional information, where the additional information includes at least one of control information, status information, request information and communication information.
  • the device 310 provided in the above embodiments belongs to the same concept as the first node in the above method embodiments.
  • the generating unit 311 is used to execute the aforementioned S103A or 103B
  • the sending unit 312 is used to execute the sending part in the aforementioned S104A or S104B.
  • the device 320 provided by the embodiment belongs to the same concept as the second node in the above method embodiment.
  • the receiving unit 321 is used to perform the receiving part in the aforementioned S104A or S104B
  • the determining unit 322 is used to perform the step S105A or S105B. Specific implementation Please refer to the above method embodiment for details of the process, and will not be described again here.
  • the devices in the communication system provided by the embodiments of the present application are only exemplified by the division of the above functional units.
  • the above function allocation can be completed by different functional units as needed, that is, the internal structure of the above device 310 is divided into into different functional units to complete all or part of the functions described above.
  • FIG. 6 shows a schematic structural diagram of a communication device provided by this application.
  • the communication device 400 includes a communication port 410, and the communication port 410 includes a sending end 411, a receiving end 412, and an elastic buffer 413.
  • the communication port 410 belongs to the same concept as the first node in the above method embodiment.
  • the communication port 410 is specifically used to perform the aforementioned S101, S102, S103A, S104A and S103B, S104B to realize the communication port 410 and other ports (for example, the second node communication port) between Serial communication.
  • the communication port 410 has the same concept as the second node in the above method embodiment, and the communication port 410 is specifically used to perform the aforementioned steps S101, S104A, S105A and S104B, S105B.
  • the communication system may include a second node and a first node.
  • the second node includes a receiving end and a buffer
  • the first node includes a sending end.
  • the first node may also include a buffer.
  • the first node is used to perform the aforementioned S101, S102, S103A, S104A and S103B, and the angle step of the first node in S104B: specifically, the sending end of the first node is used to perform the aforementioned S101, S102, S103A, S104A and In S103B and S104B, the receiving end of the second node is used to perform the angle method of the second node in the aforementioned steps S101, S104A, S105A, and S104B and S105B.
  • the computer program product includes one or more computer instructions.
  • the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.
  • the computer instructions may be stored in or transmitted from one computer-readable storage medium to another, e.g., the computer instructions may be transferred from a website, computer, server, or data center Transmission to another website site, computer, server or data center via wired (eg, coaxial cable, optical fiber, digital subscriber line or wireless (eg, infrared, wireless, microwave, etc.) means.
  • the computer-readable storage medium may It is any available media that can be accessed by a computer or a data storage device such as a server or data center integrated with one or more available media.
  • the available media can be magnetic media (for example, floppy disks, hard disks, tapes), optical media (for example, DVD), or semiconductor media (for example, SSD), etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Communication Control (AREA)

Abstract

本申请涉及一种频偏消除实现方法、设备以及系统。所述方法包括：第一节点的发送端生成长度可变填充报文，其中，所述长度可变填充报文包括频偏补偿信息，当需要附加信息时，所述长度可变填充报文还包括附加信息，当不需要附加信息时，所述长度可变填充报文不包括附加信息，其中，所述附加信息包括控制信息，状态信息，请求信息以及通信信息的至少一种；所述第一节点的发送端发送所述长度可变填充报文。上述方案能够有效减少链路带宽损失。

Description

一种频偏消除实现方法、设备以及系统
本申请要求于2022年9月7日提交中国专利局、申请号为202211098687.8、申请名称为“一种频偏消除实现方法、设备以及系统”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及通信技术领域，尤其涉及一种频偏消除实现方法、设备以及系统。
背景技术
串行通信作为一种通信方式，被广泛用于实现计算机之间、计算机内部的各部分之间的通信。为了实现串行通信，通信双方的时钟可以参照标准时钟进行设计。但在实际情况中，由于生产因素(例如，晶振的不同)或人为因素(例如：展频设置)的存在，通常会导致发送端的时钟频率与标准时钟的时钟频率存在偏差，并且接收端的时钟频率与标准时钟的时钟频率也存在偏差。如果上述两种偏差不一致，会导致通信双方的时钟频率不一致，从而出现数据丢失或资源浪费的现象。
为了弥补两个时钟频率的偏差，通常需要在发送端和接收端之间发送包含填充字段的填充报文进行消偏，但是，这种方式会导致链路带宽的损失。
发明内容
本申请公开了一种频偏消除实现方法、设备以及系统，能够调节通信双方的时钟频率的偏差的同时有效减少链路带宽的损失。
第一方面，提供了一种频偏消除的实现方法，包括：
第一节点的发送端生成长度可变填充报文，所述第一节点的发送端发送所述长度可变填充报文。相应地，第二节点的接收端接收长度可变填充报文。当所述长度可变填充报文的标识信息为第一标识符时，所述第二节点的接收端确定所述长度可变填充报文不包括附加信息，当所述长度可变填充报文的标识信息为第二标识符时，所述第二节点的接收端确定所述长度可变填充报文包括附加信息。其中，所述长度可变填充报文包括频偏补偿信息。当需要附加信息时，所述长度可变填充报文还包括附加信息，当不需要附加信息时，所述长度可变填充报文不包括附加信息。所述附加信息包括控制信息，状态信息，请求信息以及通信信息的至少一种。
上述方案中，可以根据是否需要附加信息而选择生成包括附加信息的长度可变填充报文或者生成不包括附加信息的长度可变填充报文。只有需要附加信息的时候，才长度比较长的包括附加信息的长度可变填充报文，在不需要附加信息的时候，可以生成长度比较短的不包括附加信息的长度可变填充报文，因此，既能够进行频偏补偿，又能够减少链路带宽的损失。
在一些可能的设计中，所述长度可变填充报文还包括标识信息，当所述长度可变填充报文不需要附加信息时，所述标识信息为第一标识符，当所述长度可变填充报文需要附加信息时，所述标识信息为第二标识符。
在一些可能的设计中，所述第一标识符的标识符编码不同于所述第二标识符的标识符编码。
在一些可能的设计中，需要附加信息时的长度可变填充报文的报文长度大于不要附加信息时的长度可变填充报文的报文长度。
在一些可能的设计中，所述第一节点的第一时钟和所述第二节点的第二时钟并不相同，或者，所述第一节点的第一时钟和所述第二节点的第二时钟相同。
第二方面，提供了一种频偏消除实现装置，包括生成模块以及发送模块。所述生成模块用于生成长度可变填充报文，其中，所述长度可变填充报文包括频偏补偿信息，当需要附加信息时，所述长度可变填充报文还包括附加信息，当不需要附加信息时，所述长度可变填充报文不包括附加信息，其中，所述附加信息包括控制信息，状态信息，请求信息以及通信信息的至少一种；发送模块用于发送所述长度可变填充报文。
在一些可能的设计中，所述长度可变填充报文还包括标识信息，当所述长度可变填充报文不需要附加信息时，所述标识信息为第一标识符，当所述长度可变填充报文需要附加信息时，所述标识信息为第二标识符。
在一些可能的设计中，所述第一标识符的标识符编码不同于所述第二标识符的标识符编码。
在一些可能的设计中，所述第一节点的第一时钟和所述第二节点的第二时钟并不相同，或者，所述第一节点的第一时钟和所述第二节点的第二时钟相同。
第三方面，提供了一种频偏消除实现装置，包括：接收模块以及确定模块。所述接收模块用于接收长度可变填充报文；所述确定模块用于在所述长度可变填充报文的标识信息为第一标识符的情况下，确定所述长度可变填充报文不包括附加信息，在所述长度可变填充报文的标识信息为第二标识符的情况下，确定所述长度可变填充报文包括附加信息。其中，所述附加信息包括控制信息，状态信息，请求信息以及通信信息的至少一种。
在一些可能的设计中，所述第一标识符的标识符编码不同于所述第二标识符的标识符编码。
在一些可能的设计中，需要附加信息时的长度可变填充报文的报文长度大于不要附加信息时的长度可变填充报文的报文长度。
第四方面，提供了一种通讯设备，所述通讯设备包括处理器和存储器；所述处理器用于执行所述存储器中存储的指令，以使得所述通讯设备执行如第一方面中第一节点或第二节点任一角度所述的方法。
第五方面，提供了一种包含指令的计算机程序产品，其特征在于，当所述指令被通讯设备运行时，使得所述通讯设备执行如第一方面中第一节点或第二节点任一角度所述的方法。
第六方面，提供了一种可读存储介质，包括程序指令，当所述程序指令由通讯设备执行时，所述通讯设备执行如第一方面中第一节点或者第二节点任一角度所述的方法。
第七方面，提供了一种频偏消除实现系统，包括第一节点以及第二节点，其中，所述第一节点和所述第二节点之间可以相互通信，所述第一节点用于执行如第一方面中第一节点角度的方法，所述第二节点用于执行如第一方面中第二节点角度所述的方法。
附图说明
图1是本申请涉及的两种时钟分别控制不同模块的示意图；
图2A至图2B是本申请涉及的发送端发送的填充报文以及接收端接收到的填充报文的示意图；
图3是本申请提供的一种频偏消除方法的流程示意图；
图4A至图4B是本申请提供的一些长度可变填充报文的结构示意图；
图5是本申请提供的一种通信系统的结构示意图；
图6是本申请提供的一种通信设备的结构示意图。
具体实施方式
为了便于理解本申请提供的技术方案，首先介绍本申请涉及的一些相关术语及应用场景。
计算机之间、计算机内部的各部分之间可以采用串行或并行的方式进行通信。串行通信(serial communication)是指在计算机总线或其他数据通道上，每次传输一个比特(bit)数据，并连续进行以上单次过程的通信方式，例如，快捷外围部件互连标准(peripheral component interconnect express,PCIe)、计算快速链接(compute express link,CXL)等。并行通信(parallel communication)是指多比特数据同时通过并行线进行传送的通信方式，例如，外围部件互连标准(peripheral component interconnect,PCI)。
PCIe是一种高速串行计算机扩展总线标准，属于PCI的一种。PCI是一种连接计算机主板和外部设备的总线标准，是目前个人电脑中使用最为广泛的接口，几乎所有的主板产品上都带有这种插槽。PCIe沿用了PCI的编程概念及通讯标准，但PCIe具有更强的抗干扰能力、更快的数据传输速率以及更好的兼容性等多种优点。
CXL是一种开放的互连标准。CXL技术基于成熟的PCIe基础设施，利用PCIe 5.0的物理和电子接口在三个关键领域提供高级协议：输入/输出协议、内存协议和一致性接口，这使得CXL能够为CPU和加速卡之间提供超快速互连以及更好的内存一致性，从而降低软件堆栈的复杂性和总体系统成本，进一步提高数据密集型应用程序的性能。
相较于并行通信，串行通信具有结构简单、信号线少、支持远距离的数据传输以及通信成本低等优点，因此，当前许多的计算机产品中都采用串行通信的方式来实现计算机之间、计算机内部的各部分之间的通信。
为了清楚地理解本申请提供的技术方案，首先结合图1对第一节点的发送端频偏与第二节点的接收端 的频偏不一致时，会出现数据丢失或资源浪费的原因进行说明：
如图1所示，标准时钟的时钟频率为100MHz，第一时钟的时钟频率为99.9997MHz/100.0003MHz，第二时钟的时钟频率为100.0003MHz/99.9997MHz。图中白色区域中的功能模块由第一时钟驱动，灰色区域中的功能模块由第二时钟驱动。
第一时钟驱动第一节点的第一端口的发送端110(以下简称为第一节点的发送端110)向第二端口的接收端120(以下简称为第二节点的接收端120)发送数据流，第二节点的接收端120利用时钟数据恢复(clock data recovery，CDR)模块130从第一节点的发送端110发送的数据流中恢复出与第一节点的发送端110同相位的时钟信号(即第一时钟产生的时钟信号)，然后利用这个时钟信号采集第一节点的发送端110后续发送的数据。因此，第二节点的接收端120是按照第一时钟的时钟频率来采集第一节点的发送端110发送的数据。第二节点的接收端120接收到数据后，先将数据写入第二节点的第二端口的弹性缓冲区(elastic buffer)140，之后，第二节点的第二端口的发送端150(以下简称为第二节点的发送端150)从弹性缓冲区140中读取数据，并发送给第二节点的第二端口的上层逻辑160，以便对数据进行后续的识别、处理等。
第二节点的接收端120是按照第一时钟的时钟频率将读取到的数据写入弹性缓冲区140，但是，第二节点的发送端150是按照第二时钟的时钟频率从弹性缓冲区140中读取数据。那么，当第一时钟的时钟频率与第二时钟的时钟频率不一致时，意味着第二节点的接收端120向弹性缓冲区140中写入数据的速度与第二节点的发送端150从弹性缓冲区140中读取数据的速度不同。如果第一时钟的时钟频率大于第二时钟的时钟频率，说明第二节点的接收端120向弹性缓冲区140写入数据的速度大于第二节点的发送端150从弹性缓冲区140中读取数据的速度，随着时间的推移，弹性缓冲区140就可能会出现溢出的现象，从而造成数据丢失。如果第一时钟的时钟频率小于第二时钟的时钟频率，说明第二节点的接收端120向弹性缓冲区140写入数据的速度小于第二节点的发送端150从弹性缓冲区140读取数据的速度，随着时间的推移，弹性缓冲区140中可能出现空闲状态，从而可能使得从弹性缓冲区读取数据时踩空，造成下溢。
因此，可以在第一节点的发送端110和第二节点的接收端120之间发送包含填充字段的填充报文。如图2A所示，第一节点的发送端110发送的包含填充字段的填充报文大概为40个字节(byte)，可以包括3个填充字段(body)、一个结束字段(end)以及一个控制字段(control)。其中，填充字段通常包括4个字节，用于承载填充码，填充码可以是于第一节点以及第二节点之间的通信无关的数据；结束字段通常包括4个字节，用于承载结束码；控制字段通常24个字节，用于承载控制码，控制码用于供第一节点对第二节点进行控制。第一节点的发送端发送的填充报文可以经过两个节点再到第二节点的接收端，每个节点可以在填充报文的基础上增加或者删除1个填充字段。因此，如图2B所示，第二节点的接收端接收到的填充报文可以是32字节、36字节、40字节、44字节或者48字节。例如，当两个节点都删除了一个填充字段时，第二节点的接收端接收到的填充报文可以是32字节；当两个节点中的其中一个删除了一个填充字段时，第二节点的接收端接收到的填充报文可以是36字节；当两个节点都不删除或者增加填充字段时，第二节点的接收端接收到的填充报文可以是40字节；当两个节点中的其中一个增加了一个填充字段时，第二节点的接收端接收到的填充报文可以是44字节；当两个节点都增加了一个填充字段时，第二节点的接收端接收到的填充报文可以是48字节。可以理解，上述例子中是以填充报文包括40个字节，包括3个填充字段，每个填充字段为4个字节，经过最多两个中间节点，并且，每个中间节点可以增加或者删除一个填充字段为例进行说明的，在实际应用中，填充报文可以包括更多或者更少的字节，可以包括更多或者更少的填充字段，经过更多或者更少的中间节点，每个填充字段可以包括更多或者更少的字节，每个中间节点可以增加或者删除更多的填充字段，填充报文的发送间隔可以是不相同的等等，此处不作具体限定。
当第一时钟的频率大于第二时钟的频率时，第二节点的接收端120向弹性缓冲区140写入数据的速度大于第二节点的发送端150从弹性缓冲区140中读取数据的速度，此时，可以增加填充报文中的填充字段，或者增加填充报文的数量，如此，第一节点的发送端110发送至第二节点的接收端120的数据中可以包含更多的与第一节点和第二节点的通信无关的数据，第二节点的接收端120可以将这些数据丢弃，从而减少需要缓冲在弹性缓冲区140中的数据，以避免弹性缓冲区140出现溢出状态。相反，当第一时钟的频率小于第二时钟的频率时，第二节点的接收端120向弹性缓冲区140写入数据的速度小于第二节点的发送端150从弹性缓冲区140中读取数据的速度，此时，可以减少填充报文中的填充字段，或者减少填充报文的数量，如此，第一节点的发送端110发送至第二节点的接收端120的数据中可以包含更少的与第一节点和第二节点的通信无关的数据，从而增加缓冲在弹性缓冲区140中的数据，以避免弹性缓冲区140出现空闲状态。
但是，链路两端可能始终存在频率差，因此需要一直发送填充报文提供消除频偏的功能，比如，每隔 640个字节就需要插入一个上述例子中的填充报文，则链路带宽损失为40/(40+640)＝5.88％。
本申请适用于串行通信的场景。具体地，在串行通信的场景中，通信双方(包括第一节点的发送端和第二节点的接收端)都需要被时钟驱动才能完成数据的传输。但是，当第一节点的发送端的频偏与第二节点的接收端的频偏不一致时，可能出现数据丢失或资源浪费的现象，因而需要发送填充报文进行调节。本申请中，第一节点的发送端的频偏是指第一节点的发送端的时钟频率与标准时钟的时钟频率的偏差，第二节点的接收端的频偏是指第二节点的接收端的时钟频率与标准时钟的时钟频率的频偏。
本申请提供了一种频偏消除实现方法，该方法适用于串行通信系统，即第一节点的发送端和第二节点的接收端采用串行的方式进行通信，并且，在不需要附加信息时，通过发送不包括附加信息的长度可变填充报文进行频偏消除，在需要附加信息时，通过发送包括附加信息的长度可变填充报文执行频偏消除以及附加信息。利用上述方法，能够减少链路带宽的损失。
参见图3，图3是本申请提供的一种频偏消除实现方法的流程示意图。如图3所示，本申请提供的频偏消除实现方法，包括：
S101：第二节点的发送端向第一节点的接收端发送报文。相应地，第一节点的接收端接收第二节点的发送端发送的报文。
在一种可能的实施方式中，报文用于指示第一节点发送填充报文进行频偏补偿。具体地，当第二节点中的弹性缓冲区接近溢出的时候，第二节点向第一节点发送用于指示第一节点增加填充报文的报文。当第二节点中的弹性缓冲区接近空闲的时候，第二节点向第一节点发送用于指示第一节点减少填充报文的报文。
S102：第一节点确定是否需要承载附加。如果不需要，进入步骤S103A，生成不包括附加信息的长度可变填充报文，如果需要，进入步骤S103B，生成包括附加信息的长度可变填充报文。
在一种可能的实施方式中，长度可变填充报文包括频偏补偿信息，当需要附加信息时，所述长度可变填充报文还包括附加信息，当不需要附加信息时，所述长度可变填充报文不包括附加信息。也就是，长度可变填充报文包括两种不同的形态，在需要附加信息时，长度可变填充报文包括频偏补偿信息以及附加信息；在不需要附加信息时，长度可变填充报文包括频偏补偿信息，不包括附加信息。
在一种可能的实施方式中，长度可变填充报文还包括标识信息，当所述长度可变填充报文不需要附加信息时，所述标识信息为第一标识符，当所述长度可变填充报文需要附加信息时，所述标识信息为第二标识符。
在一种可能的实施方式中，附加信息包括控制信息、请求信息、状态信息以及通讯信息中的至少一种。其中，控制信息是指第一节点对第二节点进行控制信息，例如，第一节点指示第二节点进行操作处理的信息。请求信息是指第一节点请求第二节点对第一节点进行控制的信息。状态信息为关于第一节点的状态的信息，例如，第一节点的位置、功耗等等。通讯信息可以包括其他需要进行通信的信息等等。
在一种可能的实施方式中，长度可变填充报文是按照周期进行发送的，在每个周期中，第一节点可以根据是否需要承载附加信息选择生成不同的长度可变填充报文。
S103A：第一节点生成不包括附加信息的长度可变填充报文。
在一种可能的实施方式中，长度可变填充报文包括频偏补偿信息，第一标识符，不包括附加信息。在一具体的实施方式中，如图4A所示，此形态下的长度可变填充报文可以包括：第一填充字段以及第一标识字段。其中，第一填充字段用于承载频偏补偿信息，所述频偏补偿信息可以是与第一节点和第二节点之间进行通信无关的信息。第一标识字段用于承载第一标识符，第一标识符用于识别本报文为不包括附加信息的可变填充报文。第一填充字段的长度可以是4字节，第一标识字段可以是4字节。当然，第一填充字段的长度还可以是更多或者更少的字节，第一标识字段的长度还可以是更多或者更少的字节，此处不作具体限定。
在一种可能的实施方式中，当报文指示第一节点增加填充报文，并且不需要附加信息时，增加生成的不包括附加信息的长度可变填充报文的数量，从而增加插入数据流中的不包括附加信息的长度可变填充报文的数量，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更多的与第二节点和第一节点的通信无关的数据，第二节点的接收端可以将这些数据丢弃，从而减少需要缓冲在弹性缓冲区中的数据，以避免弹性缓冲区出现溢出状态。当报文指示第一节点减少填充报文，并且不需要附加信息时，减少生成不包括附加信息的长度可变填充报文的数量，从而减少插入数据流中的不包括附加信息的长度可变填 充报文的数量，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更少的与第二节点和第一节点的通信无关的数据，以避免弹性缓冲区出现空闲状态。
在一种可能的实施方式中，当报文指示第一节点增加填充报文，并且不需要附加信息时，增加生成的不包括附加信息的长度可变填充报文的填充字段的数量，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更多的与第二节点和第一节点的通信无关的数据，第二节点的接收端可以将这些数据丢弃，从而减少需要缓冲在弹性缓冲区中的数据，以避免弹性缓冲区出现溢出状态。当报文指示第一节点减少填充报文，并且不需要附加信息时，减少生成的不包括附加信息的长度可变填充报文的填充字段的数量，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更少的与第二节点和第一节点的通信无关的数据，以避免弹性缓冲区出现空闲状态。
在一种可能的实施方式中，当报文指示第一节点增加填充报文，并且不需要附加信息时，缩短发送不包括附加信息的长度可变填充报文的周期，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更多的与第二节点和第一节点的通信无关的数据，第二节点的接收端可以将这些数据丢弃，从而减少需要缓冲在弹性缓冲区中的数据，以避免弹性缓冲区出现溢出状态。当报文指示第一节点减少填充报文，并且不需要进行附加额外功能时，增加发送不包括附加信息的长度可变填充报文的周期，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更少的与第二节点和第一节点的通信无关的数据，以避免弹性缓冲区出现空闲状态。
S104A：第一节点向第二节点发送不包括附加信息的长度可变填充报文。相应地，第二节点接收第一节点发送的不包括附加信息的长度可变填充报文。
S105A：第二节点识别不包括附加信息的长度可变填充报文，并执行频偏补偿。
在一种可能的实施方式中，第二节点可以根据长度可变填充报文中的标识信息为第一标识符识别本报文为不包括附加信息的长度可变填充报文，并执行频偏补偿。
S103B：第一节点生成包括附加信息的长度可变填充报文。
在一种可能的实施方式中，包括附加信息的长度可变填充报文包括频偏补偿信息、第二标识符以及附加信息。在一具体的实施方式中，如图4B所示，此形态下长度可变填充报文可以包括：第二填充字段、第二标识字段以及附加信息字段。其中，第二填充字段用于承载频偏补偿信息，所述频偏补偿信息是与第一节点和第二节点之间进行通信无关的信息。第二标识字段用于承载第二标识符，第二标识符用于识别本报文为包括附加信息的长度可变填充报文。附加信息字段用于承载附加信息。附加信息包括控制功信息能、请求信息、状态信息以及通讯信息中的至少一种。第二填充字段的长度可以是4字节，第二标识字段可以是4字节，附加信息字段可以是24字节。当然，第二填充字段的长度还可以是更多或者更少的字节，第二标识字段的长度还可以是更多或者更少的字节，附加信息字段也可以是更多或者更少字节，此处不作具体限定。
在一种可能的实施方式中，当报文指示第一节点增加填充报文，并且需要附加信息时，第一节点的发送端增加生成包括附加信息的长度可变填充报文的数量，从而增加插入数据流中的包括附加信息的长度可变填充报文的数量，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更多的与第二节点和第一节点的通信无关的数据，第二节点的接收端可以将这些数据丢弃，从而减少需要缓冲在弹性缓冲区中的数据，以避免弹性缓冲区出现溢出状态。当报文指示第一节点减少填充报文，并且需要附加信息时，第一节点的发送端减少生成包括附加信息的长度可变填充报文的数量，从而减少插入数据流中的包括附加信息的长度可变填充报文的数量，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更少的与第二节点和第一节点的通信无关的数据，从而增加需要缓冲在弹性缓冲区中的数据，以避免弹性缓冲区出现空闲状态。
在一种可能的实施方式中，当报文指示第一节点增加填充报文，并且需要附加信息时，第一节点的发送端增加生成的包括附加信息的长度可变填充报文的填充字段的数量，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更多的与第二节点和第一节点的通信无关的数据，第二节点的接收端可以将这些数据丢弃，从而减少需要缓冲在弹性缓冲区中的数据，以避免弹性缓冲区出现溢出状态。当报文指示第一节点减少填充报文，并且需要附加信息时，第一节点的发送端减少生成的包括附加信息的长度可变填充报文中的填充字段的数量，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更少的与第二节点和第一节点的通信无关的数据，从而增加需要缓冲在弹性缓冲区中的数据，以避免弹性缓冲区出现空闲状态。
在一种可能的实施方式中，当报文指示第一节点增加填充报文，并且需要进行附加额外功能时，减少 发送包括附加信息的长度可变填充报文的周期，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更多的与第二节点和第一节点的通信无关的数据，第二节点的接收端可以将这些数据丢弃，从而减少需要缓冲在弹性缓冲区中的数据，以避免弹性缓冲区出现溢出状态。当报文指示第一节点减少填充报文，并且需要附加信息时，增加发送包括附加信息的长度可变填充报文的周期，如此，第一节点的发送端发送至第二节点的接收端的数据中可以包含更少的与第二节点和第一节点的通信无关的数据，从而增加需要缓冲在弹性缓冲区中的数据，以避免弹性缓冲区出现空闲状态。
S104B：第一节点向第二节点发送包括附加信息的长度可变填充报文。相应地，第二节点接收第一节点发送的包括附加信息的长度可变填充报文。
S105B：第二节点识别包括附加信息的长度可变填充报文，并执行频偏补偿和附加信息对应的功能。
在一种可能的实施方式中，第二节点可以根据长度可变填充报文中的标识信息为第二标识符识别本报文为包括附加信息的长度可变填充报文，并执行频偏补偿和附加信息对应的功能。
其中，第一标识符的标识符编码不同于第二标识符的标识符编码，不包括附加信息的长度可变填充报文的报文长度少于包括附加信息的长度可变填充报文的报文长度。
应理解，上述例子中是以第一节点的第一时钟和第二节点的第二时钟并不相同为例进行说明的，在实际应用中，第一节点的第一时钟和第二节点的第二时钟也可以是相同的，此时，第一节点向第二节点发送填充报文的周期可以变得更长，以减少不必要的带宽消耗，同时又可以在需要时携带额外信息。
在上述方案中，在链路不需要附加信息的时候，第一节点的发送端发送不包括附加信息的长度可变填充报文进行频偏消除，此时链路损耗是：16/656＝2.44％；在链路需要附加信息的时候，第一节点的发送端发送包括附加信息的长度可变填充报文，此时链路损耗是：40/680＝5.88％。因此，比起链路一直发送包括附加信息的长度可变填充报文，能够有效地减少链路的带宽损失。
前述内容详细介绍了本申请提供的通信系统中第二节点(即接收端)和第一节点(即发送端)之间的频偏消除实现方法，为了更好地实施上述方法，接下来介绍本申请提供的用于配合实施上述方法的相关装置、设备及系统。
如图5所示，图5示出了本申请提供的一种通信系统的结构示意图。如图5所示，通信系统可以包括装置310以及装置320。装置310可以部署在上述第一节点中，装置320可以部署在上述第二节点中，从而能够调节第二节点和第一节点的时钟偏差，并发送附加信息。装置310包括生成单元311、发送单元312。
所述生成单元311用于生成长度可变填充报文，其中，所述长度可变填充报文包括频偏补偿信息，当需要附加信息时，所述长度可变填充报文还包括附加信息，当不需要附加信息时，所述长度可变填充报文不包括附加信息，其中，所述附加信息包括控制信息，状态信息，请求信息以及通信信息的至少一种。
所述发送单元312用于发送所述长度可变填充报文。
装置320包括接收单元321、确定单元322。
接收单元321，用于接收端接收长度可变填充报文。
确定单元322，用于当所述长度可变填充报文的标识信息为第一标识符时，确定所述长度可变填充报文不包括附加信息，当所述长度可变填充报文的标识信息为第二标识符时，确定所述长度可变填充报文包括附加信息，其中，所述附加信息包括控制信息，状态信息，请求信息以及通信信息的至少一种。
上述实施例提供的装置310与上述方法实施例中的第一节点属于同一构思，例如，生成单元311用于执行前述S103A或者103B、发送单元312用于执行前述S104A或者S104B中的发送部分，上述实施例提供的装置320与上述方法实施例中的第二节点属于同一构思，例如，接收单元321用于执行前述S104A或者S104B中的接收部分，确定单元322用于执行步骤S105A或者S105B，具体实现过程详见上述方法实施例，这里不再赘述。
本申请实施例提供的通信系统中的装置仅以上述各功能单元的划分进行举例说明，实际应用中，可以根据需要而将上述功能分配由不同的功能单元完成，即将上述装置310的内部结构划分成不同的功能单元，以完成以上描述的全部或者部分功能。
本申请还提供了一种通信设备，如图6所示，图6示出了本申请提供的一种通信设备的结构示意图。其中，通信设备400包括通信端口410，通信端口410包括发送端411、接收端412、弹性缓冲区413。通信端口410与上述方法实施例中的第一节点属于同一构思，通信端口410具体用于执行前述S101、S102、S103A、S104A以及S103B，S104B以实现通信端口410与其他端口(例如，第二节点的通信端口)之间的 串行通信。或者，通信端口410与上述方法实施例中的第二节点属于同一构思，通信端口410具体用于执行前述步骤S101、S104A、S105A以及S104B、S105B。
本申请还提供了一种通信系统，通信系统可以包括第二节点和第一节点，第二节点包括接收端以及缓冲区，第一节点包括发送端。可选的，第一节点也可以包括缓冲区。
其中，第一节点用于执行前述S101、S102、S103A、S104A以及S103B，S104B中的第一节点的角度步骤：具体地，第一节点的发送端用于执行前述S101、S102、S103A、S104A以及S103B，S104B中第一节点的角度的方法，第二节点的接收端用于执行前述步骤S101、S104A、S105A以及S104B、S105B中第二节点的角度的方法。
上述各个附图对应的流程的描述各有侧重，某个流程中没有详述的部分，可以参见其他流程的相关描述。
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机(计算设备)上加载和执行所述计算机指令时，全部或部分地产生按照本申请实施例所述的流程或功能。所述计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，所述计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如，同轴电缆、光纤、数字用户线或无线(例如，红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质，(例如，软盘、硬盘、磁带)、光介质(例如，DVD)、或者半导体介质(例如，SSD)等。

Claims (13)

  1. 一种频偏消除的实现方法，其特征在于，包括：
    第一节点的发送端生成长度可变填充报文，其中，所述长度可变填充报文包括频偏补偿信息，当需要附加信息时，所述长度可变填充报文还包括附加信息，当不需要附加信息时，所述长度可变填充报文不包括附加信息，其中，所述附加信息包括控制信息，状态信息，请求信息以及通信信息的至少一种；
    所述第一节点的发送端发送所述长度可变填充报文。
  2. 根据权利要求1所述的方法，其特征在于，所述长度可变填充报文还包括标识信息，当所述长度可变填充报文不需要附加信息时，所述标识信息为第一标识符，当所述长度可变填充报文需要附加信息时，所述标识信息为第二标识符。
  3. 根据权利要求2所述的方法，其特征在于，所述第一标识符的标识符编码不同于所述第二标识符的标识符编码。
  4. 根据权利要求1至3任一权利要求所述的方法，其特征在于，需要附加信息时的长度可变填充报文的报文长度大于不要附加信息时的长度可变填充报文的报文长度。
  5. 根据权利要求1至4任一权利要求所述的方法，其特征在于，所述第一节点的第一时钟和所述第二节点的第二时钟并不相同，或者，所述第一节点的第一时钟和所述第二节点的第二时钟相同。
  6. 一种频偏消除的实现方法，其特征在于，包括：
    第二节点的接收端接收长度可变填充报文；
    当所述长度可变填充报文的标识信息为第一标识符时，所述第二节点的接收端确定所述长度可变填充报文不包括附加信息，当所述长度可变填充报文的标识信息为第二标识符时，所述第二节点的接收端确定所述长度可变填充报文包括附加信息，其中，所述附加信息包括控制信息，状态信息，请求信息以及通信信息的至少一种。
  7. 根据权利要求6所述的方法，其特征在于，所述第一标识符的标识符编码不同于所述第二标识符的标识符编码。
  8. 根据权利要求5或6所述的方法，其特征在于，需要附加信息时的长度可变填充报文的报文长度大于不要附加信息时的长度可变填充报文的报文长度。
  9. 根据权利要求6至8任一权利要求所述的方法，其特征在于，所述第一节点的第一时钟和所述第二节点的第二时钟并不相同，或者，所述第一节点的第一时钟和所述第二节点的第二时钟相同。
  10. 一种频偏消除装置，其特征在于，包括能够执行如权利要求1至9任一方法的步骤的模块。
  11. 一种通讯设备，其特征在于，所述通讯设备包括处理器和存储器；所述处理器用于执行所述存储器中存储的指令，以使得所述通讯设备执行如权利要求1至9任一权利要求所述的方法。
  12. 一种可读存储介质，其特征在于，包括程序指令，当所述程序指令由通讯设备执行时，所述通讯设备执行如权利要求1至9任一权利要求所述的方法。
  13. 一种频偏消除实现系统，其特征在于，包括第一节点以及第二节点，其中，所述第一节点和所述第二节点之间可以相互通信，所述第一节点用于执行如权利要求1至5任一权利要求所述的方法，所述第二节点用于执行如权利要求6至9任一权利要求所述的方法。
PCT/CN2023/115261 2022-09-07 2023-08-28 一种频偏消除实现方法、设备以及系统 WO2024051511A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211098687.8 2022-09-07
CN202211098687.8A CN117714248A (zh) 2022-09-07 2022-09-07 一种频偏消除实现方法、设备以及系统

Publications (1)

Publication Number Publication Date
WO2024051511A1 true WO2024051511A1 (zh) 2024-03-14

Family

ID=90097433

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/115261 WO2024051511A1 (zh) 2022-09-07 2023-08-28 一种频偏消除实现方法、设备以及系统

Country Status (2)

Country Link
CN (2) CN117675465A (zh)
WO (1) WO2024051511A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170163744A1 (en) * 2010-06-27 2017-06-08 Valens Semiconductor Ltd. Network for transporting Ethernet and time sensitive data
CN113475112A (zh) * 2019-02-28 2021-10-01 高通股份有限公司 用于无线通信网络的聚集式控制信息
CN113993009A (zh) * 2020-07-27 2022-01-28 上海诺基亚贝尔股份有限公司 时间同步用于多域传输的方法、装置、olt及onu
CN114097298A (zh) * 2019-07-01 2022-02-25 高通股份有限公司 用于无线局域网(wlan)中的多链路通信的信令

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115336217A (zh) * 2020-03-14 2022-11-11 韦勒斯标准与技术协会公司 无线通信系统中发送或接收数据的无线通信终端以及方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170163744A1 (en) * 2010-06-27 2017-06-08 Valens Semiconductor Ltd. Network for transporting Ethernet and time sensitive data
CN113475112A (zh) * 2019-02-28 2021-10-01 高通股份有限公司 用于无线通信网络的聚集式控制信息
CN114097298A (zh) * 2019-07-01 2022-02-25 高通股份有限公司 用于无线局域网(wlan)中的多链路通信的信令
CN113993009A (zh) * 2020-07-27 2022-01-28 上海诺基亚贝尔股份有限公司 时间同步用于多域传输的方法、装置、olt及onu

Also Published As

Publication number Publication date
CN117675465A (zh) 2024-03-08
CN117714248A (zh) 2024-03-15

Similar Documents

Publication Publication Date Title
US10552352B2 (en) Methods and apparatus for synchronizing uplink and downlink transactions on an inter-device communication link
US7898956B2 (en) Credit-based rate control for high-speed interfaces
CN109857690B (zh) 驱动器的应用系统、驱动器和数据传输方法
WO2020001365A1 (zh) 一种FlexE组中PHY的调整方法、相关设备及存储介质
US20200259936A1 (en) Device-to-device link training
CN101409675A (zh) 网络分组净荷压缩
KR101820826B1 (ko) 복수의 통신 채널들 사이의 직렬 통신 포트를 공유하기 위한 시스템들 및 방법들
JP2007517334A (ja) ポイントツーポイントリンクを処理する非データ符号を介したバッファ管理
WO2021147044A1 (zh) 一种时间敏感网络时间同步方法及装置
CN108614792B (zh) 1394事务层数据包存储管理方法及电路
WO2024051511A1 (zh) 一种频偏消除实现方法、设备以及系统
CN106494466B (zh) 列车运行监控的数据换装装置及其数据换装方法
CN115150383A (zh) 文件发送方法、设备及计算机可读存储介质
CN102035733B (zh) 通过以太网建立串行数据透明传输通道的方法
TWI446766B (zh) 管理執行之方法及具有執行管理能力之裝置
CN111404842A (zh) 数据传输方法、装置及计算机存储介质
WO2021189267A1 (zh) 一种时钟展频协商方法、高速外围组件互联设备及系统
WO2015024477A1 (zh) 一种路由更新的实现方法及装置
WO2022021409A1 (zh) 灵活以太网组中物理接口调整方法和设备
WO2024198573A1 (zh) 一种流量控制方法、装置及系统
JP2014222466A (ja) 情報処理装置、情報処理システムおよび情報処理システムの通信方法
CN118368293B (zh) 一种数据传输方法、计算机设备及介质
CN110601784B (zh) Tdm接口扩展方法及装置、设备、可读存储介质
WO2024160157A1 (zh) 补偿链路两端时钟频率偏差的方法及通信端口
US12069513B2 (en) Aggregating multiple data streams on a communication link

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23862207

Country of ref document: EP

Kind code of ref document: A1