WO2018224052A1 - 一种轨道交通领域的具有容错功能的实时数据调度方法 - Google Patents

一种轨道交通领域的具有容错功能的实时数据调度方法 Download PDF

Info

Publication number
WO2018224052A1
WO2018224052A1 PCT/CN2018/095177 CN2018095177W WO2018224052A1 WO 2018224052 A1 WO2018224052 A1 WO 2018224052A1 CN 2018095177 W CN2018095177 W CN 2018095177W WO 2018224052 A1 WO2018224052 A1 WO 2018224052A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
real
time
module
fault tolerance
Prior art date
Application number
PCT/CN2018/095177
Other languages
English (en)
French (fr)
Inventor
万海
于庆涵
李星
徐燕芬
Original Assignee
清华大学
中车信息技术有限公司
中车青岛四方车辆研究所有限公司
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 清华大学, 中车信息技术有限公司, 中车青岛四方车辆研究所有限公司 filed Critical 清华大学
Publication of WO2018224052A1 publication Critical patent/WO2018224052A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
    • G06F11/0706—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment
    • G06F11/0736—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment in functional embedded systems, i.e. in a data processing system designed as a combination of hardware and software dedicated to performing a certain function
    • G06F11/0739—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment in functional embedded systems, i.e. in a data processing system designed as a combination of hardware and software dedicated to performing a certain function in a data processing system embedded in automotive or aircraft systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/10—Active monitoring, e.g. heartbeat, ping or trace-route
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00—Packet switching elements
    • H04L49/10—Packet switching elements characterised by the switching fabric construction
    • H04L49/112—Switch control, e.g. arbitration
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00—Packet switching elements
    • H04L49/30—Peripheral units, e.g. input or output ports
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00—Packet switching elements
    • H04L49/40—Constructional details, e.g. power supply, mechanical construction or backplane
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00—Packet switching elements
    • H04L49/55—Prevention, detection or correction of errors
    • H04L49/557—Error correction, e.g. fault recovery or fault tolerance

Definitions

  • the invention relates to a real-time data scheduling method based on Ethernet technology in the field of rail transit, in particular to a real-time data scheduling method with fault tolerance capability in the field of rail transit.
  • Industrial Ethernet refers to the application of Ethernet related components and technologies in the automation control and program control of industrial environments, providing deterministic and real-time performance for industrial networks.
  • Industrial Ethernet is compatible with the IEEE 802.3 standard, and has the characteristics of fast communication speed, wide application, low development and production cost, and easy integration. It is widely used in aviation, railway transportation, industrial manufacturing and other fields.
  • the real-time Ethernet switch is the backbone of an Ethernet-based real-time communication network, and its interconnection with the terminal devices constitutes the entire Ethernet topology.
  • the real-time data scheduling method is responsible for forwarding data generated by the terminal in the network and ensuring real-time data in real time.
  • the real-time nature of the data means that the data needs to reach the destination terminal within the specified time.
  • the data with real-time requirements is called real-time data.
  • the real-time data scheduling method is the key to ensuring the real-time performance of industrial Ethernet.
  • the existing real-time data scheduling methods in China mainly focus on guaranteeing real-time functions and cannot tolerate switch failure.
  • the faulty switch is short-circuited, causing the real-time data that has passed through the faulty switch to arrive at the next switch earlier, thereby causing the data to not match the corresponding receiving time window, thereby causing network packet loss.
  • the loss of real-time data can cause problems such as brake failure, which can have serious consequences. Therefore, it is necessary to invent a real-time data scheduling method with fault tolerance in the field of rail transit.
  • the technical solution of the present invention provides a real-time data scheduling method with fault tolerance in the field of rail transit, which is characterized in that it comprises the following steps:
  • Step 1 Initialize the real-time Ethernet switch system with fault tolerance; after the hardware starts, initialize the fault-tolerant module, clear the data in the storage resource pool, and set the value of the address list to 1;
  • Step 2 Monitor the switch failure; the Linux system on the CPU board monitors whether the surrounding switch is faulty through the heartbeat packet. When a port of the switch does not receive the specified heartbeat packet, the CPU determines that the switch connected to the corresponding port is faulty. When no switch failure is detected, the arbitration module directly transmits the data frame to the real-time scheduling module;
  • Step 3 In response to the switch failure, when the CPU board detects a fault of the surrounding switch, the CPU sends an instruction to open the fault tolerance function to the switch board, thereby opening a fault tolerance function of connecting the faulty switch port;
  • Step 4 Determine whether the arbitration module determines the flow direction of the data packets from each port according to whether the fault tolerance function of each port is turned on; if the fault tolerance function of the port is in the closed state, the arbitration module forwards the data packet from the port to the real-time scheduling module; The fault tolerance function is turned on, and the arbitration module forwards the data packet from the port to the storage resource allocation module;
  • Step 5 When the receiving window of the data packet is reached, the control logic module determines the address space of the data packet according to the data stream identification field; if the address space is empty, the control logic module switches to the retrieval operation of the next data packet; Otherwise, the control logic module retrieves the data in the changed address space and then sets the value of the address list corresponding to the address space to zero.
  • the invention also provides a real-time Ethernet switch system with real-time data scheduling method with fault tolerance function of rail transit, comprising a CPU board, a switch board and a PHY board;
  • the CPU board exchanges control information and data information with the switch board through the PCI-e interface, and the switch board exchanges data information through the interface and the PHY version;
  • the CPU board includes a central processing unit, a storage unit, and a PCI-e interface;
  • the real-time Linux system runs on the CPU board, and data is exchanged with the switch board through the PCI-e interface;
  • the switch board is a hardware board that implements a real-time scheduling algorithm with fault tolerance
  • the PHY board contains the PHY ASIC chip, which is the physical interface transceiver responsible for implementing the physical layer.
  • the invention has the beneficial effects that the invention makes it possible to prevent the real-time data passing through the faulty switch from reaching the next switch early when the switch in which the train fails is short-circuited, thereby causing the data to not match the corresponding receiving time window, and thus The problem that caused network packet loss.
  • FIG. 1 is a schematic diagram of a real-time data scheduling system with fault tolerance in the field of rail transit according to the present invention
  • FIG. 2 is a diagram showing an example of the internal structure of a two-port switch card according to the present invention.
  • FIG. 3 is a diagram showing an example of a structure of a four-way fault-tolerant module according to the present invention.
  • FIG. 4 is a flow chart of a fault tolerant module program of a real-time data scheduling method with fault tolerance function in the field of rail transit according to the present invention
  • this embodiment provides a real-time data scheduling method with fault tolerance in the field of rail transit, which specifically includes the following steps:
  • Step 1 Initialize a real-time Ethernet switch system with fault tolerance.
  • Step 2 Monitor the switch failure.
  • the Linux system on the CPU board monitors whether the surrounding switch is faulty through the heartbeat packet. When a port of the switch does not receive the specified heartbeat packet, the CPU determines that the switch connected to the corresponding port is faulty; when no switch failure is detected.
  • the arbitration module directly transmits the data frame to the real-time scheduling module;
  • Step 3 In response to the switch failure, when the CPU board detects the failure of the surrounding switch, the CPU sends an instruction to open the fault tolerance function to the switch board, thereby opening the fault tolerance function of connecting the faulty switch port.
  • the instruction contains a 32-bit value that corresponds to 1-32 ports from low to high.
  • the position of the command value corresponding to the port is set to 1.
  • the position of the corresponding command value is set to 0.
  • Step 4 The arbitration module determines the flow of data packets from each port according to whether the fault tolerance function of each port is turned on. If the fault tolerance function of the port is off, the arbitration module forwards the packet from the port to the real-time scheduling module; if the fault tolerance function of the port is enabled, the arbitration module forwards the packet from the port to the storage resource allocation module.
  • the storage resource allocation module allocates a storage resource pool address space for the real-time data packet according to the data flow identification field. If the data has been stored in the address space corresponding to the data flow flag field, the storage resource allocation module clears the data of the address space, sets the value in the corresponding address list to 0, and then re-addresses the real-time data. Allocation; if the address space corresponding to the data flow flag field is empty, the storage resource allocation module stores the data in the address space and sets the value of its corresponding address list to 1.
  • Step 5 When the receiving window of the data packet is reached, the control logic module determines the address space of the data packet according to the data flow identification field. If the address space is empty, the control logic module switches to the fetch operation of the next data packet; otherwise, the control logic module retrieves the data in the changed address space, and then sets the value of the address list corresponding to the address space to 0.
  • control logic module prefetches the data packet, that is, the data is taken out in advance for a certain time in a given receiving time window.
  • the embodiment also provides a real-time data scheduling method with fault tolerance function of rail transit, which is implemented by a real-time Ethernet switch system with fault tolerance function, wherein the real-time Ethernet switch system with fault tolerance includes a CPU board. , switch board and PHY board.
  • the CPU board exchanges control information and data information with the switch board through the PCI-e interface, and the switch board exchanges data information with the PHY version through the S3MII interface.
  • the CPU board includes a central processing unit (CPU) of a PowerPC architecture, a storage unit, and a PCI-e interface.
  • the real-time Linux system runs on the CPU board, and data is exchanged with the switch board through the PCI-e interface.
  • the switch board is the core hardware board that implements a real-time scheduling algorithm with fault tolerance, including physical devices and digital circuits.
  • the physical components of the switch board include Xilinx vertex 7 chip, Flash Rom, QDR external memory, etc.
  • the digital circuit of the switch board is designed by FPGA and burned into the Xilinx chip.
  • the PHY board contains Broadcom's PHY ASIC chip, which is the physical interface transceiver responsible for implementing the physical layer.
  • FIG 2 is an internal structure diagram of the switchboard of the present invention, which controls the physical device to implement forwarding of common data frames, ensures real-time transmission of real-time data frames, and provides fault tolerance.
  • the internal structure of the switch card includes MAC, classifier, normal data cache, clock data cache, real-time data channel, common data channel, multiplexer and PCI-E interface, among which:
  • the MAC is the interface for data and is responsible for sending and receiving data frames in the Ethernet.
  • the classifier distinguishes the type of the Ethernet data frame according to the EtherType field of the Ethernet frame data.
  • 0x88D7 is a real-time data type
  • 0x8801 is a clock data type.
  • the types of data frames are normal data frames, real-time data frames, and temporal data frames.
  • Real-time data frames are transmitted directly to the real-time data channel; normal data frames are stored in the normal data cache; time-based data is stored in the clock data cache.
  • the normal data cache is responsible for storing the normal data frames passed in by the classifier.
  • the clock data cache is responsible for storing the time-based data frames passed in by the classifier.
  • the real-time data channel is responsible for forwarding real-time data and providing fault tolerance.
  • the fault-tolerant module is responsible for implementing the fault tolerance function of the switch;
  • the real-time scheduling module is responsible for scheduling real-time data frames to ensure real-time transmission of data frames;
  • the clock synchronization module is responsible for providing a synchronized global clock for devices of the entire network.
  • the normal data channel is responsible for implementing the forwarding function of ordinary data frames.
  • the multiplexer is responsible for selecting the data frame that should be sent based on the priority.
  • the real-time data frame has the highest priority, and the normal data frame has the lowest priority.
  • FIG. 3 is a structural diagram of a fault tolerant module of the present invention, wherein the module for implementing the fault tolerance function includes an arbitration module, a storage resource allocation logic module, a storage resource pool, and a control logic module.
  • the function of the arbitration module is to determine the flow of data into the fault tolerant module.
  • the arbitration module responds to commands from the central processor to turn fault tolerance on or off.
  • the arbitration module transmits the data to the storage resource allocation module.
  • the arbitration module transmits the data directly to the real-time data scheduling module.
  • the storage resource allocation module is responsible for allocating its location in the storage resource pool. In real-time data, each data stream has a specific field for identification, and the field value is 1-4096. The module determines the storage location of the data stream in the storage resource pool by parsing the identity field.
  • the storage resource pool is responsible for caching real-time data that needs to be fault tolerant.
  • the storage resource pool in the present invention is implemented by an external storage device QDR.
  • the address space in the storage pool is 1-4096, and each space can store only one Ethernet longest frame.
  • the storage address space has a one-to-one correspondence with the data flow identification field.
  • the storage resource pool maintains a 4096-bit address list for recording whether data has been stored in the 1-4096 address space. If the address space stores data, the address bit is identified as 1 in the list, otherwise the flag is 0;
  • the control logic module is responsible for retrieving data from the storage resource pool at a specified time.
  • Each real-time data corresponds to an entry of a schedule in which the reception time window and the transmission time of the data are recorded.
  • Each real-time data should arrive in a defined time window and be sent out at a specified point in time.
  • the module acquires the data stream identification field and obtains the data from the storage resource pool.
  • the module first queries the address list. If the address list value of the corresponding address of the data flow field is 1, the module acquires data from the storage resource pool, and then transmits the data to the real-time scheduling module. Otherwise, the module does not respond to the retrieval request for this data.
  • the data is received by the mac module and arrives at the classifier.
  • the classifier forwards the real-time data with a field value of 0x88D7 to the fault-tolerant module by parsing the EtherType field of the Ethernet data.
  • the real-time data first passes through the arbitration module. If the input port fault tolerance function corresponding to the real-time data is closed, the arbitration module forwards to the real-time scheduling module; if the input port fault tolerance function corresponding to the real-time data is turned on, the arbitration module forwards to the storage resource allocation. In the module.
  • the storage resource allocation module allocates storage space in the storage resource pool for the real-time data according to the data identification field.
  • the real-time data is taken out from the storage resource pool and sent to the real-time scheduling module.
  • the fault tolerance function of the port is opened, the real-time data is cached in the resource storage pool due to the data advance, and is taken out at the specified receiving time, so the real-time data is not discarded because it arrives at the real-time scheduling module in advance, thereby realizing the fault tolerance function.

Abstract

本发明提供了一种轨道交通领域的具有容错功能的实时数据调度方法，其特征在于：其特征在于包括以下步骤：步骤1、对具有容错功能的实时以太网交换机系统进行初始化；步骤2、监测交换机故障；步骤3、响应交换机故障，当CPU板检测到周围交换机故障时，CPU向交换板发送打开容错功能的指令，从而打开连接故障交换机端口的容错功能；步骤4、判断仲裁模块根据各端口的容错功能是否打开；步骤5、当到达数据包的接收窗口时，控制逻辑模块根据数据流标识字段，确定该数据包的地址空间，本发明能够有效防止原本经过该故障交换机的实时数据提早到达下一个交换机，从而造成数据与其对应的接收时间窗口不匹配，进而导致网络丢包的问题。

Description

一种轨道交通领域的具有容错功能的实时数据调度方法
相关申请
本申请主张于2017年6月6日提交的、名称为“一种轨道交通领域的具有容错功能的实时数据调度方法”的中国发明专利申请：201710417001.X的优先权。
技术领域
本发明涉及一种轨道交通领域的基于以太网技术的实时数据调度方法，尤其涉及一种轨道交通领域的具有容错能力的实时数据调度方法。
背景技术
工业以太网是指在工业环境的自动化控制及程序控制中应用以太网的相关元件及技术，为工业网络提供确定性和实时性。工业以太网与IEEE 802.3标准兼容，并且具有通信速度快、应用广泛、开发和生产成本低、易于集成的特点，被广泛应用于航空、铁路交通、工业制造等领域。
实时以太网交换机是基于以太网的实时通信网络的骨架，其与终端设备的相互连接构成了整个以太网拓扑结构。实时数据调度方法负责对网络中的由终端产生的数据进行转发，并且保证实时数据的实时性。数据的实时性是指确保数据需要在规定的时间内到达目的终端，具有实时性要求的数据称为实时数据。实时数据调度方法是保证工业以太网实时性的关键。
国内现有的实时数据调度方法主要集中在保障实时功能上，并不能容忍交换机发生故障。在列车网络中，发生故障的交换机被短接，导致原本经过该故障交换机的实时数据提早到达下一个交换机，从而造成数据与其对应的接收时间窗口不匹配，进而导致网络丢包问题。列车上，实时数据的丢失会导致例如刹车失灵等问题，进而造成严重的后果。因此，发明一种轨道交通领域的具有容错功能的实时数据调度方法十分必要。
发明内容
本发明的目的是提供一种能够容忍交换机发生故障的实时数据调度方法。
本发明的技术方案在于提供了一种轨道交通领域的具有容错功能的实时数据调 度方法，其特征在于：其特征在于包括以下步骤：
步骤1、对具有容错功能的实时以太网交换机系统进行初始化；硬件启动后，对容错模块进行初始化，清空存储资源池中的数据，将地址列表的值都置为1；
步骤2、监测交换机故障；CPU板上搭载的Linux系统，通过心跳数据包监测周围交换机是否发生故障，当交换机的某端口接收不到指定的心跳数据包时，CPU判定对应端口连接的交换机发生故障；当没有检测到交换机故障时，仲裁模块直接将数据帧传输到实时调度模块中；
步骤3、响应交换机故障，当CPU板检测到周围交换机故障时，CPU向交换板发送打开容错功能的指令，从而打开连接故障交换机端口的容错功能；
步骤4、判断仲裁模块根据各端口的容错功能是否打开，决定来自各端口数据包的流向；如果端口的容错功能处于关闭状态，仲裁模块将来自该端口的数据包转发到实时调度模块；如果端口的容错功能处于开启状态，仲裁模块将来自该端口的数据包转发到存储资源分配模块；
步骤5、当到达数据包的接收窗口时，控制逻辑模块根据数据流标识字段，确定该数据包的地址空间；如果该地址空间为空，控制逻辑模块切换到下一个数据包的取回操作；否则，控制逻辑模块取回改地址空间中的数据，然后将该地址空间对应的地址列表的值设置为0。
本发明还提了一种用轨道交通的具有容错功能的实时数据调度方法的实时以太网交换机系统，包括CPU板、交换板和PHY板；其特征在于：
CPU板通过PCI-e接口与交换板进行控制信息和数据信息的交互，交换板通过接口与PHY版进行数据信息的交互；
CPU板包括中央处理器、存储单元和PCI-e接口；
CPU板上运行实时Linux系统，并且通过PCI-e接口与交换板进行数据的交换；
交换板是实现具有容错功能的实时调度算法的硬件板；
PHY板上包含PHY ASIC芯片，是物理接口收发器，负责实现物理层。
本发明的有益效果在于：本发明使得在列车发生故障的交换机被短接时，有效防 止原本经过该故障交换机的实时数据提早到达下一个交换机，从而造成数据与其对应的接收时间窗口不匹配，进而导致网络丢包的问题。
附图说明
图1是本发明所述的轨道交通领域的具有容错功能的实时数据调度系统的原理图；
图2本发明所述的两口交换卡内部结构示例图；
图3本发明所述的四路容错模块结构示例图；
图4本发明本发明所述的轨道交通领域的具有容错功能的实时数据调度方法的容错模块程序流程图；
具体实施方式
下面结合附图1-4对本发明的技术方案进行进一步描述。
如图4所示，该实施例提供了一种轨道交通领域的具有容错功能的实时数据调度方法，其具体包括以下步骤：
步骤1、对具有容错功能的实时以太网交换机系统进行初始化；
硬件启动后，对容错模块进行初始化，清空存储资源池中的数据，将地址列表的值都置为1；
步骤2、监测交换机故障；
CPU板上搭载的Linux系统，通过心跳数据包监测周围交换机是否发生故障，当交换机的某端口接收不到指定的心跳数据包时，CPU判定对应端口连接的交换机发生故障；当没有检测到交换机故障时，仲裁模块直接将数据帧传输到实时调度模块中；
步骤3、响应交换机故障，当CPU板检测到周围交换机故障时，CPU向交换板发送打开容错功能的指令，从而打开连接故障交换机端口的容错功能。
该指令包含一个32位的值，从低位到高位分别对应1-32个端口。当端口打开容错功能时，该端口对应的指令值的位置设为1；当端口关闭容错功能时，该端口对应的指令值的位置设为0。
步骤4、仲裁模块根据各端口的容错功能是否打开，决定来自各端口数据包的流 向。如果端口的容错功能处于关闭状态，仲裁模块将来自该端口的数据包转发到实时调度模块；如果端口的容错功能处于开启状态，仲裁模块将来自该端口的数据包转发到存储资源分配模块。
存储资源分配模块根据数据流标识字段，为实时数据包分配存储资源池地址空间。如果该数据流标志字段对应的地址空间中已经存储数据，存储资源分配模块清空该地址空间的数据，并将其对应的地址列表中的值设置为0，然后重新对该实时数据进行地址空间的分配；如果该数据流标志字段对应的地址空间为空，存储资源分配模块将数据存储在该地址空间中，并且将其对应的地址列表的值设置为1。
步骤5、当到达数据包的接收窗口时，控制逻辑模块根据数据流标识字段，确定该数据包的地址空间。如果该地址空间为空，控制逻辑模块切换到下一个数据包的取回操作；否则，控制逻辑模块取回改地址空间中的数据，然后将该地址空间对应的地址列表的值设置为0。
为了减少串行访问存储资源池导致的数据竞争为题，控制逻辑模块对数据包进行预取，即提前于给定的接收时间窗口一定时间将数据取出。
该实施例还提供了一种用轨道交通的具有容错功能的实时数据调度方法，该方法通过具有容错功能的实时以太网交换机系统来实现，其中，具有容错功能的实时以太网交换机系统包括CPU板、交换板和PHY板。
如图1所示，其中：CPU板通过PCI-e接口与交换板进行控制信息和数据信息的交互，交换板通过S3MII接口与PHY版进行数据信息的交互。
CPU板包括PowerPC架构的中央处理器(CPU)、存储单元和PCI-e接口。CPU板上运行实时Linux系统，并且通过PCI-e接口与交换板进行数据的交换。
交换板是实现具有容错功能的实时调度算法的核心硬件板卡，包括物理器件和数字电路。交换板的物理器件包括Xilinx vertex 7芯片、Flash Rom、QDR外接存储器等；交换板的数字电路由FPGA设计，烧录到Xilinx芯片中。
PHY板上包含Broadcom的PHY ASIC芯片，是物理接口收发器，负责实现物理层。
图2是本发明的交换板内部结构图，控制物理设备实现普通数据帧的转发,保障实时数据帧的实时性传输,并且提供容错功能。
交换卡的内部结构包括MAC、分类器、普通数据缓存、时钟数据缓存、实时数据通道、普通数据通道、多路选择器和PCI-E接口，其中：
MAC是数据的接口，负责发送和接收以太网中的数据帧。
分类器根据以太网帧数据的EtherType字段，区分以太网数据帧的类型。0x88D7为实时数据类型，0x8801是时钟数据类型。
数据帧的类型为普通数据帧、实时数据帧和时统数据帧。实时数据帧被直接传输到实时数据通道；普通数据帧存储在普通数据缓存中；时统数据存储在时钟数据缓存中。
普通数据缓存负责存储分类器传入的普通数据帧。
时钟数据缓存负责存储分类器传入的时统数据帧。
实时数据通道负责转发实时数据，并且提供容错功能。
其包括容错模块、实时调度模块和时间同步模块。容错模块负责实现交换机容错功能；实时调度模块负责调度实时数据帧，以保障数据帧的实时性传输；时钟同步模块负责为整个网络的设备提供同步的全局时钟。
普通数据通道负责实现普通数据帧的转发功能。
多路选择器负责根据优先级，选择应该发送的数据帧。其中实时数据帧的优先级最高，普通数据帧的优先级最低。
图3是本发明的容错模块结构图，其中实现容错功能的模块包括仲裁模块、存储资源分配逻辑模块、存储资源池和控制逻辑模块。
仲裁模块的功能是决定进入容错模块数据的流向。仲裁模块响应中央处理器的命令，打开或者关闭容错功能。当容错功能开启时，仲裁模块将数据传输到存储资源分配模块。当容错功能关闭时，仲裁模块将数据直接传输到实时数据调度模块。
存储资源分配模块负责为数据分配其在存储资源池中的位置。在实时数据中，每个数据流具有一个特定的字段进行标识，字段取值为1-4096。该模块通过解析标识字段来决定数据流在存储资源池中的存储位置。
存储资源池负责缓存需要进行容错的实时数据。本发明中的存储资源池通过外接 存储设备QDR实现。存储池中的地址空间为1-4096，每一个空间能且仅能存储一个以太网的最长帧。存储地址空间与数据流标识字段一一对应。存储资源池中维护一个4096位的地址列表，用于记录1-4096地址空间中是否已经存储了数据。如果地址空间存储了数据，在列表中该地址位标识为1，否则标识为0；
控制逻辑模块负责在规定的时间从存储资源池中取回数据。每个实时数据对应一个时间表的条目，条目中记录该数据的接收时间窗口和发送时间。每个实时数据应该在规定的时间窗口中到达，并且在规定的时间点发送出去。当到达数据的接收时间窗口时，该模块获取数据流标识字段，从存储资源池中获取该数据。在获取过程中，该模块首先查询地址列表，如果数据流字段对应地址的地址列表值为1，该模块从存储资源池中获取数据，之后将数据传输到实时调度模块。否则，该模块不响应该数据的取回请求。
数据通过mac模块的接收，到达分类器。分类器通过解析以太网数据的EtherType字段，将字段值为0x88D7的实时数据转发到容错模块。实时数据首先通过仲裁模块，如果该实时数据对应的输入端口容错功能关闭，则被仲裁模块转发到实时调度模块；如果该实时数据对应的输入端口容错功能打开，则被仲裁模块转发到存储资源分配模块中。存储资源分配模块根据数据标识字段为该实时数据在存储资源池中分配存储空间。当控制逻辑模块判定该实时数据的接收时间到达时，从存储资源池中将该实时数据取出，并发送到实时调度模块。打开端口的容错功能时，因数据提前到达实时数据缓存在资源存储池中，并在规定的接收时间被取出，所以实时数据不会因为提前到达实时调度模块而被丢弃，从而实现了容错功能。
虽然上面结合本发明的优选实施例对本发明的原理进行了详细的描述，本领域技术人员应该理解，上述实施例仅仅是对本发明的示意性实现方式的解释，并非对本发明包含范围的限定。实施例中的细节并不构成对本发明范围的限制，在不背离本发明的精神和范围的情况下，任何基于本发明技术方案的等效变换、简单替换等显而易见的改变，均落在本发明保护范围之内。

Claims (6)

  1. 一种轨道交通领域的具有容错功能的实时数据调度方法，其特征在于：其特征在于包括以下步骤：
    步骤1、对具有容错功能的实时以太网交换机系统进行初始化；硬件启动后，对容错模块进行初始化，清空存储资源池中的数据，将地址列表的值都置为1；
    步骤2、监测交换机故障；CPU板上搭载的Linux系统，通过心跳数据包监测周围交换机是否发生故障，当交换机的某端口接收不到指定的心跳数据包时，CPU判定对应端口连接的交换机发生故障；当没有检测到交换机故障时，仲裁模块直接将数据帧传输到实时调度模块中；
    步骤3、响应交换机故障，当CPU板检测到周围交换机故障时，CPU向交换板发送打开容错功能的指令，从而打开连接故障交换机端口的容错功能；
    步骤4、判断仲裁模块根据各端口的容错功能是否打开，决定来自各端口数据包的流向；如果端口的容错功能处于关闭状态，仲裁模块将来自该端口的数据包转发到实时调度模块；如果端口的容错功能处于开启状态，仲裁模块将来自该端口的数据包转发到存储资源分配模块；
    步骤5、当到达数据包的接收窗口时，控制逻辑模块根据数据流标识字段，确定该数据包的地址空间；如果该地址空间为空，控制逻辑模块切换到下一个数据包的取回操作；否则，控制逻辑模块取回改地址空间中的数据，然后将该地址空间对应的地址列表的值设置为0。
  2. 根据权利要求1所述的轨道交通领域的具有容错功能的实时数据调度方法，其特征在于，步骤4中，存储资源分配模块根据数据流标识字段，为实时数据包分配存储资源池地址空间；
    如果该数据流标志字段对应的地址空间中已经存储数据，存储资源分配模块清空该地址空间的数据，并将其对应的地址列表中的值设置为0，然后重新对该实时数据进行地址空间的分配；
    如果该数据流标志字段对应的地址空间为空，存储资源分配模块将数据存储在该地址空间中，并且将其对应的地址列表的值设置为1。
  3. 一种实现权利要求1所述的用轨道交通的具有容错功能的实时数据调度方法的实时以太网交换机系统，包括CPU板、交换板和PHY板；其特征在于：
    CPU板通过PCI-e接口与交换板进行控制信息和数据信息的交互，交换板通过接口与PHY版进行数据信息的交互；
    CPU板包括中央处理器、存储单元和PCI-e接口；
    CPU板上运行实时Linux系统，并且通过PCI-e接口与交换板进行数据的交换；
    交换板是实现具有容错功能的实时调度算法的硬件板；
    PHY板上包含PHY ASIC芯片，是物理接口收发器，负责实现物理层。
  4. 根据权利要求3所述的实时以太网交换机系统，其特征在于：交换卡的内部结构包括MAC、分类器、普通数据缓存、时钟数据缓存、实时数据通道、普通数据通道、多路选择器，其中：
    MAC是数据的接口，负责发送和接收以太网中的数据帧；
    分类器根据以太网帧数据的EtherType字段，区分以太网数据帧的类型；
    数据帧的类型为普通数据帧、实时数据帧和时统数据帧；实时数据帧被直接传输到实时数据通道；普通数据帧存储在普通数据缓存中；时统数据存储在时钟数据缓存中；
    普通数据缓存负责存储分类器传入的普通数据帧；
    时钟数据缓存负责存储分类器传入的时统数据帧；
    实时数据通道负责转发实时数据，并且提供容错功能；
    多路选择器负责根据优先级，选择应该发送的数据帧；其中实时数据帧的优先级最高，普通数据帧的优先级最低。
  5. 根据权利要求4所述的实时以太网交换机系统，其特征在于：实现容错功能的模块包括仲裁模块、存储资源分配逻辑模块、存储资源池和控制逻辑模块；
    其中，仲裁模块的功能是决定进入容错模块数据的流向；仲裁模块响应中央处理器的命令，打开或者关闭容错功能；当容错功能开启时，仲裁模块将数据传输到存储资源分配模块；当容错功能关闭时，仲裁模块将数据直接传输到实时数据调度模块；
    存储资源分配模块负责为数据分配其在存储资源池中的位置；
    存储资源池负责缓存需要进行容错的实时数据；控制逻辑模块负责在规定的时间从存储资源池中取回数据；
    每个实时数据对应一个时间表的条目，条目中记录该实时数据的接收时间窗口和发送时间；每个实时数据应该在规定的时间窗口中到达，并且在规定的时间点发送出去。
  6. 根据权利要求5所述的实时以太网交换机系统，其特征在于：实时数据首先通过仲裁模块，如果该实时数据对应的输入端口容错功能关闭，则被仲裁模块转发到实时调度模块；
    如果该实时数据对应的输入端口容错功能打开，则被仲裁模块转发到存储资源分配模块中；
    存储资源分配模块根据数据标识字段为该实时数据在存储资源池中分配存储空间；
    当控制逻辑模块判定该实时数据的接收时间到达时，从存储资源池中将该实时数据取出，并发送到实时调度模块；
    打开端口的容错功能时，因数据提前到达实时数据缓存在资源存储池中，并在规定的接收时间被取出，所以实时数据不会因为提前到达实时调度模块而被丢弃，从而实现了容错功能。
PCT/CN2018/095177 2017-06-06 2018-07-10 一种轨道交通领域的具有容错功能的实时数据调度方法 WO2018224052A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710417001.XA CN107276845B (zh) 2017-06-06 2017-06-06 一种轨道交通领域的具有容错功能的实时数据调度方法
CN201710417001.X 2017-06-06

Publications (1)

Publication Number Publication Date
WO2018224052A1 true WO2018224052A1 (zh) 2018-12-13

Family

ID=60064647

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/095177 WO2018224052A1 (zh) 2017-06-06 2018-07-10 一种轨道交通领域的具有容错功能的实时数据调度方法

Country Status (2)

Country Link
CN (1) CN107276845B (zh)
WO (1) WO2018224052A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107276845B (zh) * 2017-06-06 2019-11-15 清华大学 一种轨道交通领域的具有容错功能的实时数据调度方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140022912A1 (en) * 2012-07-23 2014-01-23 Broadcom Corporation Flexray communications using ethernet
CN103608762A (zh) * 2013-05-09 2014-02-26 华为技术有限公司 存储设备、存储系统及数据发送方法
CN204272154U (zh) * 2014-12-31 2015-04-15 北京纵横机电技术开发公司 高速动车组网络系统的以太网交换机
CN104796356A (zh) * 2015-04-16 2015-07-22 青岛四方车辆研究所有限公司 轨道车辆用车载以太网交换机和信号收发及列车重联方法
CN105978762A (zh) * 2016-04-27 2016-09-28 刘巍 冗余以太网数据传输设备、系统及方法
CN106330377A (zh) * 2016-08-25 2017-01-11 中车青岛四方车辆研究所有限公司 轨道列车宽带网络实时性数据传输系统及方法
CN107276845A (zh) * 2017-06-06 2017-10-20 清华大学 一种轨道交通领域的具有容错功能的实时数据调度方法

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140022912A1 (en) * 2012-07-23 2014-01-23 Broadcom Corporation Flexray communications using ethernet
CN103608762A (zh) * 2013-05-09 2014-02-26 华为技术有限公司 存储设备、存储系统及数据发送方法
CN204272154U (zh) * 2014-12-31 2015-04-15 北京纵横机电技术开发公司 高速动车组网络系统的以太网交换机
CN104796356A (zh) * 2015-04-16 2015-07-22 青岛四方车辆研究所有限公司 轨道车辆用车载以太网交换机和信号收发及列车重联方法
CN105978762A (zh) * 2016-04-27 2016-09-28 刘巍 冗余以太网数据传输设备、系统及方法
CN106330377A (zh) * 2016-08-25 2017-01-11 中车青岛四方车辆研究所有限公司 轨道列车宽带网络实时性数据传输系统及方法
CN107276845A (zh) * 2017-06-06 2017-10-20 清华大学 一种轨道交通领域的具有容错功能的实时数据调度方法

Also Published As

Publication number Publication date
CN107276845B (zh) 2019-11-15
CN107276845A (zh) 2017-10-20

Similar Documents

Publication Publication Date Title
US10425359B2 (en) Packet data traffic management apparatus
US9727508B2 (en) Address learning and aging for network bridging in a network processor
US6111875A (en) Apparatus and method for disabling external frame forwarding device for use with a network switch
US9755947B2 (en) Hierarchical self-organizing classification processing in a network switch
US11695708B2 (en) Deterministic real time multi protocol heterogeneous packet based transport
US9485200B2 (en) Network switch with external buffering via looparound path
EP2774336B1 (en) Real-time distributed network module, real-time distributed network and method therefor
US6091707A (en) Methods and apparatus for preventing under-flow conditions in a multiple-port switching device
US9537776B2 (en) Ethernet traffic management apparatus
US20060126517A1 (en) Loop detection method and device
CN108259127B (zh) Pcie双冗余万兆网ip核
CN105933162B (zh) 基于t型结构的低时延以太网转发器及方法
JP5957055B2 (ja) 航空機用データ通信ネットワーク
JP2015076889A (ja) 航空機用データ通信ネットワーク
US20140050078A1 (en) Communication interruption time reduction method in a packet communication network
WO2016101488A1 (zh) 一种三态内容可寻址存储器tcam查表方法及装置
JP4072583B2 (ja) 共有媒体アクセス制御回路を有する統合マルチポートスイッチ
US10484238B2 (en) Radio communication system for an industrial automation system, method for operating said radio communication system, and radio transceiver station
US6084878A (en) External rules checker interface
US20100128726A1 (en) Data transmitter and data transmission method
EP2549690A1 (en) Media access control address keep-alive device and method for convergence port
WO2018224052A1 (zh) 一种轨道交通领域的具有容错功能的实时数据调度方法
US20140112338A1 (en) Multi-Core Processor and Method for Multiplexing Network Management Port Thereof
US20220400031A1 (en) Master device, arithmetic processing device, programmable logic controller, network, and method
KR101544592B1 (ko) 능동적 큐 할당 방식의 고가용성 분산임베디드 네트워크 송수신 방법 및 장치

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: 18813708

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18813708

Country of ref document: EP

Kind code of ref document: A1