IPv6 over AOS space communication system based on NDP protocol
Through FPGA gateway and MPLS technology, the problem of differences in spatial network and ground network protocols is solved, the conversion of IPv6 and AOS protocols and packet label exchange is realized, the efficiency and accuracy of data transmission are improved, and the large amount and amount of data are met. Types of business requirements.
Patent Information
- Application Number
- CN202211178744.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-09-27
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2042-09-27
Smart Images

Figure CN115665272B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to an IPv6 over AOS system based on NDP protocol with FPGA as a gateway, belonging to the technical field of laser communication, suitable for a space-ground integrated system based on relay satellites and ground base stations, and capable of realizing end-to-end Internet data transmission. Background Art
[0002] With the development of space communication networks, ground communication networks will no longer be the main body of communication. In the future, there will be more and more information interactions between ground communication networks and space communication networks. Space networks have the advantages of large coverage, high networking flexibility, and high resource utilization. Ground networks have the advantages of high efficiency, reliability, scalability, low bit error rate, and fast transmission speed. In order to better integrate the advantages of the two networks, realize the information interaction between ground networks and space networks, and ensure the fast and efficient transmission of data, how to integrate ground networks with space networks and build a space-ground integrated system has gradually become the focus of future aerospace engineering research. In addition, people's demand for high-speed and large-capacity data transmission is increasing day by day. For voice, telemetry and other services, they can be transmitted through low-speed channels, while for videos, files, experimental data, these services with large amounts, multiple types, and fast development require high-speed communication channels. Therefore, the unification of space networks and ground networks is very urgent and necessary.
[0003] In order to achieve the integration of space and earth, CCSDS (Consultative Committee for Space Data Systems) has put forward a series of data transmission proposals, among which the AOS (Advanced Orbiting System) protocol has diversified business processing capabilities and flexible data processing capabilities. The AOS protocol can meet the needs of two-way link communications with a wide range of communication rates between air and ground and between air and air. At the same time, AOS adopts the OSI model, provides a rich interface for upper-layer protocols, can realize network access services, and can transmit IP data packets, providing technical support for the integration of space and earth. At present, the AOS protocol has become the preferred solution for most new spacecraft data system systems.
[0004] As the next generation network protocol to replace IPv4, IPv6 has a hierarchical address structure, a large address space, high security and good scalability, which effectively reduces the amount of equipment resources occupied and improves the equipment operation speed and bandwidth utilization. With the increasing amount of Internet information transmission, the IPv6 protocol has become the mainstream protocol used in future ground networks. In addition, in order to ensure the reliable transmission of IPv6 data packets in the network, the gateway needs to implement address resolution, so as to update the neighbor cache table of the gateway port in real time. The NDP (Neighbor Discovery Protocol) protocol, as a key protocol of IPv6, has the function of address resolution, which not only solves the problem of automatic configuration of addresses between different nodes, but also improves the accuracy of routing selection and ensures the real-time availability of links.
[0005] With the increasing amount of data in communication and the diversification of data types, how to achieve fast forwarding of data packets between different nodes has become a problem of concern. Traditional IP networks use hop-by-hop forwarding to send and receive data packets. Each router in the path must look up the target IP address of the data packet in the IP routing table and then forward the data packet to the next hop router. This method has a slow data forwarding speed. In order to increase the forwarding rate of datagrams and enhance the flexibility of links, IP datagrams can be forwarded at the link layer through hardware based on MPLS (Multi-protocol Label Switching). MPLS technology was proposed by IETF in 1997. It enables IP technology to provide users with more real-time services (voice transmission, video conferencing, etc.). In order to better study MPLS technology, the MPLS Forum was established in March 2000 and conducted research on MPLS technology, service level, traffic engineering and other fields. MPLS solves the link congestion problem of traditional IP packet switching and improves service processing capabilities. It has received increasing attention in the field of optical communications.
[0006] At present, space networks differ from ground networks in many aspects. For example, space networks have a high bit error rate, and due to the increase in distance, there is a high delay in transmission. At the same time, there is intermittent information transmission between satellites and the ground, and intersatellite information transmission, and link interruption will cause packet loss. At the same time, the computing power and CPU processing power on satellites are limited and cannot be compared with the computing power on the ground. In order to improve the data processing rate, the IPv6 protocol is still used for data processing on the satellite and ground networks, and the AOS protocol is used for forwarding between satellites. However, the IPv6 protocol is not suitable for long delays, high bit error rates, and link interruptions in space links. Therefore, in order to build a space-ground integrated system and overcome the differences in protocols between space networks and ground networks, gateways are required to convert between the AOS protocol and the IPv6 protocol, so as to achieve accurate transmission and reception of air-ground information at the IP layer. The gateway needs to have an address resolution function to achieve real-time updates of the neighbor cache table, improve the stability of the link, and improve the accuracy of route selection. In addition, the forwarding rate needs to be increased when data packets are transmitted between different nodes to meet the needs of large-capacity business information and multiple data types. Summary of the invention
[0007] In view of this, the present invention provides an IPv6 over AOS space communication system based on the NDP protocol. The system uses FPGA as a gateway to realize the mutual conversion between the IPv6 protocol and the AOS protocol, uses AOS frames to carry IPv6 data packets to realize real-time and effective data transmission, realizes address resolution between the gateway system and the local subnet based on the NDP protocol, realizes high-speed and stable forwarding of data packets between different nodes based on the MPLS technology, and connects with the radio frequency system through Gigabit and 10 Gigabit optical fiber interfaces to realize point-to-point space link transmission.
[0008] In order to achieve the above object, the technical solution adopted by the present invention is:
[0009] An IPv6 over AOS space communication system based on the NDP protocol, comprising a source node digital signal cross-processing board, an intermediate node digital signal cross-processing board and a destination node digital signal cross-processing board, each board comprising a JTAG interface module, a crystal oscillator, a system clock module, a programmable clock module, a FLASH module, an SFP transceiver module, a high-speed serial-to-parallel conversion module, a serial port driver module, a data storage module and an FPGA chip;
[0010] The JTAG interface module is used to download the program or store the program in the FLASH module.
[0011] There are two crystal oscillators on each board, with frequencies of 200M and 50M respectively. The crystal oscillator is used as the starting device of the equipment to ensure that the circuit works within a stable frequency range;
[0012] The system clock module is provided by the 200M crystal oscillator on the board, which is used to provide a clock signal for the system to keep the modules of the system synchronized;
[0013] The programmable clock module is used to provide a reference clock for the high-speed transceiver and controller, and its input clock is provided by a 50M crystal oscillator on the board;
[0014] The FLASH module is a program and register configuration storage device, which is used to burn the program stored in the FLASH into the FPGA chip when the board is powered on; in addition, the FLASH module also stores the register configuration program of the programmable clock chip, and completes the configuration of the programmable clock chip when the device is powered on to achieve synchronous output of the reference clock;
[0015] There are four SFP transceiver modules on each board, which are used to realize analog / digital conversion or digital / analog conversion. The SFP transceiver module is connected to the external hardware platform or host through the Ethernet port with optical fiber or network cable to realize the transmission and reception of data packets;
[0016] The high-speed serial-to-parallel conversion module has 8 high-speed data I / O channels for realizing bidirectional point-to-point data transmission;
[0017] The serial port driver module is used to receive external input control messages;
[0018] The data storage module is an external resource of the FPGA chip and is used to store Ethernet data at a high transmission rate;
[0019] The FPGA chip performs data processing through embedded program development to achieve respective functions.
[0020] Furthermore, the high-speed serial-to-parallel conversion module uses eight TLK2711 chips to form eight high-speed data I / O channels;
[0021] The control message includes information about the source IP address, Mac address, input label, output label, output port number, message type, message length, and reserved bytes, which are used to construct a data packet transmission path and store label information when the data packet enters and leaves the node.
[0022] Furthermore, the FPGA chip of the source node digital signal cross-processing board includes the following program modules:
[0023] The first NDP autonomous learning module is used to configure the IP address and Mac address of the four-way SFP transceiver module, send NS messages and NA messages to each other with the external hardware platform or host, obtain the real-time address information of the device connected to the four-way SFP transceiver module, realize the address resolution function, and accurately send the Ethernet data packet to the adapter module; when the external hardware platform or host sends the Ethernet data to the SFP transceiver module through the optical fiber or network cable, the SFP transceiver module performs analog / digital conversion, and then sends the digital signal to the first NDP autonomous learning module. When there is no destination address in the neighbor cache table of the external hardware platform or host, When the IP address corresponds to the Mac address, address resolution is required. At this time, the first NDP autonomous learning module extracts the source IP address and the source Mac address from the received NS message, and combines the IP address and the Mac address of the configured SFP transceiver module to form an NA message, which is fed back to the external hardware platform or host after the SFP transceiver module performs digital / analog conversion, thereby realizing the writing and updating of the neighbor cache table and completing the address resolution. In addition, the first NDP autonomous learning module periodically sends NS messages to the external hardware platform or host to search the neighbor cache table and maintain its connectivity with the SFP transceiver module.
[0024] The adapter module is used to receive the Ethernet data packet from the first NDP autonomous learning module and filter the data packet; specifically, the adapter module determines the Ethernet type field, the next header field, and the IPv6 type field of the data packet to obtain the IPv6 data packet containing the service information, and then sends the filtered IPv6 data packet to the first label query replacement module;
[0025] A first RS232 serial port transceiver module is used to receive a control message in the form of a FIB table from an external hardware platform or a host, and send it to a first label query and replacement module;
[0026] The first label query and replacement module performs label exchange based on the MPLS method to realize accurate forwarding of data packets between different nodes; specifically, the FIB table sent by the first RS232 serial port transceiver module is queried, and the label is intercepted from the corresponding field and added to the label field of the AOS frame;
[0027] The protocol conversion module further encapsulates the AOS frame received from the first label replacement module, composes an AOS frame of fixed frame length according to the AOS frame format, and sends it to the first CRC check module;
[0028] The first CRC check module is used to determine the correctness of the message according to the datagram calculation polynomial, calculate the header checksum field of the IP datagram, the UDP checksum field and the frame check sequence field of the Ethernet frame, add the checksum value to the corresponding position of the IPv6 data packet, form a complete AOS frame and send it to the first TLK2711 serial-to-parallel conversion module;
[0029] The first TLK2711 serial-to-parallel conversion module is used to realize the conversion between parallel data and serial data and the high-speed transmission of data; in addition, the first TLK2711 serial-to-parallel conversion module adopts CDR clock recovery method to realize the synchronization of the clocks at both ends of the transmitter and receiver, and regularly sends comma information to align the serial-to-parallel conversion byte boundary at the receiving end;
[0030] In the source node digital signal cross-processing board, the TLK2711 chip of the high-speed serial-to-parallel conversion module converts 16-bit parallel data into serial data through the 8b / 10b encoding and decoding function, and sends it from the FPGA chip of the source node digital signal cross-processing board to the FPGA chip of the intermediate node digital signal cross-processing board.
[0031] Furthermore, the FPGA chip of the intermediate node digital signal cross-processing board includes the following modules:
[0032] The second TLK2711 serial-to-parallel conversion module is used to receive serial data from the FPGA chip of the digital signal cross-processing board card of the source node, and send the serial data to the second label query replacement module for data processing after serial / parallel conversion, and then perform parallel / serial conversion to send it to the digital signal cross-processing board card of the destination node; the specific method is to first establish a synchronization state according to the synchronization character sent by the sending end, and then continuously detect the control signals RKLSB and RKMSB, judge the data type according to the state of RKLSB and RKMSB, discard the frame header, frame tail and synchronization character, latch the data that is really needed, and finally detect the AOS frame header and send it to the second label query replacement module;
[0033] The second RS232 serial port transceiver module is used to receive the control message in the form of LFIB table sent by the external hardware platform or the host, and send it to the second label query and replacement module;
[0034] The second label query and replacement module implements label switching based on the MPLS method. Specifically, it queries the LFIB table sent by the second RS232 serial port transceiver module, replaces the corresponding outgoing label into the label field of the AOS frame according to the value of the incoming label, and then sends the AOS frame to the second TLK2711 serial-to-parallel conversion module.
[0035] Furthermore, the FPGA chip of the destination node digital signal cross-processing board includes the following modules:
[0036] The third TLK2711 serial-to-parallel conversion module is used to receive serial data from the FPGA chip of the digital signal cross-processing board of the intermediate node, and send it to the protocol conversion module after serial / parallel conversion;
[0037] A protocol conversion module, used to decapsulate the parallel data from the third TLK2711 serial-to-parallel conversion module, verify the Ethernet type field, the next header field, and the IPv6 type field of the Ethernet frame contained therein to detect the integrity of the Ethernet frame, and send the correct Ethernet frame to the second NDP autonomous learning module;
[0038] The second NDP autonomous learning module is used to configure the IP address and Mac address of the four-way SFP transceiver module, send NS messages and NA messages to and from the external hardware platform or host, obtain the real-time address information of the device connected to the four-way SFP transceiver module, and implement the address resolution function; in addition, the obtained real-time IP address information of the connected device is compared with the IP address field of the Ethernet frame, the corresponding Mac address is found and replaced to the destination Mac address field of the Ethernet frame, and the Ethernet frame is sent to the second CRC check module; for the Ethernet frame sent back by the second CRC check module, the corresponding SFP transceiver module is selected according to the IP address, and then sent to the connected external hardware platform or host;
[0039] The second CRC check module calculates the CRC check result according to the content of the Ethernet frame sent by the second NDP autonomous learning module and adds the calculated CRC check value to the FCS field of the Ethernet frame, and then sends the Ethernet frame back to the second NDP autonomous learning module.
[0040] The beneficial effects of the present invention are:
[0041] 1. The present invention uses FPGA as a gateway to realize the conversion function between the AOS protocol and the IPv6 protocol. The IPv6 protocol is used for data processing on the satellite and the ground network, and the AOS protocol is used to carry IPv6 data packets for data transmission between satellites. By adopting a consistent network layer protocol, the complexity of protocol conversion is reduced, and the effective integration of the space network and the ground network is realized.
[0042] 2. The present invention uses FPGA as a gateway, and implements address resolution between the gateway system and the local subnet based on the NDP protocol, ensuring accurate data transmission and reception. The gateway supports automatic neighbor discovery. After the system is initialized, the node can identify the port of the opposite device interconnected with the port of this node, and realizes the initial configuration of the neighbor cache table. In addition, the gateway supports the establishment and maintenance of the control channel. After discovering a new neighbor, it automatically negotiates the position and length of the fields used in the control channel, and regularly sends NS messages to the hardware platform or host in the control channel to search the neighbor cache table and maintain its connectivity. In this way, the on and off status of the link can be detected in real time, ensuring the real-time availability of the link, and also realizing real-time updating of the neighbor cache table. In short, the gateway simplifies topology discovery and routing selection, greatly improves the scalability of the system and the accuracy of routing, and reduces the cost of space communication.
[0043] 3. The present invention combines MPLS label switching technology to achieve the transmission of services between different nodes, meet the business needs of large data volume and multiple transmission types in the current aerospace engineering, solve the problems of small business volume, poor versatility and flexibility in traditional remote control, telemetry and other communication methods, and overcome the defect that the same protocol cannot be used in the link layer in space and on the ground. The use of MPLS technology can allocate QoS functions to traffic, manage and control it, and the label switching mechanism allows users to determine the priority associated with these labels. MPLS works by applying labels to data packets when they leave the customer network and enter the MPLS network. Network elements do not look up IP header information to guide the packet, but simply read the MPLS label and transmit the data packet to other nodes along the predetermined path. Compared with traditional IP routing, the traffic in the MPLS network has higher reliability and predictability, and the data packet only needs to be transmitted along the set label switching path. The present invention changes the way of forwarding through IP addresses, and instead adopts label switching based on MPLS technology for data forwarding. According to the FIB and LFIB query label corresponding forwarding path, the rapid transmission of data packets between boards is achieved, and the network performance is improved.
[0044] In summary, the present invention studies the IPv6 protocol, CCSDS space communication standard, the conversion mechanism of the IP over AOS protocol, the NDP neighbor discovery protocol and the MPLS label exchange mechanism, uses FPGA as a gateway, and solves the problem of inconsistency between the space network and the ground network protocol through conversion between the IPv6 protocol and the AOS protocol, improves network transmission performance, reduces the packet loss rate, and increases the communication rate; by regularly sending NS and NA messages in the NDP protocol, the address resolution between the gateway and its connected devices and the dynamic configuration of the routing table are realized, and the accuracy of route selection is improved; by means of MPLS label exchange, the rapid forwarding of data packets between different nodes is realized, and the forwarding efficiency is improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] Figure 1 Schematic diagram of the protocol conversion process in space network and ground network communications.
[0046] Figure 2 This is a schematic diagram of the NDP protocol address resolution process principle.
[0047] Figure 3 This is the NDP protocol address resolution process and the routing information stored by host A and host B.
[0048] Figure 4 It is the NS and NA message format in the NDP protocol.
[0049] Figure 5 This is a schematic diagram of the working process of label switching.
[0050] Figure 6 The figure is a schematic diagram of the data transmission process based on MPLS.
[0051] Figure 7 This is the data transmission flow chart of IPv6 over AOS space communication system based on NDP protocol.
[0052] Figure 8 A schematic diagram of control messages (FIB table and LFIB table).
[0053] Fig. 9 This is a schematic diagram of the AOS frame structure.
[0054] Fig.10 Schematic diagram of intermediate node label replacement. DETAILED DESCRIPTION
[0055] The present invention will be further described below in conjunction with the accompanying drawings.
[0056] An IPv6 over AOS space communication system based on NDP protocol involves the following technologies:
[0057] 1) Protocol conversion technology.
[0058] In space network and ground network communications, the protocol conversion process is as follows Figure 1 As shown, SCPS-TP is a standard protocol of the International Organization for Standardization, which improves a series of problems existing in the TCP protocol in wireless communications while maintaining interoperability with the IP protocol. The intersatellite and satellite-to-ground links use the AOS protocol to communicate through radio frequency or laser, while the traditional TCP / IP protocol is used for communication on the satellite or in the ground network. In view of the development of Internet technology, the application of the IPv6 protocol is becoming more and more extensive, so the present invention uses the IPv6 protocol in the TCP / IP protocol cluster as the network layer protocol.
[0059] 2) NDP autonomous learning technology.
[0060] NDP (Neighbor Discovery Protocol) is a sub-protocol of ICMPv6 and an important part of the IPv6 protocol system. It combines and improves the ARP protocol, ICMP router discovery, and ICMP redirection protocols of IPv4. The NDP protocol has the functions of address resolution, tracking neighbor status, duplicate address detection, router discovery, redirection, router discovery, prefix discovery, address automatic configuration, parameter discovery, and next hop address determination. Compared with the ARP protocol of IPv4, it strengthens the independence of the address resolution protocol and the underlying link, enhances security, and reduces the propagation range of the message, thereby saving network bandwidth. In this system, FPGA can use NDP autonomous learning technology to send NS messages and NA messages to the connected devices to write, search, and update the neighbor cache table, and obtain the real-time address information of the device connected to the network interface. The system configures the IP and Mac addresses for each SFP transceiver module, and actively replies to the NA message to the connected device, thereby realizing the address resolution function and ensuring the accurate transmission and reception of IPv6 data packets.
[0061] The address resolution process of the NDP protocol is as follows Figure 2As shown in the figure, host A needs to obtain the link layer address of host B before sending. In order to realize the dynamic configuration of the routing table, the link layer address of the device can be dynamically obtained through the address resolution function of the NDP protocol. Similar to the ARP protocol, host A first sends an NS message to the connected device (taking host B as an example). The message uses the IPv6 address of host A as the source address and the requested node multicast address of host B as the destination address to resolve the Mac address corresponding to the IPv6 address of host B. When host B receives the NS message, it will respond with an NA message. The source address of the NA message is the IPv6 address of host B, and the destination address is the IPv6 address of host A. Different from the multicast of the NS message, the NA message is unicast, which completes a resolution process.
[0062] When the host sends an IPv6 data packet to the FPGA, since there is no Mac address corresponding to the destination IP in the routing table, it will first send an NS message to inquire about the Mac address corresponding to the destination IP. After the FPGA replies with the Mac address of the interface, the host starts to send data packets. In addition, the FPGA will periodically send the NDP protocol to detect whether the device connected to the port is connected normally. The NDP address resolution process based on FPGA is as follows Figure 3 As shown, assuming that the IPv6 address of host A is FE80:0:0:30::, and the Mac address is C4-00-AD-74-96-C5, the IPv6 address of host B is FE80:0:0:10::, and the Mac address is C4-00-AD-90-9F-76, the IPv6 address and Mac address of the SFP transceiver module interface of FPGA1 connected to host A are FE80:0:0:01:: and 00-00-AD-90-9F-01 respectively, and the IPv6 address and Mac address of the SFP transceiver module interface of FPGA2 connected to host B are FE80:0:0:02:: and 00-00-AD-90-9F-02 respectively. First, host A and host B send NS messages to FPGA1 and FPGA2 respectively. After receiving the NS messages, FPGA sends NA messages to host A and host B respectively, but fills in the Mac address field with the Mac address of the SFP transceiver module interface of FPGA1 and the Mac address of the SFP transceiver module interface of FPGA2 respectively. This is equivalent to host A and host B asking for the Mac address corresponding to the destination IP FE80:0:0:10:: and FE80:0:0:30::, and FPGA1 and FPGA2 reply with the Mac address of the interface respectively.
[0063] The routing information stored by host A and host B through the NDP protocol is as follows Figure 3As shown in the figure, it can be seen that through the address resolution function of the NDP protocol, the routing tables of host A and host B store the target IP address and the Mac address of the SFP transceiver module interface of the FPGA to which they are connected. After obtaining the routing information, the host can start sending IPv6 data packets. In addition, in order to maintain the connectivity of the device connected to the port, the FPGA periodically sends NS messages to the connected device to detect the device connection status. If the NA message from the device can be received, the routing table information is maintained. If the NA message is not received for a long time, the corresponding device IP address and Mac address are deleted from the routing table. The formats of NS and NA messages are as follows: Figure 4 shown.
[0064] 3)MPLS Multi-Protocol Label Switching technology.
[0065] MPLS technology combines the second layer switching with the third layer routing technology, avoids multiple routing on the network through label switching, and simplifies the process of packet transmission. The forwarding method based on MPLS technology has more advantages than traditional IP forwarding, changing the hop-by-hop forwarding method used in traditional IP forwarding. It does not need to check the destination IP of the IP message, but only needs to forward according to the content in the AOS label field, which improves the data transmission rate.
[0066] In traditional IP forwarding, the physical layer receives a message from a port of the switch and sends it to the data link layer. The data link layer removes the link layer encapsulation and sends it to the corresponding network layer according to the protocol field of the message. The network layer first checks whether the message is sent to the local machine. If so, it removes the network layer encapsulation and sends it to its upper layer protocol. If not, it searches the routing table and sends the message to the data link layer of the corresponding port according to the destination IP address. This is performed once for each node passed, so the data packet sending rate is slow.
[0067] MPLS (Multi-protocol Label Switching) adds a fixed-length label to each IP datagram, and then forwards the IP datagram at the link layer through hardware. This technology has the advantages of flow control, load balancing, path backup, and fault recovery. It changes the link congestion problem that may occur in traditional IP networks. By flexibly controlling the transmission path of datagrams, it realizes the rational allocation of resources, thereby greatly speeding up the forwarding rate of datagrams. In addition, two or more LSPs (label switching paths) can be configured, and several of them can be selected as backups, thereby realizing the rational allocation of business flows and ensuring business continuity. The MPLS network consists of label switching routers (LSR: Label Switching Router) and label edge routers (LER: Label Edge Router), such as Figure 5As shown. LER is located at the edge and is responsible for adding and removing labels. LSR is located at the core of the MPLS network and is responsible for label switching. Each LSR transmits data packets based on the label distribution protocol LDP and finds the path corresponding to a specific label, namely the label switching path LSP. When an IP datagram enters the MPLS domain, a label is pressed into the entry node (LER) and forwarded to the next LSR according to the forwarding table. All LSRs will forward according to the label in the future.
[0068] The MPLS protocol includes the Label Distribution Protocol (LDP), the Label Mapping Table (LIB) and the Forwarding Information Base (FIB). As the core technology of MPLS, LDP realizes label distribution and forwarding through the signaling mechanism. The FIB (Forwarding Information Table) is responsible for forwarding ordinary IP packets, while the LFIB (Label Forwarding Information Table) is responsible for forwarding packets with MPLS labels. The present invention uses the IPv6 address as the Forwarding Equivalence Class (FEC), establishes the Label Switching Path (LSP) according to the FIB and LFIB, and uses the bidirectional ping packet method to simulate data forwarding and study the data processing process.
[0069] The data transmission process is as follows Figure 6 As shown, in the source node FPGA, after the NDP autonomous learning completes the address resolution, the host A starts to send IPv6 data packets, and combines the FIB table issued by the RS232 serial port transceiver module to form an AOS frame and send it to the TLK2711 serial-to-parallel conversion module, and then sends it to the intermediate node FPGA in the form of serial data. The intermediate node FPGA first performs serial-to-parallel conversion. When receiving a message with a label, it searches the LFIB table issued by the RS232 serial port transceiver module. If the corresponding outgoing label is a normal label, the AOS frame is forwarded to the next intermediate node. If the corresponding outgoing label is a special label (the present invention stipulates that this label is 16'hFFFF), it indicates that the data is to be sent to the Ethernet port, and its next hop is the destination node. After the AOS frame is sent to the destination node, the label in the message is removed, and the AOS frame is decapsulated and sent to the host B after the NDP autonomous learning module. There can be multiple intermediate nodes in the system. In order to verify the system function, only one intermediate node is set in the present invention.
[0070] 4) Serial-to-parallel conversion technology based on TLK2711 chip
[0071] The TLK2711 chip is used to convert between serial data and parallel data to achieve high-speed data transmission. The TLK2711 chip converts 16-bit parallel data into serial data through the 8b / 10b encoding and decoding function, and forwards it between different digital signal cross-processing boards. Data transmission and reception are all performed at the data link layer.
[0072] The Tlk2711 high-speed serial transceiver uses clock recovery technology (CDR) to synchronize the clocks at both ends of the transmitter and receiver. In addition, in order to align the serial-to-parallel conversion byte boundary at the receiver, the transmitter needs to send commas regularly. The 16-bit wide data is divided into 8 bits for high and low, and the control character consists of two bytes of D code and K code (frame header: K28.2 / K27.7, frame tail: K29.7 / K30.7, synchronization character: D5.6 / K28.5, synchronization response character: D12.1 / K28.5, idle character: D10.2 / K28.5). The system is in a state of out-of-step when it is powered on. In order to establish synchronization, the transmitter will send a 1ms synchronization character, and then the transmitter starts to send data frames, and synchronization characters need to be sent at the beginning and end of the frame to maintain the synchronization state of the link. When the system is in synchronization state, the receiving end continuously detects the control signals RKLSB and RKMSB, determines the data type according to the status of RKLSB and RKMSB, discards the frame header, frame tail and synchronization characters, latches the data that is really needed, and finally detects the AOS frame header.
[0073] This system consists of a source node digital signal cross-processing board, an intermediate node digital signal cross-processing board and a destination node digital signal cross-processing board. The board models are all XIKINX KINTEX-7 (XC7K325). Each board includes a JTAG interface module, a crystal oscillator, a system clock module, a programmable clock module, a FLASH module, an SFP transceiver module, a high-speed serial-to-parallel conversion module, a serial port driver module, a data storage module and an FPGA chip.
[0074] The JTAG interface module is mainly used to download and burn the program or to store the bin file generated by the program in the FLASH module.
[0075] The board uses two crystal oscillators with frequencies of 200M and 50M as the starting devices of the equipment to ensure that the circuit works within a stable frequency range.
[0076] The system clock module is provided by the 200M crystal oscillator on the board, which is used to provide the system clock signal for the FPGA chip to keep the modules of the system synchronized;
[0077] The programmable clock module is used to provide reference clocks for high-speed transceivers and controllers, and drives the programmable clock chip (Silicon Labs SI5338) to generate 125M and 156.25M reference clocks for use by Gigabit and 10G IP cores, and its input clock is provided by the 50M crystal oscillator on the board;
[0078] The FLASH module is a program and register configuration storage device, which is used to burn the program stored in the FLASH into the FPGA chip when the board is powered on; in addition, the FLASH module also stores the register configuration program of the programmable clock chip, and completes the configuration of the programmable clock chip when the device is powered on to achieve synchronous output of the reference clock;
[0079] The SFP transceiver module on each board has four Ethernet interfaces (LAN interfaces) for analog / digital conversion or digital / analog conversion. The SFP transceiver module is connected to the Ethernet port of the external hardware platform or host through the SFP optical module, optical fiber, and network cable to realize the transmission and reception of data packets. The SFP optical module is a hot-swappable small package module. It is an interface device that converts electrical signals into optical signals and realizes photoelectric conversion. The transmission rates of the SFP optical module are 100M and 1000M, while the SFP+ optical module is an enhanced version of the SFP optical module with a transmission rate of 10G. The SFP transceiver module of the XIKINX KINTEX-7 board has four Ethernet interfaces (LAN interfaces) for data communication with the hardware platform or host. Two of the interfaces use the SFP electrical port optical module to receive Gigabit Ethernet data, and two interfaces use the SFP+ optical port optical module to receive 10G Ethernet data. The received Ethernet data is sent to the FPGA chip through four G-bit transceivers (GTX).
[0080] The high-speed serial-to-parallel conversion module includes 8 TLK2711 chips, which constitute 8 data transmission links (I / O channels) to realize the serial-to-parallel conversion function of data. The interface for external data output (baseband interface) is divided into 6 TLK2711 electrical ports and 2 TLK2711 optical ports. The electrical port uses LVDS signal line to transmit data, while the optical port uses optical fiber to transmit data. The TLK2711 high-speed serial transceiver has a serial transmission efficiency of up to 2.7Gbps, and has the advantages of high transmission rate, strong anti-interference ability, low power consumption, hot plug protection, etc. It is very suitable for high-speed two-way data transmission system on satellites. Using serial communication between different boards can effectively improve the data transmission rate and transmission bandwidth.
[0081] The serial port driver module uses 2 RS232 (control and debug interface) to realize serial port communication with the FPGA chip, and is used to receive external input control messages (FIB table and LFIB table). The chip model is MAX3232.
[0082] The data storage module is an external resource of the FPGA chip. The module uses a DDR3 chip (model MT41J256M16HA-125) with the advantages of low power consumption and heat generation, higher operating frequency, and good versatility. It is suitable for large-capacity data storage under high transmission rates of gigabit or 10 gigabit.
[0083] The FPGA chip realizes functions such as protocol conversion, address resolution, label switching, serial / parallel and parallel / serial conversion, data packet forwarding, and control message reception through embedded program development. The FPGA chips of the source node digital signal cross-processing board, the intermediate node digital signal cross-processing board, and the destination node digital signal cross-processing board have different program functions, and the specific functions are as follows:
[0084] The FPGA chip of the source node digital signal cross-processing board includes the following program modules: a first NDP autonomous learning module, an adaptation module, a first RS232 serial port transceiver module, a first label replacement query module, a protocol conversion module, a first CRC check module, and a first TLK2711 serial-to-parallel conversion module.
[0085] The FPGA chip of the intermediate node digital signal cross-processing board includes the following modules: a second TLK2711 serial-to-parallel conversion module, a second RS232 serial port transceiver module, and a second label replacement query module.
[0086] The FPGA chip of the destination node digital signal cross-processing board includes the following modules: a third TLK2711 serial-to-parallel conversion module, a protocol conversion module, a second NDP autonomous learning module, and a second CRC check module.
[0087] Specifically, the transmission principle of the IPv6 over AOS space communication system based on the NDP protocol is as follows: Figure 7As shown, this system is a duplex unit, which has both the function of receiving data and the function of sending data, and uses the ping packet to simulate the service sending process. Before PC A sends the service, it first needs to complete the address resolution based on the NDP protocol. PC A sends the NS message to the first NDP autonomous learning module of the FPGA1 (source node digital signal cross-processing board) chip. The first NDP autonomous learning module stores the IPv6 address and Mac address of PCA and replies with the NA message. PCA stores the Mac address under the queried destination IPv6 address in the routing table, and then PCA can start sending IPv6 data packets. The IPv6 data packet is sent from the first NDP autonomous learning module to the adaptation module, and the qualified ICMPv6 data packet is screened out and sent to the first label query replacement module. After receiving the FIB table sent by the first RS232 serial port transceiver module, the first label query replacement module pushes the label according to the content in the table (the pressed label is obtained by the PCA through the FIB table sent by the first RS232 serial port transceiver module) and preliminarily encapsulates it into an AOS frame. After the protocol conversion module and the first CRC check module form a complete AOS frame, the first TLK2711 serial-to-parallel conversion module converts the parallel data into serial data through 8b / 10b encoding and transmits it through physical channels such as microwaves or lasers, and forwards it to FPGA2 (digital signal intersection of intermediate nodes) according to the label based on MPLS technology. The second TLK2711 serial-to-parallel conversion module of the FPGA3 (destination node digital signal cross-processing board) chip converts the data into parallel data and sends it to the second label replacement query module. The second label replacement query module performs label exchange according to the LFIB table received by the second RS232 serial port transceiver module, and then sends it back to the second TLK2711 serial-to-parallel conversion module. After converting it into serial data, it is sent to the third TLK2711 serial-to-parallel conversion module of the FPGA3 (destination node digital signal cross-processing board) chip. After serial-to-parallel conversion, the AOS frame is sent to the protocol conversion module, and the Ethernet frame is sent to the second NDP autonomous learning module after decapsulation. After replacing the Mac address, the second CRC check module checks whether the FCS (frame check sequence) of the Ethernet frame is correct, and finally the second NDP autonomous learning module sends it to the corresponding network port. The same is true for the process from PC B to PC A. In this process, FPGA1 serves as the destination node digital signal cross-processing board and FPGA3 serves as the source node digital signal cross-processing board. In order to verify the system function, the present invention uses LVDS signal lines to simulate the high-speed transmission and reception of link data. The program modules included in each node are as follows:
[0088] The source node digital signal cross-processing board FPGA chip includes the following program modules:
[0089] 1. The first NDP self-learning module:
[0090] It is used to configure the IP address and Mac address of the four-way SFP transceiver module, send NS messages and NA messages to and from the hardware platform or host, obtain the real-time address information of the device connected to the four-way SFP transceiver module, realize the address resolution function, and accurately send the Ethernet data packet to the adapter module; when the external hardware platform or host sends the Ethernet data to the SFP transceiver module through the optical fiber or network cable, the SFP transceiver module performs analog / digital conversion, and then sends the digital signal to the first NDP autonomous learning module for address resolution. When there is no destination IP address in the neighbor cache table of the external hardware platform or host, the SFP transceiver module performs analog / digital conversion, and then sends the digital signal to the first NDP autonomous learning module for address resolution. When the address corresponds to the Mac address, address resolution is required. At this time, the first NDP autonomous learning module extracts the source IP address and the source Mac address from the received NS message, and combines the IP address and the Mac address of the configured SFP transceiver module to form an NA message, which is fed back to the external hardware platform or host after digital / analog conversion by the SFP transceiver module, thereby realizing the writing and updating of the neighbor cache table and completing the address resolution. In addition, the first NDP autonomous learning module periodically sends NS messages to the external hardware platform or host to search the neighbor cache table and maintain its connectivity with the SFP transceiver module.
[0091] 2. Adaptation module:
[0092] It is mainly used to receive Ethernet data packets from the first NDP autonomous learning module and filter and screen the data packets; specifically, the adaptation module determines the Ethernet type field, next header field, and IPv6 type field of the data packet to obtain an IPv6 data packet containing business information, and then sends the filtered IPv6 data packet to the first label query replacement module. The filtering content mainly includes the Ethernet frame type, the next packet header, and the ICMPv6 type. The specific filtering list is as follows:
[0093]
[0094]
[0095] 3. The first RS232 serial port transceiver module:
[0096] Used to receive control messages in the form of FIB tables from external hardware platforms or hosts, and send them to the first tag query replacement module. The RS232 interface in the present invention is a commonly used asynchronous interface, and the transmission format is: start bit + 8 data bits + stop bit, and no check bit is used. The serial port communication rate is 115.2Kbps, and the communication distance is 15m. Control messages refer to FIB tables and LFIB tables, such as Figure 8As shown. The FIB table and LFIB table consist of 160 bytes, including the start bit (0xFAFA), source IP address, Mac address, input label, output label, the corresponding input port number when the FPGA receives the control message, the corresponding output port number when the control message is sent, the message type, message length, reserved bytes, 124Byte fixed-length data field, end bit (0XF5F5) and other information, which are used to build the data packet transmission path and store the label information when the data packet enters and leaves the node. In addition, the RS232 serial port transceiver module also supports the processing of control messages, that is, the 160Byte control message is split into 16Bytes each and sent to the protocol conversion module respectively, that is, the received control message is inserted into the AOS frame for transmission.
[0097] 4. First label replacement query module:
[0098] Label switching is performed based on MPLS to achieve accurate forwarding of data packets between different nodes; specifically, the FIB table sent by the first RS232 serial port transceiver module is queried, and the label is intercepted from the corresponding field and added to the label field of the AOS frame.
[0099] 5. Protocol conversion module:
[0100] For the source node digital signal cross-processing board FPGA chip, the protocol conversion module further encapsulates the AOS frame received from the first label replacement module, composes an AOS frame with a fixed frame length according to the AOS frame format, and sends it to the first CRC check module.
[0101] The format of the AOS frame is as follows Fig. 9As shown in FIG. 1 , the length of an AOS frame is 1024 bytes, including a transmission frame header (10 bytes), a transmission frame insertion field (15 bytes), a reserved field (48 bytes), an IPOC (7 bytes), a transmission frame data field (810 bytes), and a transmission frame tail (134 bytes). The synchronization header is used to locate the time when the frame arrives and achieve frame synchronization; the version number and the spacecraft identifier constitute the main channel identifier (MCID); the virtual channel identifier (VCID) and the main channel identifier constitute the global virtual channel (GVCID), which is used to mark each virtual channel; the label field, the DCN field, and The OAM domain is included in the reserved field. The label domain is used to forward AOS frames between different nodes. The frame transmission path is constructed in combination with the FIB table and the LFIB table. The DCN domain is used to receive the control messages sent by the RS232 serial port transceiver module. Every 16 bytes are stored in this field and transmitted to other nodes through 10 AOS frames. The receiving end extracts the control messages inserted in the domain and merges them to obtain a complete 160-byte control message. The OAM domain can be used to implement link management and maintenance, including the extraction and insertion of OAM overhead information, the generation and control of OAM frames, performance detection, fault warning and other functions.
[0102] 6. The first CRC check module:
[0103] It is used to determine the correctness of the message according to the datagram calculation polynomial, calculate the header checksum field of the IP datagram, the UDP checksum field and the frame check sequence (FCS) field of the Ethernet frame, add the checksum value to the corresponding position of the IPv6 data packet, form a complete AOS frame and send it to the first TLK2711 serial-to-parallel conversion module. The calculation method of each field is as follows:
[0104] 1. The IP datagram header checksum only checks the datagram header, not the data part. First, divide the IP datagram header into a sequence of many 16-bit words, set the checksum field to 0, and use the inverse arithmetic operation to add all the 16-bit words. The inverse of the sum is the IP datagram header checksum.
[0105] 2. UDP checksum, UDP checksum verification content includes 12 bytes pseudo header (including 4 bytes source IP address, 4 bytes destination IP address, 1 byte 0, 1 byte protocol, 2 bytes UDP datagram length), 2 bytes source port, 2 bytes destination port, 2 bytes UDP datagram length, 2 bytes UDP checksum (filled with 0 when calculating), and the data part of the UDP datagram. The UDP checksum verifies both the header and the data, while the IP datagram header checksum only verifies the IP datagram header.
[0106] 3. There are three verification methods for FCS. The first method is to calculate the CRC32 check code bit by bit. The second method is to use a non-flipped lookup table for fast calculation and calculate the CRC32 check code by byte. The third method is to use a flipped lookup table for fast calculation and calculate the CRC check code by byte. The present invention uses the first method to calculate the value of FCS. First, a CRC generator polynomial G(x) is selected, the transmitted data is shifted left by k bits, and 0 is added to the right side (where k is the highest power of the generator polynomial). Then, the data after shifting and adding 0 is used to perform a modulo-2 division (actually an XOR operation) on G(x). The remainder obtained is the CRC check code of the data.
[0107] 7. The first TLK2711 serial-to-parallel conversion module:
[0108] It is used to realize the conversion between parallel data and serial data and the high-speed transmission of data. In addition, the first TLK2711 serial-to-parallel conversion module uses CDR clock recovery to synchronize the clocks at both ends of the transmitter and receiver, and periodically sends comma information to align the serial-to-parallel conversion byte boundary at the receiving end. In the source node digital signal cross-processing board FPGA chip, the TLK2711 chip of the high-speed serial-to-parallel conversion module converts the 16-bit parallel data into serial data through the 8b / 10b encoding and decoding function, and sends it from the FPGA chip of the source node digital signal cross-processing board to the FPGA chip of the intermediate node digital signal cross-processing board.
[0109] The program modules included in the intermediate node digital signal cross-processing board FPGA chip are:
[0110] 1. The second TLK2711 serial-to-parallel conversion module,
[0111] It is used to receive serial data from the FPGA chip of the digital signal cross-processing board of the source node, send it to the second label query and replacement module for data processing after serial / parallel conversion, and then perform parallel / serial conversion to send it to the digital signal cross-processing board of the destination node; the specific method is to first establish a synchronization state according to the synchronization character sent by the sending end, and then continuously detect the control signals RKLSB and RKMSB, judge the data type according to the state of RKLSB and RKMSB, discard the frame header, frame tail and synchronization character, latch the data that is really needed, and finally detect the AOS frame header and send it to the second label query and replacement module.
[0112] 2. The second RS232 serial port transceiver module,
[0113] Used to receive the control message in the form of LFIB table sent by the external hardware platform or host, and send it to the second label query and replacement module.
[0114] 3. The second label replacement query module.
[0115] The label switching is realized based on the MPLS method. Specifically, the LFIB table sent by the second RS232 serial port transceiver module is queried, and the corresponding outgoing label is replaced into the label field of the AOS frame according to the value of the incoming label, and then the AOS frame is sent to the second TLK2711 serial-to-parallel conversion module.
[0116] The program modules included in the destination node digital signal cross-processing board FPGA chip are:
[0117] 1. The third TLK2711 serial-to-parallel conversion module:
[0118] It is used to receive serial data from the FPGA chip of the digital signal cross-processing board of the intermediate node, and send it to the protocol conversion module after serial / parallel conversion.
[0119] 2. Protocol conversion module:
[0120] For the destination node digital signal cross-processing board FPGA chip, the protocol conversion module is used to decapsulate the parallel data (AOS frame) from the third TLK2711 serial-to-parallel conversion module, and check the Ethernet type field, next header field, and IPv6 type field of the Ethernet frame contained therein to detect the integrity of the Ethernet frame, and send the correct Ethernet frame to the second NDP autonomous learning module
[0121] 3. Second NDP self-learning module:
[0122] It is used to configure the IP address and Mac address of the four-way SFP transceiver module, send NS messages and NA messages to and from the external hardware platform or host, obtain the real-time address information of the device connected to the four-way SFP transceiver module, and realize the address resolution function; in addition, the obtained real-time IP address information of the connected device is compared with the IP address field of the Ethernet frame, the corresponding Mac address is found and replaced to the destination Mac address field of the Ethernet frame, and the Ethernet frame is sent to the second CRC check module; for the Ethernet frame sent back by the second CRC check module, the corresponding SFP transceiver module is selected according to the IP address, and then sent to the connected external hardware platform or host.
[0123] 4. Second CRC check module:
[0124] For the destination node digital signal cross-processing board FPGA chip, the CRC check module is used to determine the correctness of the message, recalculate the FCS field, and then send it to the NDP autonomous learning module.
[0125] Through the above analysis, it can be seen that the present invention proposes an IPv6 over AOS system, which transmits IP data packets through AOS frames based on the NDP protocol and label exchange mechanism, realizes automatic identification of devices connected to Ethernet ports and dynamic configuration of routing tables, and the system can realize the encapsulation, forwarding and decapsulation of AOS frames, thereby realizing the normal sending and receiving of IPv6 data packets. This system has the following functions:
[0126] ①Support the conversion of IPv6 protocol messages and AOS protocol messages.
[0127] ②Support label-based AOS frame forwarding and support real-time processing of AOS frames (such as label push, label replacement, and label information forwarding).
[0128] ③Supports the processing of control packets. Control messages (FIB and LFIB) are sent and received via RS232 and can be forwarded via AOS frames.
[0129] ④Support dynamic interaction between the network port and the connected device, and dynamically update the routing cache table according to the connection status.
[0130] ⑤After initialization, the node can identify the port of the opposite device to which the node port is connected. After discovering a new neighbor, it automatically negotiates the position and length of the field used in the control channel, and regularly sends protocol packets in the control channel to maintain its connectivity.
[0131] In order to verify the functions and performance of the system, a test environment was built to test the IPv6 over AOS gateway system, and the logic of each module was verified, and the tool software was used to capture packets for analysis. The performance of this system was verified by testing the system's packet loss rate, bandwidth utilization, frame error rate, and data transmission bit rate.
[0132] When the service reaches the intermediate node, the label is replaced as follows: Fig.10 As shown in the figure, the experiment uses the 1st and 5th TLK2711 to transmit the two-way ping data. It can be seen that when the data packet arrives at the intermediate node from host A (IP address is FE80:0:0:30::) through the first TLK2711, the label is 16'h0001. After querying the LFIB table, the label is replaced with 16'hFFFF after passing through the intermediate node and sent to the destination node through the fifth TLK2711. Similarly, host B (IP address is FE80:0:0:10::) also completes the label replacement at the intermediate node. After 1 hour of testing, the experimental results are shown in Table 1. There is no packet loss during the data transmission process, which verifies the stability of the system.
[0133]
[0134] Table 1: Ping results of 2-port Gigabit Ethernet and 2-port 10 Gigabit Ethernet data transmission simultaneously
[0135] The bandwidth utilization was measured using a network analyzer. The results are shown in Table 2. After half an hour of measurement, the number of frames received and sent by the Gigabit Ethernet was 2645503, the number of frames received and sent by the 10 Gigabit Ethernet was 9920635, the number of error frames was 0, and the bandwidth utilization was 100%.
[0136]
[0137] Table 2: Number of frames received and sent, frame error rate, data bit rate, and bandwidth utilization of Gigabit and 10 Gigabit Ethernet interfaces
[0138] In summary, the system of the present invention encapsulates IPv6 data packets in AOS frames and transmits them in space through lasers, integrating ground networks and space networks to meet the current high-speed and large-capacity requirements for data transmission. The system uses FPGA as a gateway and implements the address resolution function between the gateway and its subnets and the dynamic configuration of the routing table based on the NDP protocol. The system implements the mapping between the IPv6 protocol and the AOS protocol through Ethernet frame adaptation and protocol conversion; in addition, the system combines MPLS technology and makes routing decisions based on labels, which reduces the networking cost compared to the original switching method based on IP routing for forwarding, improves the forwarding efficiency of devices with the same performance, and has good scalability. This system is a duplex unit, consisting of a source node digital signal cross-processing board, an intermediate node digital signal cross-processing board, and a destination node digital signal cross-processing board, and has the functions of receiving and sending data. Each board of the system includes a JTAG interface module, a crystal oscillator, a system clock module, a programmable clock module, a FLASH module, an SFP transceiver module, a high-speed serial-to-parallel conversion module, a serial port driver module, a data storage module, and an FPGA chip. The FPGA chip implements the following functional modules through embedded program development: an NDP autonomous learning module, an adapter module, an RS232 serial port transceiver module, a label replacement query module, a protocol conversion module, a CRC check module, and a TLK2711 serial-to-parallel conversion module. The communication rate of the 2-way Gigabit Ethernet electrical port in the present invention is ≥1Gbps, and the communication rate of the 2-way 10G Ethernet optical port is ≥2.5Gbps. The bandwidth utilization of the Gigabit and 10G Ethernet ports is close to 100% after being tested by a network analyzer.
Claims
1. An IPv6 overAOS space communication system based on NDP protocol, characterized in that: It includes a source node digital signal cross-processing board, an intermediate node digital signal cross-processing board and a destination node digital signal cross-processing board, each board includes a JTAG interface module, a crystal oscillator, a system clock module, a programmable clock module, a FLASH module, an SFP transceiver module, a high-speed serial-to-parallel conversion module, a serial port driver module, a data storage module and an FPGA chip; The JTAG interface module is used to download the program or store the program in the FLASH module. There are two crystal oscillators on each board, with frequencies of 200M and 50M respectively. The crystal oscillator is used as the starting device of the equipment to ensure that the circuit works within a stable frequency range; The system clock module is provided by the 200M crystal oscillator on the board, which is used to provide a clock signal for the system to keep the modules of the system synchronized; The programmable clock module is used to provide a reference clock for the high-speed transceiver and controller, and its input clock is provided by a 50M crystal oscillator on the board; The FLASH module is a program and register configuration storage device, which is used to burn the program stored in the FLASH into the FPGA chip when the board is powered on; in addition, the FLASH module also stores the register configuration program of the programmable clock chip, and completes the configuration of the programmable clock chip when the device is powered on to achieve synchronous output of the reference clock; There are four SFP transceiver modules on each board, which are used to realize analog / digital conversion or digital / analog conversion. The SFP transceiver module is connected to the external hardware platform or host through the Ethernet port with optical fiber or network cable to realize the transmission and reception of data packets; The high-speed serial-to-parallel conversion module has 8 high-speed data I / O channels for realizing bidirectional point-to-point data transmission; The serial port driver module is used to receive external input control messages; The data storage module is an external resource of the FPGA chip and is used to store Ethernet data at a high transmission rate; The FPGA chip performs data processing by means of embedded program development to realize respective functions; The high-speed serial-to-parallel conversion module uses 8 TLK2711 chips to form 8 high-speed data I / O channels; The control message includes information about the source IP address, Mac address, inbound label, outbound label, outbound port number, message type, message length, and reserved bytes, which are used to construct a data packet transmission path and store label information when the data packet enters and leaves the node; The FPGA chip of the source node digital signal cross-processing board includes the following program modules: The first NDP autonomous learning module is used to configure the IP address and Mac address of the four-way SFP transceiver module, send NS messages and NA messages to each other with the external hardware platform or host, obtain the real-time address information of the device connected to the four-way SFP transceiver module, realize the address resolution function, and accurately send the Ethernet data packet to the adapter module; when the external hardware platform or host sends the Ethernet data to the SFP transceiver module through the optical fiber or network cable, the SFP transceiver module performs analog / digital conversion, and then sends the digital signal to the first NDP autonomous learning module. When there is no destination address in the neighbor cache table of the external hardware platform or host, When the IP address corresponds to the Mac address, address resolution is required. At this time, the first NDP autonomous learning module extracts the source IP address and the source Mac address from the received NS message, and combines the IP address and the Mac address of the configured SFP transceiver module to form an NA message, which is fed back to the external hardware platform or host after the SFP transceiver module performs digital / analog conversion, thereby realizing the writing and updating of the neighbor cache table and completing the address resolution. In addition, the first NDP autonomous learning module periodically sends NS messages to the external hardware platform or host to search the neighbor cache table and maintain its connectivity with the SFP transceiver module. The adapter module is used to receive the Ethernet data packet from the first NDP autonomous learning module and filter the data packet; specifically, the adapter module determines the Ethernet type field, the next header field, and the IPv6 type field of the data packet to obtain the IPv6 data packet containing the service information, and then sends the filtered IPv6 data packet to the first label query replacement module; A first RS232 serial port transceiver module is used to receive a control message in the form of a FIB table from an external hardware platform or a host, and send it to a first label query and replacement module; The first label query and replacement module performs label exchange based on the MPLS method to realize accurate forwarding of data packets between different nodes; specifically, the FIB table sent by the first RS232 serial port transceiver module is queried, and the label is intercepted from the corresponding field and added to the label field of the AOS frame; The protocol conversion module further encapsulates the AOS frame received from the first label replacement module, composes an AOS frame of fixed frame length according to the AOS frame format, and sends it to the first CRC check module; The first CRC check module is used to determine the correctness of the message according to the datagram calculation polynomial, calculate the header checksum field of the IP datagram, the UDP checksum field and the frame check sequence field of the Ethernet frame, add the checksum value to the corresponding position of the IPv6 data packet, form a complete AOS frame and send it to the first TLK2711 serial-to-parallel conversion module; The first TLK2711 serial-to-parallel conversion module is used to realize the conversion between parallel data and serial data and the high-speed transmission of data; in addition, the first TLK2711 serial-to-parallel conversion module adopts CDR clock recovery method to realize the synchronization of the clocks at both ends of the transmitter and receiver, and regularly sends comma information to align the serial-to-parallel conversion byte boundary at the receiving end; In the source node digital signal cross-processing board, the TLK2711 chip of the high-speed serial-to-parallel conversion module converts the 16-bit parallel data into serial data through the 8b / 10b encoding and decoding function, and sends it from the FPGA chip of the source node digital signal cross-processing board to the FPGA chip of the intermediate node digital signal cross-processing board; The FPGA chip of the intermediate node digital signal cross-processing board includes the following modules: The second TLK2711 serial-to-parallel conversion module is used to receive serial data from the FPGA chip of the digital signal cross-processing board card of the source node, and send the serial data to the second label query replacement module for data processing after serial / parallel conversion, and then perform parallel / serial conversion to send it to the digital signal cross-processing board card of the destination node; the specific method is to first establish a synchronization state according to the synchronization character sent by the sending end, and then continuously detect the control signals RKLSB and RKMSB, judge the data type according to the state of RKLSB and RKMSB, discard the frame header, frame tail and synchronization character, latch the data that is really needed, and finally detect the AOS frame header and send it to the second label query replacement module; The second RS232 serial port transceiver module is used to receive the control message in the form of LFIB table sent by the external hardware platform or the host, and send it to the second label query and replacement module; The second label query and replacement module implements label switching based on the MPLS method. Specifically, the LFIB table sent by the second RS232 serial port transceiver module is queried, and the corresponding outgoing label is replaced into the label field of the AOS frame according to the value of the incoming label, and then the AOS frame is sent to the second TLK2711 serial-to-parallel conversion module; The FPGA chip of the destination node digital signal cross-processing board includes the following modules: The third TLK2711 serial-to-parallel conversion module is used to receive serial data from the FPGA chip of the digital signal cross-processing board of the intermediate node, and send it to the protocol conversion module after serial / parallel conversion; A protocol conversion module, used to decapsulate the parallel data from the third TLK2711 serial-to-parallel conversion module, verify the Ethernet type field, the next header field, and the IPv6 type field of the Ethernet frame contained therein to detect the integrity of the Ethernet frame, and send the correct Ethernet frame to the second NDP autonomous learning module; The second NDP autonomous learning module is used to configure the IP address and Mac address of the four-way SFP transceiver module, send NS messages and NA messages to and from the external hardware platform or host, obtain the real-time address information of the device connected to the four-way SFP transceiver module, and implement the address resolution function; in addition, the obtained real-time IP address information of the connected device is compared with the IP address field of the Ethernet frame, the corresponding Mac address is found and replaced to the destination Mac address field of the Ethernet frame, and the Ethernet frame is sent to the second CRC check module; for the Ethernet frame sent back by the second CRC check module, the corresponding SFP transceiver module is selected according to the IP address, and then sent to the connected external hardware platform or host; The second CRC check module calculates the CRC check result according to the content of the Ethernet frame sent by the second NDP autonomous learning module and adds the calculated CRC check value to the FCS field of the Ethernet frame, and then sends the Ethernet frame back to the second NDP autonomous learning module.
