Network security architecture and implementation method of dynamic reconfigurable system based on FPGA
By adopting the network security architecture of a dynamic reconfigurable system based on FPGA in the cloud computing environment, and using FPGA for data filtering and protocol offloading, the network security interleaving problem in the cloud computing environment is solved, and high bandwidth and high-performance network forwarding is achieved. Suitable for enterprise-level users.
Patent Information
- Application Number
- CN202410938908.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-15
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2044-07-15
AI Technical Summary
There is a problem of triple security interweaving of network security in the cloud computing environment. Unknown attacks caused by traditional network threat forms and unknown vulnerabilities are difficult to prevent. In the existing technology, FPGA heterogeneous execution of network security architecture is less researched and cannot meet the high level of enterprise-level users. Bandwidth and flexible security requirements.
Design a network security architecture of a dynamic reconfigurable system based on FPGA, including programmable network cards, memory and processors, FPGA is used for data filtering and protocol offloading, and realize high-performance network forwarding and security policy adjustment through dynamic local reconfiguration. Provides high bandwidth and network security capabilities.
Effectively respond to unknown threats and persistent attacks, reduce the load on the software network protocol stack, improve performance, and provide high-performance network forwarding capabilities. It is suitable for enterprise-level users and has good application prospects.
Smart Images

Figure CN118473835B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of network security technology, and in particular to a network security architecture and implementation method of a dynamically reconfigurable system based on FPGA. Background Art
[0002] Cloud computing data centers are also facing the generalized functional safety of functional safety, network security and information data security. In the cloud computing environment, there are both common problems and particularities in network security. Among them, common problems refer to the traditional forms of network threats in the cloud computing environment, such as system security, data security, network security and user privacy, among which network attacks caused by endogenous security common problems such as vulnerabilities and backdoors are still serious challenges. Under the conditions of "design defects and untrustworthy open supply chains", it is impossible to guarantee the ideal security requirements of "toxic and sterile" in the cyberspace environment from both technical and economic perspectives. From the perspective of "known unknowns" risk prevention, unknown attacks based on unknown vulnerabilities or unknown backdoors belong to "unknown unknowns", and unknown unknowns belong to uncertainty. Uncertainty threats are one of the core factors that make it difficult to reverse the imbalance of network security. In order to solve the generalized functional safety problem of triple security interweaving, the industry has proposed the classic dynamic heterogeneous redundant structure (DHR). At present, heterogeneous executors in dynamic heterogeneous redundant structure (DHR) mostly use ARM, X86, etc., and there is little research on network security architecture for FPGA heterogeneous execution. Users want security products that have the ability to provide flexible services, open and universal security components that are applicable in the existing network environment and can be configured to meet future security needs. Summary of the invention
[0003] To this end, the present invention provides a network security architecture and a network security implementation method for a dynamically reconfigurable system based on FPGA, an enterprise-level reconfigurable component that provides security services at the reconfigurable layer, and can provide high-bandwidth network access functions and network security functions for enterprise-level users.
[0004] According to the design scheme provided by the present invention, on the one hand, a network security architecture of a dynamically reconfigurable system based on FPGA is provided, comprising:
[0005] A programmable network card, used for receiving data messages and extracting target application messages from the data messages by using a data filtering mechanism;
[0006] A memory, the memory being connected to the programmable network card and configured to receive and store target application messages extracted by the programmable network card;
[0007] A processor, connected to the programmable network card and the memory, for executing a corresponding target application according to a user instruction, generating a reconfiguration task according to a security policy decision mechanism and a protocol offload control mechanism, and sending the reconfiguration task to the programmable network card;
[0008] Among them, the programmable network card includes: a network interface for receiving data messages, a cache unit for caching data messages, and a hardware processor for extracting target application messages from the cached data messages using a data filtering task mechanism, the hardware processor includes at least one FPGA, and the hardware processor configures the FPGA according to the data filtering task indication or the reconfiguration task indication so that the FPGA performs the corresponding task according to the task indication.
[0009] As the network security architecture of the FPGA-based dynamic reconfigurable system of the present invention, further, the hardware processor is connected to at least three FPGAs to respectively respond to target application task instructions and reconfiguration task instructions.
[0010] As a network security architecture of the FPGA-based dynamic reconfigurable system of the present invention, further, when responding to a reconfiguration task indication, the FPGA loads a bitstream file for protocol unloading in an external storage medium, and uses dynamic local reconfiguration to enable the bitstream file to perform protocol replacement at the smallest granularity.
[0011] On the other hand, the present invention also provides a network security implementation method of a dynamically reconfigurable system based on FPGA, which is implemented based on the above architecture, and the implementation process includes:
[0012] The processor responds to the startup task of the target application and sends the target task to the programmable network card, wherein the target task includes a target application task and / or a reconfiguration task;
[0013] The programmable network card determines at least one functional element in the target task information, wherein the target task is used to indicate the task to be performed by the FPGA in the programmable network card, and the at least one functional element is used to jointly indicate the data flow direction in the target task;
[0014] Mapping the at least one functional element to at least one FPGA in the programmable network card respectively to generate configuration information, wherein the configuration information is used to represent a connection relationship between the at least one FPGA in the programmable network card, and the connection relationship corresponds to the data flow direction;
[0015] The FPGA in the programmable network card is configured according to the configuration information and the target task is executed.
[0016] As a network security implementation method of the FPGA-based dynamic reconfigurable system of the present invention, further, when responding to a reconfiguration task indication, the FPGA in the programmable network card loads a bitstream file for protocol unloading in an external storage medium, and uses dynamic local reconfiguration to enable the bitstream file to perform protocol replacement at the smallest granularity.
[0017] As a network security implementation method of a dynamically reconfigurable system based on FPGA of the present invention, further, before loading the bit stream file for protocol unloading in the external storage medium, it also includes:
[0018] The protocol type is specified according to the upper layer application requirements, and the corresponding bitstream file is set according to the specified protocol type.
[0019] As a network security implementation method for the FPGA-based dynamic reconfigurable system of the present invention, further, when responding to a reconfiguration task indication, the FPGA in the programmable network card utilizes a reconfiguration controller IP core to load a bitstream file for protocol unloading in an external storage medium, and the reconfiguration controller IP core is composed of one or more virtual package managers connected to a bitstream extraction path.
[0020] As a network security implementation method of the FPGA-based dynamic reconfigurable system of the present invention, further, using the reconfiguration controller IP core to load the bit stream file for protocol unloading in the external storage medium, it includes:
[0021] When a virtual package trigger signal is detected, the virtual package manager will query the reconfigurable module corresponding to the trigger signal and control the configuration of the reconfigurable module. The trigger signal is a register write signal. Multiple virtual package managers run in parallel, and each virtual package manager controls the corresponding virtual package.
[0022] As a network security implementation method of the FPGA-based dynamic reconfigurable system of the present invention, further, the security policy decision mechanism includes:
[0023] Abnormal data of the target application is detected according to the security policy. When the target application is detected to be running abnormally, the filtering rules in the data filtering task mechanism are replaced and sent to the programmable network card hardware processor, and the FPGA in the programmable network card is dynamically configured.
[0024] As a network security implementation method of the FPGA-based dynamic reconfigurable system of the present invention, further, in the data filtering task mechanism, data messages are filtered based on the protocol, port, source address and destination address to extract the target application message.
[0025] Beneficial effects of the present invention:
[0026] The present invention aims at generalized functional safety issues such as unknown threats or network attacks, and uses the message screening rule processing mechanism and the information demand processing mechanism to extract target message data. When the message screening rule processing mechanism or the information demand processing mechanism in the processor receives abnormal adjudication information, it can notify the programmable network card to change the message screening rule or information demand strategy, so as to achieve the purpose of dynamically adjusting the message screening rule or information demand for abnormal information, and solve the unknown attack or continuous blind attack based on internal and external collaboration in the network space; the transport layer and network layer related affairs implemented by software are implemented by hardware, so as to reduce the load of the software network protocol stack, improve the performance of unloading the network forwarding function, provide high-performance network forwarding capability, and release the general computing resources of the CPU. It can provide high-bandwidth network access function and network security function for enterprise-level users, and has good application prospects in the field of network security. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] Figure 1 This is a schematic diagram of the network security architecture of the FPGA-based dynamically reconfigurable system in the embodiment;
[0028] Figure 2 This is a schematic diagram of the network security implementation process in the embodiment;
[0029] Figure 3 This is a schematic diagram of the working principle of the virtual package manager in the embodiment;
[0030] Figure 4 It is a schematic diagram of the working principle of the programmable network card in the embodiment;
[0031] Figure 5 This is a schematic diagram of the IPv6 data packet filtering process in the embodiment;
[0032] Figure 6 The figure is a schematic diagram of the TCP data packet filtering process in the embodiment. DETAILED DESCRIPTION
[0033] In order to make the purpose, technical solutions and advantages of the present invention clearer and more understandable, the present invention is further described in detail below in conjunction with the accompanying drawings and technical solutions.
[0034] Traditional network security architecture generally includes a processor and a network card, which are connected through a data bus. Among them, the processor is a serial processing unit, generally a central processing unit; the data bus is a serial interface data bus or a parallel interface data bus. The application runs on the processor, and the network card does not have a hardware offload function. The network card receives the message sent to the application and transmits it to the processor through the data bus. The processor's processing of the network message needs to be parsed by the protocol stack and then transmitted to the application layer for use. Protocol stack parsing includes but is not limited to protocol stack parsing of each layer such as the link layer, network layer, and transport layer. The performance of layer-by-layer parsing is affected and the delay is large. Especially for long messages, the software processing performance is seriously affected because multiple CPU memory copies are required. Although in some solutions, the network card can realize the function offload of data sending / receiving direction, it is a specific processing for a certain protocol in a certain scenario. The hardware offload flexibility is poor, and the offload function is not used in most applications, and network acceleration cannot be achieved in a true sense.
[0035] In this embodiment, in order to increase the dynamic heterogeneous redundancy characteristics of the FPGA heterogeneous execution architecture, avoid unknown attacks or continuous blind attacks based on internal and external collaboration in cyberspace, and solve the generalized functional safety problem of triple security interweaving, see Figure 1 As shown, a network security architecture of a dynamically reconfigurable system based on FPGA is provided, including:
[0036] A programmable network card, used for receiving data messages and extracting target application messages from the data messages by using a data filtering mechanism;
[0037] A memory, the memory being connected to the programmable network card and configured to receive and store target application messages extracted by the programmable network card;
[0038] A processor, connected to the programmable network card and the memory, for executing a corresponding target application according to a user instruction, generating a reconfiguration task according to a security policy decision mechanism and a protocol offload control mechanism, and sending the reconfiguration task to the programmable network card;
[0039] Among them, the programmable network card includes: a network interface for receiving data messages, a cache unit for caching data messages, and a hardware processor for extracting target application messages from the cached data messages using a data filtering task mechanism, the hardware processor includes at least one FPGA, and the hardware processor configures the FPGA according to the data filtering task indication or the reconfiguration task indication so that the FPGA performs the corresponding task according to the task indication.
[0040] The processor may be a CPU, which may be an independent chip, or a CPU integrated in a system on chip (SoC), or a CPU integrated in a microcontroller unit (MCU). A programmable network card is a network card provided with a hardware processor, which is a hardware processor that uses a hardware description language for data processing, and the hardware processor is an FPGA. Applications are running on the processor, and the number of running applications is not limited, which is 1 or more, and the corresponding policy decision modules are also 1 or more. The programmable network card receives messages sent to the application through the network interface, and the processor receives data through the programmable network card. When the processor policy decision module receives abnormal data, it will replace the packet filtering and content filtering mechanisms to realize the dynamic configuration of the filtering rules. By configuring the FPGA to dynamically configure the filtering rules, support for new security business needs can be achieved.
[0041] The hardware processor is connected to at least three FPGAs to respectively respond to target application task instructions and reconfiguration task instructions.
[0042] Furthermore, based on the above architecture, the embodiment of the present invention also provides a network security implementation method of a dynamically reconfigurable system based on FPGA, see Figure 2 As shown, the implementation process includes:
[0043] The processor responds to the startup task of the target application and sends the target task to the programmable network card, wherein the target task includes a target application task and / or a reconfiguration task;
[0044] The programmable network card determines at least one functional element in the target task information, wherein the target task is used to indicate the task to be performed by the FPGA in the programmable network card, and the at least one functional element is used to jointly indicate the data flow direction in the target task;
[0045] Mapping the at least one functional element to at least one FPGA in the programmable network card respectively to generate configuration information, wherein the configuration information is used to indicate a connection relationship between the at least one FPGA in the programmable network card, and the connection relationship corresponds to the data flow direction;
[0046] The FPGA in the programmable network card is configured according to the configuration information and the target task is executed.
[0047] Among them, the security policy adjudication mechanism includes: detecting abnormal data of target application operation according to the security policy, and when the target application operation is detected to be abnormal, changing the filtering rules in the data filtering task mechanism and sending them to the programmable network card hardware processor, and dynamically configuring the FPGA in the programmable network card. In the data filtering task mechanism, data packets can be filtered based on the protocol, port, source address and destination address to extract the target application packets.
[0048] In order to implement the transport layer and network layer related transactions implemented by software by hardware, the hardware offloading solution for applications is proposed to reduce the load of the software network protocol stack, improve performance, offload the network forwarding function, provide high-performance network forwarding capabilities, and release CPU general computing resources. Hardware offloading mainly delegates the network, storage, security, and control functions originally running on the processor to the programmable network card for processing. The programmable network card has a bitstream file for protocol offloading in external storage media such as DDR, FLASH, NAND, NOR, and QSPI. Through dynamic local reconfiguration, the bitstream file is loaded to change the protocol at the smallest granularity. Except for the function of the reconfigurable module, the rest of the functions run normally, reducing the risk of temporary functional failure caused by changing the protocol offloading.
[0049] Specifically, they are mainly Transmission Control Protocol (TCP), User Datagram Protocol (User Datagram Protocol), Internet Control Message Protocol (ICMP), Internet Group Management Protocol (IGMP), Address Resolution Protocol (ARP) and other protocols. The specified protocol type can be determined according to the requirements of the upper layer application, for example, it can be one or a combination of Transmission Control Protocol (TCP), User Datagram Protocol (User Datagram Protocol), Internet Control Message Protocol (ICMP), Internet Group Management Protocol (IGMP), Address Resolution Protocol (ARP). Depending on the specified protocol type, the bitstream file generated by the reconfiguration will be different. For example, if the specified protocol type is Transmission Control Protocol (TCP), the Transmission Control Protocol (TCP) partial reconfiguration process is executed.
[0050] The partial reconfiguration bitstream generation process uses the ISE or Vivado development environment provided by Xilinx. The specific operation process of using Vivado to generate partial reconfiguration bitstream is as follows:
[0051] (1) Generate static and reconfigurable modules separately.
[0052] (2) Create physically constrained block regions (Pblocks) to represent reconfigurable partitions.
[0053] (3) Enable each reconfigurable partition.
[0054] (4) Implement the complete design, including the static module and one reconfigurable module in each reconfigurable partition.
[0055] (5) Save design checkpoints for complete wiring design.
[0056] (6) Delete the reconfigurable modules that need to be changed from the above design, and then only keep the static design checkpoints.
[0057] (7) Lock static placement and routing.
[0058] (8) Add the modified reconfigurable module to the static logic design and implement the new configuration, saving the checkpoint of the complete wiring design.
[0059] (9) Repeat step (8) until all reconfigurable modules are implemented.
[0060] (10) Run program verification on all configurations to ensure that all configurations do not violate the reconfigurable design rules.
[0061] (11) After successful verification, create a bitstream for each configuration.
[0062] (12) Burn the generated multiple configuration bit streams into the programmable network card through the hardware management module.
[0063] The partial reconfiguration controller IP core can automatically load partial bitstreams and is a key IP core in partial reconfiguration designs. The partial reconfiguration controller IP core can provide autonomous control functions for partial reconfiguration designs. The partial reconfiguration controller IP core mainly consists of one or more virtual package managers connected to the bitstream extraction path. The virtual package manager mainly includes logic in the reconfigurable logic and static logic to assist in the configuration of reconfigurable modules, such as ensuring that the reconfigurable modules are in a safe state before changes. The reconfigurable bitstream is found by the bitstream extraction path from the external bitstream configuration library, and then the data is transferred to the internal configuration access port (Internal Configuration Access Port, ICAP). The virtual package manager runs in parallel and monitors the occurrence of trigger events. The trigger signal can be a hardware-based signal or a signal written by software through a register. When a trigger signal is detected, the virtual package manager will query the reconfigurable module corresponding to the trigger signal and control the configuration of the module.
[0064] A virtual package manager can be in one of two states: (1) Active: The virtual package manager controls the corresponding virtual package. The virtual package detects trigger signals and loads reconfigurable modules. (2) Disabled: Other signals control the corresponding virtual package. The virtual package manager does not respond to trigger signals and does not load reconfigurable modules.
[0065] Active state is the main state of each virtual package manager. Each virtual package manager follows a basic set of steps, such as Figure 3 shown.
[0066] In this embodiment, see Figure 4 As shown in the figure, network message processing is implemented based on a dynamically reconfigurable system, where the network interface unit (NIU) can implement the physical layer and data link layer of the 10 Gigabit Ethernet interface. The IP processing module implements the parsing and processing of messages received by the network interface. It distinguishes the protocol type through the Ethernet message header and can parse protocol messages such as ARP, ICMP, IGMP, TCP and UDP. Other message types are directly sent to the PCIE DMA cache interface through the user interface and reported to the software for processing. In addition, to support the address resolution protocol, an ARP Server is built in, and MAC address resolution can be performed through hardware. The ICMP function is built in, which can echo ICMP requests or generate alarm information such as destination address unreachable. The IGMP function is built in to receive IGMP messages and respond to message information. The UDP module processes all UDP messages and sends them to the PCIE DMA cache, and triggers DMA to transmit UDP messages to the application layer. The TCP module implements actions such as TCP connection establishment, maintenance, state switching and removal, and caches TCP messages in the off-chip DDRRAM, connecting the PCIE DMA interface through the user interface.
[0067] Protocol processing modules such as ARP, ICMP, IGMP, TCP and UDP can all be partially reconfigured. Depending on the usage requirements, after partial reconfiguration, all or part of the protocol processing functions can be retained. This has the advantage of saving the internal logic resources of the FPGA and time-sharing functions in the same FPGA device resources. Partial reconfiguration can realize more complex FPGA systems. FPGAs can flexibly shut down or partially use their resources, thereby achieving power optimization without sacrificing performance.
[0068] It also supports dynamic updates and online repairs. After deployment, the FPGA configuration can be updated to improve performance, fix errors, or add new features without physically replacing the hardware.
[0069] Based on specific security business requirements, the processor can respond to the start of the target application, and the load unloading control unit controls the protocol unloading function in the programmable network card through the bus interface. The bus interface can be a serial bus interface, such as a PCIe serial interface, a USB serial interface, an RS485 interface, or an RS232 interface. After receiving the trigger signal, the programmable network card first shuts down the running reconfigurable module, then loads the Transmission Control Protocol (TCP) load unloading reconfigurable module, and after the loading is completed, starts loading the Transmission Control Protocol (TCP) load unloading reconfigurable module.
[0070] In order to increase the dynamic heterogeneous redundancy characteristics of the architecture and avoid unknown attacks or continuous blind attacks based on internal and external collaboration in cyberspace, packet filtering rule processing mechanisms and content filtering processing mechanisms are proposed, through which packet filtering rules and content filtering rules can be dynamically and real-time replaced.
[0071] Packet filtering and content filtering are mainly based on protocol filtering, port filtering, source IP address and destination address filtering. They filter message data, allowing safe data packets that meet the rules to pass through and discarding data packets that do not meet the rules.
[0072] The data configured for the Content-Addressable Memory (CAM) module will be written to the CAM module; the data configured for the network mask RAM module will be written to the corresponding RAM module. The network data packets are stored in two asynchronous FIFOs. The processor sends the network data packets to the message packet extraction module, and the bus interface module writes the data into the input FIFO; the message packet extraction module writes the successfully filtered network data packets into the output FIFO and sends them to the bus through the bus read operation. The content addressable memory CAM module is responsible for matching the corresponding data fields, including the TCP and UDP data packet destination port number CAM, the TCP and UDP data packet source port number CAM, the destination IP address CAM, the source IP address CAM, and the source MAC address CAM. MIPS performs initialization configuration and real-time data update operations on the CAM module. The network mask RAM module is used to store the subnet mask data of the source IP address and the destination IP address. In the matching mode, whether to match a certain range of addresses is optional. MIPS is responsible for transmitting the signal to the firewall module whether to match the IP address range. If only a certain range of message packets are matched, the IP address is ANDed with the data stored in the RAM to obtain the IP address range, and then sent to the IP address CAM for matching. In this way, a set of data in the CAM can match a certain range of IP addresses. Because CAM is relatively resource-intensive, this mode can save a certain amount of area resources. If an accurate match is required, this module will not be enabled. Message packet extraction module, this module is responsible for parsing the corresponding data packets, and according to the information in the data packet header, extracts the field information with the corresponding meaning and sends it to the corresponding CAM module for matching to obtain the matching result. If the match is successful, the network data packet is sent to the FIFO. If it fails, the data in the FIFO is cleared.
[0073] In order to more clearly illustrate the process of replacing packet filtering rules and content filtering rules, the following takes IPv6 data packets as an example to exemplify the dynamic and real-time replacement of packet filtering rules and content filtering rules through the packet filtering rule processing mechanism and the content filtering processing mechanism of the present application.
[0074] See also Figure 5 As shown, MIPS is an open source IP core of FPGA. ARM hard core, NIOS soft core, etc. can also be used. It is mainly used to control bus data and initialize the system before work. MIPS sends the filtering strategy to the CAM module and the network mask data to the RAM module to complete the configuration of the filtering rules.
[0075] The Ethernet module generates an interrupt to occupy the bus, and MIPS obtains the IPv6 data packet from the bus and sends the IPv6 data packet to the inFIFO.
[0076] Read inFIFO, get the MAC address, and send it to the CAM module for matching. Read inFIFO, get the IPv6 data packet header information, and send it to the corresponding CAM module for matching. If it is TCP / UDP, extract the port number and send it to the CAM module for matching.
[0077] If the CAM matches successfully, the inFIFO is cleared and the next interrupt is generated by the Ethernet module. Otherwise, the IPv6 data packet is written to the outFIFO.
[0078] In order to more clearly illustrate the process of dynamic adjustment of the packet filtering rule processing mechanism and the content filtering processing mechanism provided by this embodiment, the packet filtering rule processing mechanism and the content filtering processing mechanism of the embodiment of the present application are exemplarily illustrated below using TCP data packets as an example.
[0079] See also Figure 6 As shown, in response to the start of the target application running in the processor, the message filtering policy and network mask of the target application are configured in the programmable network card of the computing device.
[0080] According to the message filtering policy and the network mask, the pending messages sent to the TCP application are screened, and the TCP application runs on the processor of the computer device.
[0081] The DMA method writes the target data into the target memory corresponding to the programmable network card, so that the target application and policy decision can read the target data from the target memory.
[0082] If the policy decision detects an anomaly, the processor of the computer device will modify the message filtering policy and the network mask in a DMA manner.
[0083] Unless otherwise specifically stated, the relative steps, numerical expressions and values of the components and steps set forth in these embodiments do not limit the scope of the present invention.
[0084] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.
[0085] The units and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. A person of ordinary skill in the art may use different methods to implement the described functions for each specific application, but such implementation is not considered to be beyond the scope of the present invention.
[0086] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits, and accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or in the form of software function modules. The present invention is not limited to any specific form of combination of hardware and software.
[0087] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims
1. A network security system based on FPGA dynamic reconfigurable system, characterized in that: Include: A programmable network card, used for receiving data messages and extracting target application messages from the data messages by using a data filtering mechanism; A memory, the memory being connected to the programmable network card and configured to receive and store target application messages extracted by the programmable network card; A processor, connected to the programmable network card and the memory, for executing a corresponding target application according to a user instruction, generating a reconfiguration task according to a security policy decision mechanism and a protocol offload control mechanism, and sending the reconfiguration task to the programmable network card; Among them, the programmable network card includes: a network interface for receiving data messages, a cache unit for caching data messages, and a hardware processor for extracting target application messages from the cached data messages using a data filtering task mechanism, the hardware processor includes at least one FPGA, the hardware processor configures the FPGA according to a data filtering task indication or a reconfiguration task indication, so that the FPGA performs a corresponding task according to the task indication; the hardware processor is connected to at least three FPGAs to respond to the target application task indication and the reconfiguration task indication respectively; when the message screening rule processing mechanism or the information demand processing mechanism in the processor receives abnormal ruling information, the programmable network card replaces the data filtering task mechanism, and when the FPGA responds to the reconfiguration task indication, it loads a bit stream file for protocol unloading in an external storage medium, and enables the bit stream file to perform protocol replacement at the smallest granularity through dynamic local reconfiguration.
2. A network security implementation method based on FPGA dynamic reconfigurable system, characterized in that: Based on the network security system implementation of claim 1, the implementation process includes: The processor responds to the startup task of the target application and sends the target task to the programmable network card, wherein the target task includes a target application task and / or a reconfiguration task; The programmable network card determines at least one functional element in the target task information, wherein the target task is used to indicate the task to be performed by the FPGA in the programmable network card, and the at least one functional element is used to jointly indicate the data flow direction in the target task; Mapping the at least one functional element to at least one FPGA in the programmable network card respectively to generate configuration information, wherein the configuration information is used to indicate a connection relationship between the at least one FPGA in the programmable network card, and the connection relationship corresponds to the data flow direction; The FPGA in the programmable network card is configured according to the configuration information and the target task is executed.
3. The network security implementation method of the FPGA-based dynamic reconfigurable system according to claim 2 is characterized in that: Before loading the bitstream file for protocol offloading in the external storage medium, it also includes: The protocol type is specified according to the upper layer application requirements, and the corresponding bitstream file is set according to the specified protocol type.
4. The network security implementation method of the FPGA-based dynamic reconfigurable system according to claim 2 is characterized in that: When responding to a reconfiguration task indication, the FPGA in the programmable network card uses a reconfiguration controller IP core to load a bitstream file for protocol unloading in an external storage medium. The reconfiguration controller IP core consists of one or more virtual package managers connected to a bitstream extraction path.
5. The network security implementation method of the FPGA-based dynamic reconfigurable system according to claim 4 is characterized in that: Use the reconfiguration controller IP core to load the bitstream file for protocol offloading from the external storage medium, including: When a virtual package trigger signal is detected, the virtual package manager will query the reconfigurable module corresponding to the trigger signal and control the configuration of the reconfigurable module. The trigger signal is a register write signal. Multiple virtual package managers run in parallel, and each virtual package manager controls the corresponding virtual package.
6. The network security implementation method of the FPGA-based dynamic reconfigurable system according to claim 2 is characterized in that: The security policy decision mechanism includes: Abnormal data of the target application is detected according to the security policy. When the target application is detected to be running abnormally, the filtering rules in the data filtering task mechanism are replaced and sent to the programmable network card hardware processor, and the FPGA in the programmable network card is dynamically configured.
7. The network security implementation method of the FPGA-based dynamic reconfigurable system according to claim 2 or 6, characterized in that: In the data filtering task mechanism, data packets are filtered based on the protocol, port, source address and destination address to extract the target application message.
