Server BMC dynamic security authentication and firmware protection method and system based on hardware root of trust

Abstract

The invention discloses a server BMC dynamic security authentication and firmware protection method and system based on a hardware root of trust, relates to the technical field of server remote management security, and discloses the server BMC dynamic security authentication and firmware protection method and system based on the hardware root of trust, which realize dynamic authentication by generating a device fingerprint through a security chip. According to the method, a multi-layer protection mechanism is constructed in combination with fragment encryption, differential hash check and memory integrity verification, and meanwhile, the data credibility is enhanced by using a block chain storage key hash value, so that the problems of static authentication risk, insufficient firmware tampering detection and missing of a trust chain in a traditional scheme are effectively solved, and the security protection capability of a BMC system can be improved.

Description

Technical Field

[0001] This application relates to the field of server remote management security technology, and in particular to a server BMC dynamic security authentication and firmware protection method and system based on hardware root of trust. Background Technology

[0002] With the increasing demand for remote server management, the Base Controller (BMC), as a core component carrying management protocols such as IPMI and Redfish, faces severe security threats. Traditional BMC security mechanisms have serious flaws: at the authentication level, reliance on static password authentication or pre-set X.509 certificates not only carries the risk of weak authentication but also results in a high risk of key leakage due to the complexity of certificate update mechanisms; at the firmware protection level, it only supports static hash verification of the entire firmware image, lacking sufficient detection capabilities against differential tampering attacks and lacking continuous runtime protection; regarding trust chain construction, it fails to establish a complete root of trust from hardware to software, making the BMC's own security configuration vulnerable to tampering. Existing technical solutions such as static certificate authentication, firmware hash verification, and access control lists cannot effectively solve core issues such as dynamic authentication, fine-grained firmware protection, and the construction of a complete trust chain. Especially when facing advanced persistent threats, traditional solutions are insufficiently effective.

[0003] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of this application is to provide a method and system for dynamic security authentication and firmware protection of server BMC based on hardware root of trust, which aims to improve the security protection capability of BMC system.

[0005] To achieve the above objectives, this application proposes a server BMC dynamic security authentication and firmware protection method based on a hardware root of trust. The method includes: The unique identifier data and hardware version data of the server motherboard are obtained through the security chip. Device fingerprint data is generated based on the unique identifier data and hardware version data, and the device fingerprint data is stored in the tamper-proof storage area of ​​the security chip. Dynamic authentication processing based on device fingerprint data: Generate a first random number and send it to the management platform through the out-of-band channel of the management interface; receive signature data containing a second random number and timestamp data returned by the management platform; verify the legality of the signature data using the security chip and device fingerprint data; if the verification is successful, generate session key data based on the elliptic curve Diffie-Hellman algorithm; upload the hash value of the session key data to the blockchain network for storage. Firmware protection based on session key data: During firmware transmission, the firmware update package data is fragmented and encrypted to obtain encrypted fragment data; checksum data of the encrypted fragment data is generated based on the SHA-3 hash algorithm; when receiving firmware update package data, the integrity of the checksum data is verified piece by piece; during firmware storage, incremental update log data is generated based on the firmware update package data and stored in the secure partition of the BMC flash memory; abnormal changes in the incremental update log data are monitored based on the differential hash algorithm; during firmware operation, a trusted execution environment is established through a security chip and a base hash value is pre-stored; periodic memory integrity checks are performed on critical processes of the BMC kernel to obtain memory checksum data; if the memory checksum data is abnormal, the management interface is frozen and an alarm notification is triggered. The security chip performs digital signature processing on BMC user permission configuration data to obtain the signature configuration data; when a permission change request is received, the hardware key and dynamic password are used to form two-factor authentication data and perform authentication processing; security event log data is collected in real time, and the security event log data is analyzed based on the anomaly detection model to trigger an automatic locking mechanism.

[0006] In one embodiment, the step of performing dynamic authentication processing based on device fingerprint data includes: The first random number is generated using the security chip; The first random number is sent to the management platform via an out-of-band channel; Receive signature data returned by the management platform. This signature data is generated by the management platform using its security chip to sign the first random number, the second random number, and the timestamp data. The signature data is verified using a security chip and device fingerprint data. If the verification is successful, the session key data is generated using the elliptic curve Diffie-Hellman algorithm based on the first and second random numbers, and the hash value of the session key data is uploaded to the blockchain network.

[0007] In one embodiment, the step of generating session key data using the elliptic curve Diffie-Hellman algorithm based on a first random number and a second random number, and uploading the hash value of the session key data to the blockchain network includes: Initial key data is generated based on the elliptic curve Diffie-Hellman algorithm and the first and second random numbers. The initial key data is processed using the key derivation function to obtain the derived key data; The derived key data is output to the blockchain network as session key data.

[0008] In one embodiment, the steps of performing fragmented encryption processing on the firmware update package data during the firmware transmission stage to obtain encrypted fragmented data; generating verification value data of the encrypted fragmented data based on the SHA-3 hash algorithm; and verifying the integrity of the verification value data fragment by fragment when receiving the firmware update package data include: Use session key data to perform fragmented encryption on firmware update package data, and obtain encrypted fragmented data; Verification value data is generated for each encrypted data fragment based on the SHA-3 hash algorithm, and a Merkle tree structure is constructed to store the verification value data; When receiving firmware update package data, the encrypted fragment data is extracted piece by piece and its checksum data is verified to match the Merkle tree structure in order to verify the integrity of the checksum data.

[0009] In one embodiment, the steps of generating incremental update log data based on firmware update package data during the firmware storage stage, storing the incremental update log data in a secure partition of the BMC flash memory, and monitoring abnormal changes in the incremental update log data based on a differential hash algorithm include: Incremental update log data is generated based on firmware update package data; Divide the BMC flash memory into a security partition to store incremental update log data; The differential hashing algorithm is used to monitor changes in incremental update log data in real time. If an anomaly is detected in the incremental update log data, an alert is generated and firmware loading is paused.

[0010] In one embodiment, the step of using the differential hashing algorithm to monitor changes in incremental update log data in real time includes: The incremental update log data is processed based on the sliding window algorithm to generate the current fingerprint data. Perform a differential comparison between the current fingerprint data and the pre-stored baseline fingerprint data; If the difference value exceeds the preset threshold rule data, the incremental update log data is determined to be abnormal.

[0011] In one embodiment, the steps of establishing a trusted execution environment and pre-storing baseline hash value data through a security chip during the firmware execution phase, performing periodic memory integrity checks on critical processes of the BMC kernel, and obtaining memory check result data; if the memory check result data is abnormal, freezing the management interface and triggering an alarm notification include: A trusted execution environment is established using a secure chip, and pre-stored baseline hash value data is loaded. Periodically scan key processes in the BMC kernel to obtain memory region data; Based on a preset range of key function segment addresses, extract key function segment data from memory region data; Calculate the hash value of the key function segment data and compare it with the baseline hash value data to generate memory verification result data.

[0012] In one embodiment, the step of calculating the hash value of the key function segment data and comparing it with the baseline hash value data to generate memory verification result data includes: Extract key function segment data; The fast hash algorithm is applied to calculate the real-time hash value of key function segment data; The real-time hash value data is compared with the pre-stored baseline hash value data. If they are inconsistent, the memory verification result data is determined to be abnormal.

[0013] In one embodiment, the steps of real-time acquisition of security event log data, analysis of the security event log data based on an anomaly detection model, and triggering an automatic locking mechanism include: Collect security event log data in real time from the system log; Security event log data is analyzed based on a Bayesian classifier model and preset threshold rules to detect abnormal behavior. If abnormal behavior is detected, a lock command is generated to trigger the automatic locking mechanism.

[0014] Furthermore, to achieve the above objectives, this application also proposes a server BMC dynamic security authentication and firmware protection system based on a hardware root of trust. The system includes a memory, a processor, and a server BMC dynamic security authentication and firmware protection program based on a hardware root of trust, stored in the memory and executable on the processor. The program is configured to implement the steps of the aforementioned server BMC dynamic security authentication and firmware protection method.

[0015] The proposed method and system for dynamic security authentication and firmware protection of server BMC based on hardware root of trust achieves dynamic authentication by generating device fingerprints through a security chip. It combines fragmented encryption, differential hash verification and memory integrity verification to build a multi-layer protection mechanism. At the same time, it uses blockchain to store key hash values ​​to enhance data credibility. It effectively solves the problems of static authentication risk, insufficient firmware tamper detection and lack of trust chain in traditional solutions, and can improve the security protection capability of BMC system. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating an embodiment of the server BMC dynamic security authentication and firmware protection method based on hardware root of trust in this application. Figure 2 This is a schematic diagram of a structural embodiment of the server BMC dynamic security authentication and firmware protection system based on hardware root of trust in this application.

[0019] Explanation of icon numbers: 10. Memory; 20. Processor.

[0020] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of this application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0022] It should be understood that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0023] In existing technologies, server remote management scenarios widely rely on baseboard management controllers for device monitoring and maintenance. Traditional baseboard management controllers use static password authentication mechanisms, which are vulnerable to default password leaks and cannot defend against phishing attacks. Firmware integrity verification only supports full-image hash verification, which is insufficient for detecting certain tampering attacks. Access control policies are based on network address filtering and lack dynamic authentication mechanisms, leading to frequent unauthorized access vulnerabilities.

[0024] To address these issues, researchers found that static authentication mechanisms are ill-suited for dynamic security environments, necessitating the introduction of authentication factors based on unique device identifiers. Firmware protection needs to cover the entire lifecycle of transmission, storage, and operation; a single verification method is insufficient to handle complex attack scenarios. The lack of a hardware root of trust makes security configurations vulnerable to tampering, requiring the establishment of an end-to-end trusted execution environment.

[0025] Based on this, the embodiments of this application provide a server BMC dynamic security authentication and firmware protection method based on hardware root of trust, referring to... Figure 1 The server BMC dynamic security authentication and firmware protection method based on hardware root of trust includes steps S100 to S400, wherein: Step S100: Obtain the unique identifier data and hardware version data of the server motherboard through the security chip, generate device fingerprint data based on the unique identifier data and hardware version data, and store the device fingerprint data in the tamper-proof storage area of ​​the security chip. Step S200: Perform dynamic authentication processing based on device fingerprint data: Generate a first random number and send it to the management platform through the out-of-band channel of the management interface; receive signature data containing a second random number and timestamp data returned by the management platform; verify the legality of the signature data using the security chip and device fingerprint data; if the verification is successful, generate session key data based on the elliptic curve Diffie-Hellman algorithm; upload the hash value of the session key data to the blockchain network for storage. Step S300: Firmware protection processing based on session key data: During the firmware transmission phase, the firmware update package data is fragmented and encrypted to obtain encrypted fragment data; checksum data of the encrypted fragment data is generated based on the SHA-3 hash algorithm; when receiving firmware update package data, the integrity of the checksum data is verified piece by piece; during the firmware storage phase, incremental update log data is generated based on the firmware update package data, and the incremental update log data is stored in the secure partition of the BMC flash memory; abnormal changes in the incremental update log data are monitored based on the differential hash algorithm; during the firmware operation phase, a trusted execution environment is established through the security chip and the baseline hash value data is pre-stored; periodic memory integrity verification processing is performed on the critical processes of the BMC kernel to obtain memory verification result data; if the memory verification result data is abnormal, the management interface is frozen and an alarm notification is triggered; Step S400: Based on the security chip, digitally sign the BMC user permission configuration data to obtain the signature configuration data; when a permission change request data is received, obtain the hardware key and dynamic password to form two-factor authentication data and perform authentication processing; collect security event log data in real time, analyze the security event log data based on the anomaly detection model, and trigger the automatic locking mechanism.

[0026] In this embodiment, device fingerprint data refers to a unique identity credential generated by combining the server motherboard's physical identifier and hardware version. Specifically, it can be implemented by hashing the motherboard serial number and version number using the encryption engine of the security chip to establish an unforgeable device identity. Dynamic authentication processing refers to achieving dynamic identity confirmation through bidirectional random number exchange and timestamp verification. Specifically, it can use the elliptic curve digital signature algorithm to verify the management platform's identity, eliminating the risk of static credential leakage. Fragmented encryption processing refers to dividing the firmware data into multiple encryption units. Specifically, it can use a session key to implement AES-GCM encryption on each data block to prevent overall data leakage during transmission. Incremental update log data refers to audit information recording firmware version differences. Specifically, it can use a binary difference algorithm to generate version change records, facilitating the detection of unauthorized tampering. Trusted execution environment refers to an isolated secure computing area. Specifically, it can use the TEE technology of the security chip to create a protected memory space, ensuring that the execution of critical processes is not interfered with.

[0027] In this embodiment, the security chip collects the motherboard's unique identifier information to generate a device fingerprint and stores it in an tamper-proof area, forming a hardware-level trust anchor. During dynamic authentication, a random number is generated and sent to the management platform. A signature response containing a timestamp is received, and the platform's identity is confirmed by verifying the signature's legality. After successful authentication, a session key is negotiated based on an elliptic curve algorithm, and the key hash value is written to the blockchain to achieve an immutable record. During firmware updates, data packets are encrypted in blocks. Each encrypted block generates an independent checksum to construct a Merkle tree. The receiving end verifies the hash value block by block to ensure transmission integrity. During the storage phase, firmware version difference logs are recorded, and a sliding window algorithm is used to calculate the log fingerprint in real time. Anomalies are detected by comparing the fingerprint with a baseline value. During the runtime phase, a baseline hash is loaded in a trusted environment, and critical process memory areas are periodically scanned. Code tampering is detected using a fast hash algorithm. When permissions change, two-factor authentication is performed using a hardware key and a dynamic password, and an automatic locking mechanism is triggered by real-time analysis of security logs.

[0028] Compared to existing technologies, dynamic authentication mechanisms replace static password authentication, achieving replay attack resistance through random number exchange and blockchain storage. Sharding encryption and Merkle tree verification, compared to full-mirror hash verification, can accurately locate tampered data blocks. Incremental log monitoring technology significantly reduces storage overhead and improves detection sensitivity compared to traditional full-scale comparison methods. The establishment of a trusted execution environment overcomes the limitations of software-layer protection, achieving hardware-level memory protection. Two-factor authentication combined with an anomaly detection model forms a multi-dimensional defense system. Through the above technical solutions, this application effectively solves the problem of traditional authentication mechanisms being easily cracked, achieving dynamic identity verification and secure session key negotiation. Firmware protection covers the entire process of transmission encryption, storage monitoring, and runtime verification, significantly improving the ability to resist complex attacks. The introduction of a hardware root of trust ensures that secure configurations cannot be tampered with, and the anomaly detection mechanism promptly blocks potential threats, forming a complete server management controller security protection system.

[0029] In one feasible implementation, the steps for dynamic authentication based on device fingerprint data include: generating a first random number using a security chip; sending the first random number to a management platform via an out-of-band channel; receiving signature data returned by the management platform, which is generated by the management platform using its security chip to sign the first random number, a second random number, and timestamp data; verifying the legality of the signature data using the security chip and the device fingerprint data; and if the verification is successful, generating session key data based on the first and second random numbers using the elliptic curve Diffie-Hellman algorithm, and uploading the hash value of the session key data to the blockchain network.

[0030] In this embodiment, the out-of-band channel refers to a management communication channel independent of the server's business network. Specifically, it can be implemented using a dedicated network interface of the baseboard management controller to avoid data transmission interference from business network attacks. The Elliptic Curve Diffie-Hellman algorithm is a key exchange protocol based on elliptic curve cryptography. Specifically, it can be implemented using NIST P-256 curve parameters, ensuring the security of temporary session key generation through the discrete logarithm problem. The blockchain network refers to a distributed ledger storage system, specifically implemented using a consortium blockchain architecture, ensuring the immutability of key hash values ​​through a multi-node consensus mechanism.

[0031] In this embodiment, during the dynamic authentication process, the security chip first generates a cryptographically strong random number as a temporary session parameter, which is transmitted to the remote management platform via a physically isolated out-of-band communication link. Upon receiving the random number, the management platform combines it with its own generated second random number and a precise timestamp, and performs a digital signature operation using a hardware security module. The server-side security chip verifies the signature validity using a pre-set device fingerprint. After confirming the legitimate identities of both communicating parties, it performs a key negotiation operation based on elliptic curve cryptography, using the two random numbers as input parameters, to generate an encryption key valid only for this session. Finally, the key hash value is written to the blockchain node, forming an audit record with timestamps and multi-party verification features.

[0032] In this embodiment, dynamically generated random numbers are used as the basic parameters of the session, combined with a timestamp signature verification mechanism, which can effectively resist replay attacks and man-in-the-middle attacks. A one-time key negotiation method is used instead of a fixed certificate, eliminating the security risks associated with long-term key storage. At the same time, blockchain technology further enhances the traceability and non-repudiation of key usage records, thus realizing a strong authentication mechanism based on dynamic random parameters and solving the security risks posed by the ease of counterfeiting and long-term validity of traditional static certificates.

[0033] In one feasible implementation, the step of generating session key data based on a first random number and a second random number using the elliptic curve Diffie-Hellman algorithm, and uploading the hash value of the session key data to the blockchain network, includes: generating initial key data based on the elliptic curve Diffie-Hellman algorithm and the first and second random numbers; processing the initial key data using a key derivation function to obtain derived key data; and outputting the derived key data as session key data to the blockchain network.

[0034] In this embodiment, the elliptic curve Diffie-Hellman algorithm refers to a key exchange protocol based on the elliptic curve discrete logarithm problem, specifically implemented using the secp256k1 curve parameters, used for secure negotiation of shared keys in insecure channels. The key derivation function is an algorithm that expands the initial key into a more complex key sequence, specifically implemented using the HKDF standard, enhancing the entropy value of the derived key through salt and context information. The derived key data refers to the session key generated after key expansion processing, specifically using the AES-256 algorithm as the encryption basis, with a length of 256 bits to meet high-strength encryption requirements. The blockchain network refers to a distributed ledger storage system, specifically implemented using the Hyperledger Fabric framework, ensuring the immutability of the key hash value through a consensus mechanism.

[0035] In this embodiment, during the dynamic authentication process, an initial shared key is calculated using the elliptic curve Diffie-Hellman algorithm with a first random number and a second random number as input parameters. Subsequently, a key derivation function is applied to iteratively process the initial key, generating a derived key with higher complexity using a pseudo-random function. This derived key is output to the blockchain node as the final session key, and its hash value is written to the immutable ledger of the blockchain via a smart contract. For example, the iteration count can be set to 10,000 during the key derivation stage, generating a 512-bit intermediate key using the HMAC-SHA512 function, and finally truncating the first 256 bits as the valid session key.

[0036] In this embodiment, by dynamically generating an initial key and superimposing key derivation processing, the key for each session is made unique and unpredictable. Simultaneously, the blockchain storage mechanism eliminates the single point of failure risk of centralized storage, ensuring that key verification records cannot be maliciously tampered with. Thus, this application solves the technical defect of traditional static keys being easily cracked. The dynamic key derivation mechanism significantly improves the randomness and complexity of the key generation process. Combined with the distributed storage characteristics of blockchain, a complete key lifecycle management chain is formed, ensuring the independence and traceability of keys for each session.

[0037] In one feasible implementation, the steps of performing fragmented encryption processing on the firmware update package data during the firmware transmission stage to obtain encrypted fragmented data; generating checksum data for each encrypted fragmented data based on the SHA-3 hash algorithm; and verifying the integrity of the checksum data piece by piece when receiving firmware update package data include: performing fragmented encryption processing on the firmware update package data using session key data to obtain encrypted fragmented data; generating checksum data for each encrypted fragmented data based on the SHA-3 hash algorithm and constructing a Merkle tree structure to store the checksum data; and extracting the encrypted fragmented data piece by piece and verifying whether its checksum data matches the Merkle tree structure when receiving firmware update package data to verify the integrity of the checksum data.

[0038] In this embodiment, fragmented encryption refers to dividing the firmware update package into multiple data blocks and encrypting them independently. Specifically, this can be achieved using the AES-GCM algorithm combined with dynamically generated session keys. Independent encryption of each fragment prevents man-in-the-middle attacks from obtaining the complete firmware image. The SHA-3 hash algorithm refers to the third-generation secure hash algorithm standard. Specifically, a variant of SHA3-512 can be used to generate a 512-bit checksum, and its collision resistance ensures the unique identification of fragmented data. The Merkle tree structure refers to a binary tree-like hash value aggregation structure. Specifically, the root hash can be generated by merging the hash values ​​of child nodes layer by layer. This structure allows the integrity of a specific fragment to be verified using a small number of hash values.

[0039] In this embodiment, during the transmission of the firmware update package, the firmware data is first encrypted in blocks using a dynamically negotiated session key. For example, a 1GB firmware file is divided into 1024 1MB segments and encrypted separately. Each encrypted segment generates an independent checksum using the SHA-3 algorithm. All checksums are used to construct a Merkle tree based on a binary tree structure, and the root hash value is stored in the secure chip. When the receiving end obtains an encrypted segment, it compares the segment hash value with the hash chain of the corresponding node in the Merkle tree. For example, to verify segment number 512, it only needs to provide the hash of that segment and the hash paths of its sibling nodes to the root node to confirm that the segment has not been tampered with during transmission.

[0040] Understandably, traditional firmware hash verification only performs a single hash check on the entire image, making it unable to locate modified fragments when subjected to differential tampering attacks. This solution, however, uses fragment encryption and a Merkle tree structure to not only achieve fragment-by-fragment integrity verification but also precisely locate tampered data blocks. For example, if a fragment's hash value is detected to be mismatched with the Merkle tree, the decryption process for that fragment can be immediately terminated. Furthermore, the use of dynamic session keys to encrypt fragments provides more effective protection against replay attacks compared to static certificate encryption schemes.

[0041] In one feasible implementation, the steps of generating incremental update log data based on firmware update package data during the firmware storage stage, storing the incremental update log data in a secure partition of the BMC flash memory, and monitoring abnormal changes in the incremental update log data based on a differential hash algorithm include: generating incremental update log data based on firmware update package data; dividing a secure partition in the BMC flash memory to store the incremental update log data; applying a differential hash algorithm to monitor changes in the incremental update log data in real time; and if an abnormality in the incremental update log data is detected, generating alarm data and pausing firmware loading.

[0042] In this embodiment, incremental update log data refers to a data structure that only records the differences between the firmware update package and the previous version. Specifically, it can be generated using a binary difference comparison algorithm to reduce storage space usage and improve tamper detection efficiency. The secure partition of the BMC flash memory refers to a protected storage area divided through hardware isolation mechanisms. Specifically, it can employ access control policies implemented by the flash controller to prevent unauthorized processes from modifying data. The differential hash algorithm is an algorithm that calculates the hash differences of data blocks using a sliding window mechanism. Specifically, it can employ a fingerprint generation method based on rolling hashing to achieve rapid detection of local data tampering.

[0043] In this embodiment, after the firmware update package completes transmission verification, the system generates incremental update log data containing only the modified parts using a binary difference comparison tool. This log data is written to a secure partition in the BMC flash memory, isolated by a hardware access control policy, such as setting up an independent storage block and configuring read-only permissions. The system continuously runs a differential hash monitoring process, for example, performing a sliding window scan of the log data every 5 seconds to generate the current data fingerprint and compare it with a baseline fingerprint. When an unexpected hash difference value is detected exceeding a preset threshold, such as three consecutive data blocks having mismatched hash values, the system immediately interrupts the firmware loading process and triggers an audible and visual alarm signal.

[0044] In this embodiment, by combining incremental log monitoring with a differential hash algorithm, local data tampering can be accurately identified, solving the technical problem of easy tampering and low detection efficiency of update logs during the firmware storage stage. Incremental storage reduces storage pressure, differential hash algorithm improves tamper detection sensitivity, and hardware protection mechanisms for secure partitions block unauthorized modifications, forming a multi-layered firmware storage protection system.

[0045] In one feasible implementation, the step of using the differential hash algorithm to monitor changes in incremental update log data in real time includes: processing the incremental update log data based on the sliding window algorithm to generate current fingerprint data; performing a differential comparison between the current fingerprint data and the pre-stored baseline fingerprint data; and determining that the incremental update log data is abnormal if the difference value exceeds a preset threshold rule data.

[0046] In this embodiment, the sliding window algorithm refers to a sliding window technique that processes incremental update log data into blocks of fixed length. Specifically, it can use a window of 512 bytes per block for data slicing, achieving continuous coverage through a sliding step size. This technique can effectively capture local data change characteristics. The differential hash algorithm is a method for calculating the hash value difference between adjacent data blocks. Specifically, it can use the SHA-256 algorithm to generate the hash value of each data block and perform an XOR operation, identifying abnormal modifications by quantifying the difference value. The preset threshold rule data refers to a pre-set tolerance range for differences. Specifically, it can employ a dynamic threshold mechanism, such as setting the threshold based on the historical data statistics, with the mean plus three standard deviations. This mechanism can adapt to normal fluctuations in different scenarios.

[0047] In this embodiment, the incremental update log data is divided into contiguous data blocks. Each data block generates a corresponding hash value as the current fingerprint data through a sliding window algorithm. The system compares the current fingerprint with the baseline fingerprint block by block, calculating the difference between their hash values. When the difference exceeds a preset threshold, it is determined that the log data has been modified without authorization. For example, during the firmware storage phase, if the hash difference value of a data block in the incremental log is detected to reach the threshold, the system will immediately pause the firmware loading process and generate an alarm. This process achieves fine-grained monitoring through a differential comparison mechanism, significantly reducing computational overhead compared to the traditional full-scale verification method.

[0048] In this embodiment, the incremental update log is monitored in real time using a differential hash algorithm, which can capture minute data changes. For example, the protection mechanism can be triggered if an abnormal offset of a single function address is detected in the firmware log. This can effectively identify covert tampering attacks during the firmware storage stage, such as malicious code injection behavior targeting security patch update logs.

[0049] In one feasible implementation, during the firmware runtime phase, a trusted execution environment is established through a security chip and a baseline hash value is pre-stored. Periodic memory integrity checks are performed on critical processes of the BMC kernel to obtain memory check result data. If the memory check result data is abnormal, the management interface is frozen and an alarm notification is triggered. This includes: establishing a trusted execution environment through the security chip and loading the pre-stored baseline hash value data; periodically scanning critical processes of the BMC kernel to obtain memory region data; extracting critical function segment data from the memory region data based on a preset critical function segment address range; calculating the hash value of the critical function segment data and comparing it with the baseline hash value data to generate memory check result data.

[0050] In this embodiment, the Trusted Execution Environment (TEX) refers to a secure execution space created through the hardware isolation mechanism of a security chip. Specifically, it can be implemented using ARM TrustZone or Intel SGX technology to isolate critical code from interference by external malicious code. The baseline hash value data refers to the pre-calculated and stored hash values ​​of legitimate process memory regions, specifically generated using the SHA-256 algorithm, serving as a baseline reference for integrity verification. Periodic memory integrity verification processing refers to scanning memory data at preset time intervals, specifically implemented using a timer trigger mechanism, for real-time detection of runtime tampering. Critical processes refer to core functional modules in the BMC kernel involved in system management and network communication, specifically including the IPMI protocol processing module or the Redfish interface service process. In this embodiment, during the firmware execution phase, a trusted execution environment is defined using the hardware isolation function of the security chip, and the pre-calculated baseline hash value of the legitimate memory region of the critical process is loaded into this environment. The system periodically triggers a memory scan operation to extract critical function segment data from the memory region of the BMC kernel, for example, by parsing the process memory mapping table to determine the function address range. The extracted data is hashed to generate a real-time hash value, which is then compared with the pre-stored baseline hash value. If a hash value mismatch is detected, it is determined that memory tampering has occurred, the network communication function of the management interface is immediately frozen, and an encrypted alarm notification is sent to the management platform through the out-of-band channel. In this embodiment, by periodically verifying the memory integrity of critical processes and combining it with the hardware protection mechanism of the Trusted Execution Environment (TEE), malicious code injection behavior can be identified in real time. This allows for precise location of critical function segments and high-frequency verification, improving the efficiency of runtime attack interception and achieving dynamic protection of BMC runtime memory data. This effectively prevents malicious code from residing and spreading during system operation. The collaborative mechanism of the TEE and periodic verification solves the technical shortcomings of traditional solutions, such as the inability to detect partial tampering and the lack of real-time protection. Furthermore, the use of a hardware root of trust ensures the security of the verification process.

[0051] In one feasible implementation, the steps of calculating the hash value of the key function segment data and comparing it with the baseline hash value data to generate memory verification result data include: extracting the key function segment data; applying a fast hash algorithm to calculate the real-time hash value data of the key function segment data; comparing the real-time hash value data with the pre-stored baseline hash value data, and determining that the memory verification result data is abnormal if they are inconsistent.

[0052] In this embodiment, critical function segment data refers to specific memory region data of critical processes in the BMC kernel. Specifically, it can be implemented using memory address mapping technology to extract the function entry addresses and code segment length ranges of the process, used to locate the core code regions that need to be verified. The fast hash algorithm refers to a hash function based on low computational complexity, specifically XXH64 or FarmHash algorithms, used to quickly generate hash values ​​for memory data under limited computing resources. Real-time hash value data refers to the instantaneous hash calculation result of the memory content of critical function segments in the current running state. Specifically, it can be implemented by obtaining the real-time memory mapping data of the process through memory image snapshot technology and performing hash operations, used to reflect the real-time integrity status of the memory code. Baseline hash value data refers to the hash values ​​of legitimate function segments pre-stored in the trusted execution environment. Specifically, it can be implemented by the security chip performing hash calculations on the verified original code during the trusted boot phase and encrypting and storing it, serving as a reference standard for integrity verification.

[0053] In this embodiment, during the firmware execution phase, a pre-stored baseline hash value is loaded into the trusted execution environment established by the security chip. The system periodically scans the critical processes of the BMC kernel, obtains their memory region data, and extracts the corresponding code segment data based on the preset critical function segment address range. After extracting the critical function segment data, a fast hash algorithm is applied to calculate its real-time hash value, which is then compared with the baseline hash value. If the two are inconsistent, the memory verification result is determined to be abnormal, triggering the management interface freeze and alarm mechanism. This process achieves high-frequency verification by processing critical function segment data in stages, combined with a fast hash algorithm, while relying on the trusted execution environment to ensure the immutability of the baseline data.

[0054] In this embodiment, by dividing key function segments and implementing periodic fast hash calculations, abnormal code regions can be accurately located. For example, when an attacker injects malicious code fragments, even if the overall firmware hash remains unchanged, the real-time hash value of a specific function segment will deviate, thus solving the problem of low efficiency in runtime memory tampering detection and achieving fine-grained real-time monitoring of critical processes. The combination of fast hashing algorithms and key function segment location technology enables rapid identification of malicious behaviors such as injection attacks and code hijacking while maintaining system performance stability. Based on benchmark data stored in the trusted execution environment, the security of the verification reference system is ensured, preventing attackers from circumventing detection by tampering with the benchmark value. Furthermore, the automatic freezing mechanism triggered after anomaly detection can promptly block the spread of potential attack chains, improving the security of the BMC operating environment.

[0055] In one feasible implementation, the steps of real-time collection of security event log data, analysis of security event log data based on an anomaly detection model, and triggering an automatic locking mechanism include: real-time collection of security event log data from system logs; analysis of security event log data based on a Bayesian classifier model and preset threshold rules to detect abnormal behavior; and if abnormal behavior is detected, generation of locking command data to trigger the automatic locking mechanism.

[0056] In this embodiment, security event log data refers to the recorded data generated during the operation of the BMC system, including user logins, permission changes, and firmware operation behaviors. Specifically, it can be implemented by real-time collection of system kernel logs and audit logs using the syslog protocol, serving as the data foundation for anomaly detection. The Bayesian classifier model refers to a classification algorithm model built based on probability and statistics theory. Specifically, it can be implemented using the Naive Bayes algorithm combined with a historical security event training dataset, used to dynamically identify abnormal behavior patterns in the logs. Threshold rule data refers to a set of preset abnormal behavior judgment conditions, specifically implemented using a combination of multi-dimensional rules, such as a threshold for the number of permission changes within a single hour or a threshold for the frequency of cross-region login attempts, used to quantify the judgment criteria for abnormal behavior. The automatic locking mechanism refers to a security control strategy that restricts access to the management interface, specifically implemented using dynamic updates of firewall policies or forced termination of session connections, used to immediately block potential attacks when anomalies are detected.

[0057] In this embodiment, the system continuously captures BMC kernel events and user operation records through a log collection module, forming a structured security event log data stream. This data stream is input into a Bayesian classifier model for real-time analysis. The model calculates the deviation between the event feature vector and the probability distribution of normal behavior, and makes a comprehensive judgment based on preset threshold rules. For example, when the system detects that the same user account triggers 5 permission change operations within 10 minutes and the IP address of the operation is abnormal, it will determine this as abnormal behavior. At this time, the security control module will generate a locking command, immediately freezing the corresponding account permissions and closing the relevant management ports by calling the BMC underlying interface.

[0058] In some specific implementations, the log acquisition module can be configured to simultaneously collect hardware sensor data and firmware operation logs. The Bayesian classifier model can load different versions of the training dataset to adapt to the security policies of a specific environment. Threshold rule data can employ a dynamic adjustment mechanism, such as automatically adjusting alarm sensitivity based on system load status. The automatic locking mechanism can be set to a tiered response mode, such as a temporary lock triggered by the first anomaly and a permanent lock triggered by repeated anomalies.

[0059] In this embodiment, dynamic behavior modeling is achieved through a Bayesian classifier model, which can adaptively learn normal operating patterns and significantly improve the detection capability of unknown threats. Simultaneously, the synergistic effect of threshold rule data and automatic locking mechanism solves the problem of lag in response in traditional solutions, achieving closed-loop control from anomaly detection to security protection. Thus, this application realizes intelligent monitoring and real-time response to security events in the BMC system, effectively identifying malicious behaviors disguised as normal operations, promptly blocking unauthorized access and configuration tampering, and improving the security protection level of the server management and control unit.

[0060] In the embodiments of this application, the server BMC dynamic security authentication and firmware protection method based on hardware root of trust achieves dynamic authentication by generating device fingerprints through a security chip. It combines fragmented encryption, differential hash verification and memory integrity verification to build a multi-layer protection mechanism. At the same time, it uses blockchain to store key hash values ​​to enhance data credibility. It effectively solves the problems of static authentication risk, insufficient firmware tampering detection and lack of trust chain in traditional solutions, and can improve the security protection capability of the BMC system.

[0061] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the server BMC dynamic security authentication and firmware protection method based on hardware root of trust in this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0062] This application also provides a server BMC dynamic security authentication and firmware protection system based on a hardware root of trust, see reference. Figure 2 The hardware root-based server BMC dynamic security authentication and firmware protection system includes: a memory 10, a processor 20, and a hardware root-based server BMC dynamic security authentication and firmware protection program stored on the memory 10 and executable on the processor 20. The hardware root-based server BMC dynamic security authentication and firmware protection program is configured to implement the steps of the hardware root-based server BMC dynamic security authentication and firmware protection method.

[0063] The hardware root-based server BMC dynamic security authentication and firmware protection system provided in this application, employing the hardware root-based server BMC dynamic security authentication and firmware protection method described in the above embodiments, can enhance the security protection capabilities of the BMC system. Compared with the prior art, the beneficial effects of the hardware root-based server BMC dynamic security authentication and firmware protection system provided in this application are the same as those of the hardware root-based server BMC dynamic security authentication and firmware protection method provided in the above embodiments, and other technical features of the hardware root-based server BMC dynamic security authentication and firmware protection system are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0064] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0065] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. All equivalent structural transformations made under the technical concept of this application using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included within the scope of patent protection of this application.

Claims

1. A method for dynamic security authentication and firmware protection of server BMC based on hardware root of trust, characterized in that, The method includes: The unique identifier data and hardware version data of the server motherboard are obtained through the security chip. Device fingerprint data is generated based on the unique identifier data and hardware version data, and the device fingerprint data is stored in the tamper-proof storage area of ​​the security chip. Dynamic authentication processing based on device fingerprint data: Generate a first random number and send it to the management platform through the out-of-band channel of the management interface; receive signature data containing a second random number and timestamp data returned by the management platform; verify the legality of the signature data using the security chip and device fingerprint data; if the verification is successful, generate session key data based on the elliptic curve Diffie-Hellman algorithm; upload the hash value of the session key data to the blockchain network for storage. Firmware protection based on session key data: During firmware transmission, the firmware update package data is fragmented and encrypted to obtain encrypted fragment data; checksum data of the encrypted fragment data is generated based on the SHA-3 hash algorithm; when receiving firmware update package data, the integrity of the checksum data is verified piece by piece; during firmware storage, incremental update log data is generated based on the firmware update package data and stored in the secure partition of the BMC flash memory; abnormal changes in the incremental update log data are monitored based on the differential hash algorithm; during firmware operation, a trusted execution environment is established through a security chip and a base hash value is pre-stored; periodic memory integrity checks are performed on critical processes of the BMC kernel to obtain memory checksum data; if the memory checksum data is abnormal, the management interface is frozen and an alarm notification is triggered. The security chip performs digital signature processing on BMC user permission configuration data to obtain the signature configuration data; when a permission change request is received, the hardware key and dynamic password are used to form two-factor authentication data and perform authentication processing; security event log data is collected in real time, and the security event log data is analyzed based on the anomaly detection model to trigger an automatic locking mechanism.

2. The server BMC dynamic security authentication and firmware protection method based on hardware root of trust as described in claim 1, characterized in that, The steps for dynamic authentication based on device fingerprint data include: The first random number is generated using the security chip; The first random number is sent to the management platform via an out-of-band channel; Receive signature data returned by the management platform. This signature data is generated by the management platform using its security chip to sign the first random number, the second random number, and the timestamp data. The signature data is verified using a security chip and device fingerprint data. If the verification is successful, the session key data is generated using the elliptic curve Diffie-Hellman algorithm based on the first and second random numbers, and the hash value of the session key data is uploaded to the blockchain network.

3. The server BMC dynamic security authentication and firmware protection method based on hardware root of trust as described in claim 2, characterized in that, The steps of generating session key data based on a first random number and a second random number using the elliptic curve Diffie-Hellman algorithm, and uploading the hash value of the session key data to the blockchain network, include: Initial key data is generated based on the elliptic curve Diffie-Hellman algorithm and the first and second random numbers. The initial key data is processed using the key derivation function to obtain the derived key data; The derived key data is output to the blockchain network as session key data.

4. The server BMC dynamic security authentication and firmware protection method based on hardware root of trust as described in claim 1, characterized in that, During the firmware transmission phase, the firmware update package data is fragmented and encrypted to obtain encrypted fragment data. Generate verification data for encrypted fragmented data based on the SHA-3 hash algorithm; When receiving firmware update package data, the steps for verifying the integrity of the checksum data piece by piece include: Use session key data to perform fragmented encryption on firmware update package data, and obtain encrypted fragmented data; Verification value data is generated for each encrypted data fragment based on the SHA-3 hash algorithm, and a Merkle tree structure is constructed to store the verification value data; When receiving firmware update package data, the encrypted fragment data is extracted piece by piece and its checksum data is verified to match the Merkle tree structure in order to verify the integrity of the checksum data.

5. The server BMC dynamic security authentication and firmware protection method based on hardware root of trust as described in claim 1, characterized in that, The steps of generating incremental update log data based on firmware update package data during the firmware storage phase, storing the incremental update log data in the security partition of BMC flash memory, and monitoring abnormal changes in the incremental update log data based on a differential hash algorithm include: Incremental update log data is generated based on firmware update package data; Divide the BMC flash memory into a security partition to store incremental update log data; The differential hashing algorithm is used to monitor changes in incremental update log data in real time. If an anomaly is detected in the incremental update log data, an alert is generated and firmware loading is paused.

6. The server BMC dynamic security authentication and firmware protection method based on hardware root of trust as described in claim 5, characterized in that, The steps for using the differential hash algorithm to monitor changes in incremental update log data in real time include: The incremental update log data is processed based on the sliding window algorithm to generate the current fingerprint data. Perform a differential comparison between the current fingerprint data and the pre-stored baseline fingerprint data; If the difference value exceeds the preset threshold rule data, the incremental update log data is determined to be abnormal.

7. The server BMC dynamic security authentication and firmware protection method based on hardware root of trust as described in claim 1, characterized in that, During the firmware execution phase, a trusted execution environment is established through a security chip and baseline hash value data is pre-stored. Periodic memory integrity verification is performed on key processes of the BMC kernel to obtain memory verification result data. If the memory verification result is abnormal, the steps to freeze the management interface and trigger an alarm notification include: A trusted execution environment is established using a secure chip, and pre-stored baseline hash value data is loaded. Periodically scan key processes in the BMC kernel to obtain memory region data; Based on a preset range of key function segment addresses, extract key function segment data from memory region data; Calculate the hash value of the key function segment data and compare it with the baseline hash value data to generate memory verification result data.

8. The server BMC dynamic security authentication and firmware protection method based on hardware root of trust as described in claim 7, characterized in that, The step of calculating the hash value of the key function segment data and comparing it with the baseline hash value data to generate memory verification result data includes: Extract key function segment data; The fast hash algorithm is applied to calculate the real-time hash value of key function segment data; The real-time hash value data is compared with the pre-stored baseline hash value data. If they are inconsistent, the memory verification result data is determined to be abnormal.

9. The server BMC dynamic security authentication and firmware protection method based on hardware root of trust as described in claim 1, characterized in that, The steps of real-time collection of security event log data, analysis of the security event log data based on the anomaly detection model, and triggering of the automatic locking mechanism include: Collect security event log data in real time from the system log; Security event log data is analyzed based on a Bayesian classifier model and preset threshold rules to detect abnormal behavior. If abnormal behavior is detected, a lock command is generated to trigger the automatic locking mechanism.

10. A server BMC dynamic security authentication and firmware protection system based on hardware root of trust, characterized in that, The hardware root-based server BMC dynamic security authentication and firmware protection system includes: a memory, a processor, and a hardware root-based server BMC dynamic security authentication and firmware protection program stored on the memory and executable on the processor. The hardware root-based server BMC dynamic security authentication and firmware protection program is configured to implement the steps of the hardware root-based server BMC dynamic security authentication and firmware protection method as described in any one of claims 1 to 9.

Images