Generating the identity of a computing device using physically unclonable functions

Abstract

A method comprises: generating a device secret by a computing device, the generating comprising: providing at least one value by at least one physically unclonable function (PUF); and generating the device secret using a key derivation function (KDF), wherein the at least one value provided by the at least one PUF is an input to the KDF; and storing the generated device secret in a memory of the computing device.

Description

[0001] Related Applications

[0002] This application claims priority to U.S. patent application serial number 16 / 363,204, filed on March 25, 2019, and entitled “GENERATING AN IDENTITY FOR A COMPUTING DEVICE USING A PHYSICALUNCLONABLE FUNCTION,” the entire disclosure of which is hereby incorporated herein by reference.

[0003] This application is related to U.S. non-provisional application serial number 15 / 970,660, entitled “KEY GENERATION AND SECURE STORAGE IN A NOISY ENVIRONMENT,” filed by Pisasale et al. on May 3, 2018, the entire contents of which are incorporated by reference as if fully set forth herein.

[0004] This application is related to U.S. non-provisional application serial number 15 / 853,498, entitled “PHYSICAL UNCLONABLE FUNCTION USING MESSAGE AUTHENTICATION CODE,” filed by Mondello et al. on December 22, 2017, the entire contents of which are incorporated by reference as if fully set forth herein.

[0005] This application is related to U.S. non-provisional application serial number 15 / 965,731, entitled “SECURE DISTRIBUTION OF SECRET KEY USING A MONOTONIC COUNTER,” filed by Mondello et al. on April 27, 2018, the entire contents of which are incorporated by reference as if fully set forth herein. Technical Field

[0006] At least some embodiments disclosed herein relate generally to identities of computing devices, and more particularly, but not limited to, generating identities of computing devices using physically unclonable functions. Background Art

[0007] A Physical Unclonable Function (PUF) provides a digital value that can serve as a unique identity for a semiconductor device, such as a microprocessor, for example. For example, a PUF is based on physical changes that occur naturally during semiconductor manufacturing and permit differentiation between otherwise identical semiconductor chips.

[0008] PUFs are commonly used in cryptography. A PUF may be a physical entity, for example, embodied in a physical structure. PUFs are often implemented in integrated circuits and are often used in applications with high security requirements. For example, PUFs can be used as unique and unalterable device identifiers. PUFs can also be used for secure key generation and as a source of randomness.

[0009] In an example related to device identification, The IoT platform is a collection of cloud services provided by Microsoft. The IoT platform supports the Device Identity Composition Engine (DICE) and many different kinds of Hardware Security Modules (HSM). DICE is an upcoming standard at the Trusted Computing Group (TCG) for device identification and authentication that enables manufacturers to use silicon gates to establish device identities that reside in hardware. HSMs are used to ensure device identity and provide advanced functionality such as hardware-based device authentication and zero touch provisioning.

[0010] DICE provides a scalable security framework that uses the HSM footprint to anchor trust for building security solutions such as authentication, secure boot, and remote attestation. DICE can be used to characterize the current environment of constrained computing for IoT devices and provides an alternative to more traditional security framework standards such as the Trusted Computing Group (TCG) and Trusted Platform Module (TPM). The IoT platform has HSM support for DICE in HSMs from some silicon vendors.

[0011] In one example related to trust services, the Robust Internet of Things (RIoT) is an architecture for providing trust services to computing devices. Trust services include device identity, authentication, and data integrity. The RIoT architecture can be used to remotely re-establish trust in a device that has been compromised by malware. Furthermore, RIoT services can be provided on even very small devices at low cost.

[0012] Improved security technologies have created a need for more frequent software updates for products in the field. However, these updates must be managed and verified without human intervention. RIoT can be used to solve these technical problems.

[0013] RIoT provides the foundation for cryptographic operations and key management for many security scenarios. Authentication, integrity verification, and data protection require cryptographic keys to encrypt and decrypt, as well as mechanisms to hash and sign data. Most networked devices also use cryptography to secure communications with other devices.

[0014] Cryptographic services provided by RIoT include device identity, data protection, and authentication. Regarding device identity, a device typically authenticates itself by proving possession of encryption keys. If the keys associated with a device are extracted and cloned, the device may be counterfeit.

[0015] Regarding data protection, devices typically use cryptography to encrypt and fully protect locally stored data. If the encryption key is only accessible by authorized code, unauthorized software cannot decrypt or modify the data.

[0016] Regarding authentication, a device sometimes needs to report the code it is running and its security configuration. For example, authentication is used to prove that the device is running the latest code.

[0017] If keys are managed only in software, then bugs in software components can lead to key leaks. For software-only systems, the primary way to restore trust after a key leak is to install updated software and provision new keys for the device. This is time-consuming for servers and mobile devices, and is not possible when the device is physically inaccessible.

[0018] Some methods to ensure remote re-provisioning use hardware-based security. Software-level attacks can allow hackers to use hardware-protected keys but not extract them, so hardware-protected keys are a useful building block for re-provisioning of compromised systems. A Trusted Platform Module, or TPM, is an example of a security module that provides hardware protection for keys and also allows a device to report (certify) the software it is running. Therefore, a compromised TPM-equipped device can securely issue new keys and can provide an attestation report.

[0019] TPMs are widely available on computing platforms (e.g., firmware TPMs using SoC integration and processor mode isolation). However, TPMs are often not feasible. For example, small IoT devices cannot support TPMs without significantly increasing cost and power requirements.

[0020] RIoT can be used to provide device security for small computing devices, but it can also be applied to any processor or computer system. If software components outside the RIoT kernel are compromised, then RIoT provides security patching and re-provisioning. RIoT also uses different methods for encryption key protection. The most protected encryption keys used by the RIoT framework can only be obtained briefly during startup. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Embodiments are illustrated by way of example and not limitation in the figures of the accompanying drawings in which like reference numerals indicate similar elements.

[0022] Figure 1 A host device is shown verifying the identity of a computing device according to one embodiment.

[0023] Figure 2 An example computing system having an identity component and an authentication component is shown according to one embodiment.

[0024] Figure 3 An example computing device of a vehicle is shown according to one embodiment.

[0025] Figure 4 An example host device is shown in communication with an example computing device of a vehicle according to one embodiment.

[0026] Figure 5A An application board is shown that generates an identifier, certificates, and keys for a host device according to one embodiment.

[0027] Figure 5B An example computing system is shown that uses tier segmented booting according to one embodiment.

[0028] Figure 6 An example computing device is shown that generates identifiers, certificates, and keys using an asymmetric generator according to one embodiment.

[0029] Notice

[0030] Figure 7 An authentication component is shown that authenticates the identity of a computing device using a decryption operation according to one embodiment.

[0031] Figure 8 A block diagram is shown of an example process to verify credentials according to one embodiment.

[0032] Fig. 9 A method of verifying the identity of a computing device using an identifier, a certificate, and a key is shown according to one embodiment.

[0033] Fig.10 A system for generating a unique key from the output of a message authentication code (MAC) that receives input from a Physically Unclonable Function (PUF) device is presented according to one embodiment.

[0034] Fig.11 A system for generating a unique key from the output of a MAC that receives input from one or more PUF devices selected by a selector module is shown according to one embodiment.

[0035] Fig.12 A system for generating a unique key from the output of a MAC that receives input from one or more PUF devices and input from a monotonic counter (and / or input from another freshness mechanism such as a NONCE, timestamp, etc.) is shown according to one embodiment.

[0036] Fig.13 A method of generating an output from a MAC using one or more input values ​​provided by one or more PUFs is shown according to one embodiment.

[0037] Fig.14 A system for generating a root key from the output of a MAC that receives input from one or more PUF devices and input from a monotonic counter (and / or input from another freshness mechanism such as a NONCE, timestamp, etc.), and adds additional MACs to generate session keys, according to one embodiment, is shown.

[0038] Fig.15 A computing device for storing an obfuscated key in a non-volatile memory is shown according to one embodiment.

[0039] Fig.16 An example of an intermediate key generated during the obfuscation process according to one embodiment is shown.

[0040] Fig.17 Demonstrating an embodiment of Fig.16 An instance of another intermediate key generated during the obfuscation process.

[0041] Fig.18 A method for generating and storing an obfuscation key in a non-volatile memory according to one embodiment is presented.

[0042] Fig.19 A computing device for generating an initial key based on key injection, obfuscating the initial key, and storing the obfuscated key in a non-volatile memory is shown according to one embodiment.

[0043] Fig. 20 A computing device for generating an identity using a Physically Unclonable Function (PUF) according to one embodiment is shown.

[0044] Fig.21 A system is presented that sends an initial value provided by a monotonic counter of the system for use in determining whether tampering with the system has occurred, according to one embodiment.

[0045] Fig. 22 A method for generating an identity of a computing device using a Physically Unclonable Function (PUF) according to one embodiment is presented. DETAILED DESCRIPTION

[0046] At least some embodiments herein relate to verification of the identity of one or more computing devices. In various embodiments, a host device verifies the identity of a computing device by sending a message to the computing device. The computing device uses the message to generate an identifier, certificate, and key that are sent to the host device. The host device uses the generated identifier, certificate, and key to verify the identity of the computing device.

[0047] Other embodiments relate to generating an identity for a computing device using a physically unclonable function (PUF). In various embodiments described below, the computing device may generate its own identity using at least one PUF before the host device verifies the identity as described above. Various embodiments regarding generating an identity using one or more PUFs are described below in the section entitled "Generating an Identity for a Computing Device Using a PUF".

[0048] In some instances related to verification of identity, the computing device may be a flash memory device. In some instances, flash memory is utilized to add a strong level of security capabilities in a computing system (eg, an application controller of an autonomous vehicle).

[0049] Flash memory is used in many computer systems. There are various types of flash memory available today, including serial NOR, parallel NOR, serial NAND, parallel NAND, e.MMC, UFS, etc. These interfaces are used in most embedded systems across various industries and applications.

[0050] For example, serial NOR is used in a large number of applications such as medical devices, factory automation boards, automotive ECUs, smart meters, and Internet gateways. Given the diversity of chipset architectures (processors, controllers, or SoCs), operating systems, and supply chains used across these applications, flash memory is a common denominator building block in these systems.

[0051] Today's computer system resiliency is often characterized by the location of a root of trust integrated into the device and the solution utilized by the security functions it provides. For more information on the root of trust, see the definition produced by the National Institute of Technology (NIST) in Special Publication 800-164. Existing industry uses different implementations of system-level root of trust, using a mix of hardware and software capabilities, resulting in technical problems of fragmented approaches and confusing security levels. This complex array of options also suffers from the key limitation of how to defend the non-volatile memory in which critical code and data are stored.

[0052] Existing approaches rely on processors and other security elements (e.g., hardware security modules (HSMs)) to provide critical security services to their systems. This has created a security gap in many systems at the lowest level of boot, where discrete flash memory components store system-critical code and data. Flash memory has become a target for many hackers to establish advanced persistent threats (APTs) that can mask themselves from higher-level code and resist removal. In many of these cases, the flash memory is reimaged or overwritten with new malicious code, which can compromise the integrity of the device.

[0053] Various embodiments of the present disclosure related to identity verification provide technical solutions to the above technical problems. In some embodiments, a computing device integrates a hardware-based root of trust into a flash memory device, thereby achieving strong encryption identity and health management of IoT devices. By moving basic security primitives within the memory, it becomes simpler to protect the integrity of the code and data contained in the memory itself. This approach can significantly enhance system-level security while minimizing the complexity and cost of the implementation.

[0054] In one embodiment, the new IoT device management capability utilizes flash memory by using flash memory and associated software to implement Device configuration onboarding and management by the IoT cloud. In one example, the solution provides a cryptographic identity that becomes the basis for key device provisioning services, such as the Azure IoT Hub Device Provisioning Service (DPS). In one instance, this DPS, along with enabled storage, enables zero-touch configuration of devices to the correct IoT hub, as well as other services.

[0055] In some embodiments, to implement the above capabilities, the Device Identity Composition Engine (DICE) is used (DICE is an upcoming standard from the Trusted Computing Group (TCG)). In one example, the enabled memory only allows trusted hardware to gain access to the Microsoft Azure IoT cloud. In one example, the health and identity of the IoT device is verified in the memory, where critical code is typically stored. The unique identity of each IoT device can now provide end-to-end device integrity at a new level, starting at the boot process. This can enable additional functionality, such as hardware-based device authentication and provisioning, and managing the remediation of the device when necessary.

[0056] In one embodiment, the method includes: receiving, by a computing device (e.g., a serial NOR flash memory device) a message from a host device (e.g., a CPU, a GPU, an FPGA, or an application controller of a vehicle); generating, by the computing device, an identifier (e.g., a public identifier ID L1public), certificates (e.g., ID L1 certificate) and the key (for example, K L1 public), wherein the identifier is associated with the identity of the computing device and the certificate is generated using the message; and sending the identifier, certificate, and key by the computing device to a host device, wherein the host device is configured to verify the identity of the computing device using the identifier, certificate, and key.

[0057] In some embodiments, the above computing device (e.g., a flash memory device) integrates DICE-RIoT functionality for generating identifiers, certificates, and keys described above and used by a host device to verify the identity of the computing device. In one example, the computing device stores a device secret that acts as a primitive key, and the sequence of identification steps between the layers of the DICE-RIoT protocol is based on the primitive key. In one example, layers L0 and L1 of the DICE-RIoT functionality are implemented in the computing device using hardware and / or software. In one example, layer L0 is implemented only in hardware.

[0058] Figure 1 A host device 151 is shown verifying the identity of a computing device 141 according to one embodiment. The host device 151 sends a message to the computing device 141. In one embodiment, the host device 151 includes a freshness mechanism (not shown) that generates a freshness for use in sending messages to the computing device 141 to avoid replay attacks. In one example, each message sent to the computing device 141 includes a freshness generated by a monotonic counter.

[0059] In one example, the message is an empty string, a conventional known string (e.g., an alphanumeric string known to the manufacturer or operator of host device 151), or may be another value (e.g., an identity value assigned to a computing device). In one example, the message is a unique identity (UID) of the device.

[0060] In response to receiving the message, computing device 141 generates an identifier, certificate, and key. The identifier is associated with the identity of computing device 141. Computing device 141 includes one or more processors 143 that control the operation of identity component 147 and / or other functions of computing device 141.

[0061] The identifiers, certificates, and keys are generated by the identity component 147 and are based on the device secret 149. In one example, the device secret 149 is a unique device secret (UDS) stored in a memory of the computing device 141. In one example, the identity component 147 uses the UDS as a primitive key for implementing the DICE-RIoT protocol. The identifiers, certificates, and keys are output from layer L1 of the DICE-RIoT protocol (see, e.g., Figure 6). In one embodiment, the identity of layer L1 corresponds to the identity of the computing device 141 itself, the manufacturer of the computing device 141, the manufacturer of a thing that includes the computing device 141 as a component, and / or an application or other software stored in the memory of the computing device 141. In one example, the application identity (e.g., ID number) is used for a mobile phone, TV, STB, etc., for which a unique combination of characters and numbers is used to identify the thing.

[0062] In one example, the identity of layer L1 is an ASCII string. For example, the identity can be a manufacturer name concatenated with a thing name (e.g., LG|TV_model_123_year_2018, etc.). In one example, the identity can be represented in hexadecimal form (e.g., 53 61 6D 73 75 6E 67 20 7C 20 54 56 5F 6D 6F 64 65 6C 5F 31 32 335F 79 65 61 72 5F 32 30 31 38).

[0063] In one embodiment, a manufacturer may use a UDS for a class or group of items being produced. In other embodiments, each item may have its own unique UDS. For example, the UDS for a TV may be UDS=0x12234...4444, and the UDS for a laptop may be UDS=0xaabb...00322.

[0064] In one embodiment, device secret 149 is a secret key stored by computing device 141 in memory 145. Identity component 147 uses the secret key as an input to a message authentication code (MAC) to generate a derived secret. In one example, the derived secret is a fused derived secret (FDS) in the DICE-RIoT protocol.

[0065] In one example, memory 145 includes a read-only memory (ROM) storing an initial boot code for booting computing device 141. FDS is a key provided to the initial boot code by processor 143 during a boot operation. In one example, ROM corresponds to layer L0 of the DICE-RIoT protocol.

[0066] Host device 151 uses the identifier, certificate, and key as input to verification component 153, which verifies the identity of computing device 141. In one embodiment, verification component 153 uses the identifier to perform at least one decryption operation to provide a result. The result is compared to the key to determine whether the identity of computing device 141 is valid. If so, host device 151 performs further communications with computing device 141 using the key received from computing device 141. For example, once host device 151 verifies the "triple" (identifier, certificate, and key), the key can be used to authenticate any other information exchanged between computing device 141 and host device 151.

[0067] In one embodiment, digital identities are assigned to a large number of "things" (e.g., according to the Internet of Things). In one example, the things are physical objects, such as vehicles or physical items that reside inside vehicles. In one example, the things are people or animals. For example, each person or animal can be assigned a unique digital identifier.

[0068] In some cases, the manufacturer of a product wants to be able to verify that each product is genuine. Currently, this problem is solved by only buying things from trusted sellers, or from others who have some kind of legal certificate that ensures that the purchased things are genuine. However, in the case of stolen things, if the things do not have electronic identities, it is difficult to block or locate the things so that they are not used inappropriately. In one example, the location is based on the identity of the things when they try to interact with the public infrastructure. In one example, the blocking is based on the inability to verify the identity of the things that want to use the public infrastructure.

[0069] In one embodiment, computing device 141 implements the DICE-RIoT protocol using identity component 147 in order to associate a unique signature with a chain of trust corresponding to computing device 141. Computing device 141 establishes layers L0 and L1. The chain of trust continues through host device 151 which establishes layers L2, ... In one example, a unique identifier may be assigned to each object, person, and animal in any defined environment (e.g., a trust zone defined by geographic parameters).

[0070] In one embodiment, computing device 141 is a component in a thing to which an identity is desired to be assigned. For example, the thing may be an autonomous vehicle that includes computing device 141. For example, computing device 141 may be a flash memory used by an application controller of the vehicle.

[0071] When computing device 141 is manufactured, the manufacturer may inject the UDS into memory 145. In one example, the UDS may be agreed upon and shared with customers who will perform additional manufacturing operations using computing device 141. In another example, the UDS may be randomly generated by the original manufacturer and then communicated to the customer using a secure infrastructure (e.g., via a network such as the Internet).

[0072] In one example, the customer may be the manufacturer of a vehicle incorporating computing device 141. In many cases, the vehicle manufacturer wishes to change the UDS so that it is unknown to the vendor of computing device 141. In these cases, the customer may replace the UDS using an authenticated replacement command provided by host device 151 to computing device 141.

[0073] In some embodiments, the customer may inject customer immutable information into the memory 145 of the computing device 141. In one example, the immutable information is used to generate a unique FDS and is not used alone as a differentiator. The customer immutable information is used to distinguish between various items manufactured by the customer. For example, the customer immutable information may be a combination of letters and / or numbers to define primitive information (e.g., a combination of some or all of the following information: date, time, batch location, wafer location, x, y position in a wafer, etc.).

[0074] For example, in many cases, the immutable information also includes data from a cryptographic feature configuration performed by a user (e.g., a customer who receives a device from a manufacturer). This configuration or setup can be accomplished simply by using an authentication command (a command that requires knowledge of the key to be executed). The user has knowledge of the key (e.g., based on provisioning via a secure infrastructure from the manufacturer). The immutable information represents a form of cryptographic identity for the computing device that is distinct from the device's unique ID (UID). In one example, including the cryptographic configuration in a set of immutable information provides users with tools that can be used to customize the immutable information themselves.

[0075] In one embodiment, computing device 141 includes a freshness mechanism that generates freshness. Freshness can be provided with an identifier, certificate, and key when sent to host device 151. Freshness can also be used with other communications of host device 151.

[0076] In one embodiment, computing device 141 is a component on an application board. Another component on the application board (not shown) can use knowledge of device secret 149 (e.g., knowledge of the injected UDS) to verify the identity of computing device 141. The component requests computing device 141 to produce an output using a message authentication code in order to prove possession of the UDS. For example, the message authentication code may be as follows: HMAC(UDS, "application board message|freshness")

[0077] In another embodiment, FDS can also be used as a criterion to prove possession of a device (e.g., knowledge of a secret key). FDS is derived from UDS in this way: FDS = HMAC-SHA256[UDS, SHA256(identity of "L1")]

[0078] Therefore, the message authentication code can be as follows: HMAC(FDS, "AppPadMessage|Freshness")

[0079] Figure 2 An example computing system with an identity component 107 and an authentication component 109 is shown according to one embodiment. A host system 101 communicates with a memory system 105 via a bus 103. A processing device 111 of the memory system 105 has read / write access to memory areas 111, 113, ... 119 of a non-volatile memory 121. In one example, the host system 101 also reads data from and writes data to the volatile memory 123. In one example, the identity component 107 supports layers L0 and L1 of the DICE-RIoT protocol. In one example, the non-volatile memory 121 stores boot code.

[0080] Verification component 109 is used to verify the identity of memory system 105. Verification component 109 uses a triptych including an identifier certificate and a key generated by identity component 107 in response to receiving a host message from host system 101, for example as described above.

[0081] The identity component 107 is Figure 1 The authentication component 109 is an instance of the identity component 147. Figure 1 An instance of the verification component 153.

[0082] The memory system 105 includes a key store 157 and a key generator 159. In one example, the key store 157 can store root keys, session keys, UDS (DICE-RIoT), and / or other keys used for cryptographic operations by the memory system 105.

[0083] In one example, key generator 159 generates a public key that is sent to host system 101 for verification by verification component 109. The public key is sent as part of a triplet that also includes an identifier and a certificate, as described above.

[0084] The memory system 105 includes a freshness generator 155. In one example, the freshness generator 155 can be used for authenticated commands. In one example, multiple freshness generators 155 can be used. In one example, the freshness generator 155 can be used by the host system 101.

[0085] In one example, processing device 111 and memory regions 111, 113, ... 119 are on the same chip or die. In some embodiments, memory regions store data for use by host system 101 and / or processing device 111 during machine learning processing and / or other data generated by software processing programs executed on host system 101 or on processing device 111.

[0086] The computing system may include a write component in the memory system 105 that selects a memory region 111 (e.g., a recording segment of flash memory) for recording new data from the host system 101. The computing system 100 may further include a write component in the host system 101 that coordinates with a write component 107 in the memory system 105 to at least facilitate the selection of the memory region 111.

[0087] In one example, volatile memory 123 is used as system memory for a processing device (not shown) of host system 101. In one embodiment, a processing program of host system 101 selects a memory region for writing data. In one example, host system 101 may select a memory region based in part on data from sensors and / or software processing programs executing on the autonomous vehicle. In one example, the aforementioned data is provided by host system 101 to processing device 111 which selects a memory region.

[0088] In some embodiments, the host system 101 or the processing device 111 includes at least a portion of the identity component 107 and / or the verification component 109. In other embodiments, or in combination, the processing device 111 and / or the processing device in the host system 101 includes at least a portion of the identity component 107 and / or the verification component 109. For example, the processing device 111 and / or the processing device of the host system 101 may include logic circuits that implement the identity component 107 and / or the verification component 109. For example, a controller or processing device (e.g., a CPU, FPGA, or GPU) of the host system 101 may be configured to execute instructions stored in a memory for performing the operations of the identity component 107 and / or the verification component 109 described herein.

[0089] In some embodiments, identity component 107 is implemented in an integrated circuit chip disposed in memory system 105. In other embodiments, authentication component 109 in host system 101 is part of an operating system, device driver, or application of host system 101.

[0090] An example of a memory system 105 is a memory module connected to a central processing unit (CPU) via a memory bus. Examples of memory modules include dual in-line memory modules (DIMMs), small outline DIMMs (SO-DIMMs), non-volatile dual in-line memory modules (NVDIMMs), etc. In some embodiments, the memory system may be a hybrid memory / storage system that provides both memory functions and storage device functions. In general, a host system may utilize a memory system that includes one or more memory areas. The host system may provide data to be stored at the memory system and may request data to be retrieved from the memory system. In one example, the host may access various types of memory including volatile and non-volatile memory.

[0091] The host system 101 may be a computing device, such as a controller in a vehicle, a network server, a mobile device, a cellular phone, an embedded system (e.g., an embedded system with a system on chip (SOC) and internal or external memory), or any computing device that includes a memory and a processing device. The host system 101 may include or be coupled to a memory system 105, so that the host system 101 can read data from the memory system 105 or write data to the memory system 105. The host system 101 may be coupled to the memory system 105 via a physical host interface. As used herein, "coupled to" generally refers to a connection between components, which may be an indirect communication connection or a direct communication connection (e.g., without intermediate components), whether wired or wireless, including connections such as electrical connections, optical connections, magnetic connections, etc. Examples of physical host interfaces include, but are not limited to, a Serial Advanced Technology Attachment (SATA) interface, a Peripheral Component Interconnect Express (PCIe) interface, a Universal Serial Bus (USB) interface, Fibre Channel, Serial Attached SCSI (SAS), a Double Data Rate (DDR) memory bus, etc. The physical host interface may be used to transmit data between the host system 101 and the memory system 105. The physical host interface may provide an interface for passing control, address, data, and other signals between the memory system 105 and the host system 101.

[0092] Figure 2 Memory system 105 is described as an example. In general, host system 101 can access multiple memory systems via the same communication connection, multiple separate communication connections, and / or a combination of communication connections.

[0093] The host system 101 may include a processing device and a controller. The processing device of the host system 101 may be, for example, a microprocessor, a central processing unit (CPU), a processing core of a processor, an execution unit, etc. In some cases, the controller of the host system may be referred to as a memory controller, a memory management unit, and / or an initiator. In one example, the controller controls communications between the host system 101 and the memory system 105 via the bus 103. These communications include sending host messages for verifying the identity of the memory system 105 as described above.

[0094] The controller of the host system 101 may communicate with the controller of the memory system 105 to perform operations such as reading data, writing data, or erasing data at a memory area of ​​the non-volatile memory 121. In some cases, the controller is integrated into the same package of the processing device 111. In other cases, the controller is separate from the package of the processing device 111. The controller and / or the processing device may include hardware, such as one or more integrated circuits and / or discrete components, buffer memory, cache memory, or a combination thereof. The controller and / or the processing device may be a microcontroller, a dedicated logic circuit (such as a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), etc.), or another suitable processor.

[0095] In one embodiment, memory regions 111, 113, ... 119 may include any combination of different types of non-volatile memory components. In addition, memory cells of a memory region may be grouped into memory pages or data blocks, which may refer to units for storing data. In some embodiments, volatile memory 123 may be, but is not limited to, random access memory (RAM), dynamic random access memory (DRAM), and synchronous dynamic random access memory (SDRAM).

[0096] In one embodiment, one or more controllers of the memory system 105 may communicate with the memory areas 111, 113, ... 119 to perform operations, such as reading data, writing data, or erasing data. Each controller may include hardware, such as one or more integrated circuits and / or discrete components, buffer memory, or a combination thereof. Each controller may be a microcontroller, a dedicated logic circuit (such as a field programmable gate array (FPGA), an application-specific integrated circuit (ASIC), etc.), or another suitable processor. The controller may include a processing device (processor) configured to execute instructions stored in a local memory. In one example, the local memory of the controller includes an embedded memory configured to store instructions for executing various processing programs, operations, logic flows, and routines for controlling the operation of the memory system 105 (including controlling the communication between the memory system 105 and the host system 101). In some embodiments, the local memory may include memory registers that store memory pointers, extracted data, etc. The local memory may also include a read-only memory (ROM) for storing microcode.

[0097] In general, the controller of the memory system 105 may receive commands or operations from the host system 101 and / or the processing device 111 and may convert the commands or operations into instructions or appropriate commands that implement the selection of the memory area based on the data write counter for the memory area. The controller may also be responsible for other operations, such as wear leveling, garbage collection operations, error detection and error correction code (ECC) operations, encryption operations, cache operations, and address conversion between logical block addresses and physical block addresses associated with the memory area. The controller may further include a host interface circuit to communicate with the host system 101 via a physical host interface. The host interface circuit may convert commands received from the host system into command instructions to access one or more of the memory areas, and convert responses associated with the memory areas into information for the host system 101.

[0098] The memory system 105 may also include additional circuits or components not illustrated. In some embodiments, the memory system 105 may include a cache or buffer (e.g., DRAM or SRAM) and address circuits (e.g., row decoders and column decoders) that can receive addresses from one or more controllers and decode the addresses to access memory regions.

[0099] In some embodiments, a controller and / or processing device 111 in the host system 101 or the memory system 105 includes at least a portion of the identity component 107 and / or the verification component 109. For example, the controller and / or processing device 111 may include logic circuits that implement the identity component 107 and / or the verification component 109. For example, the processing device (processor) may be configured to execute instructions stored in a memory for performing operations that provide read / write access to a memory area of ​​the identity component 107 as described herein. In some embodiments, the verification component 109 is part of an operating system, a device driver, or an application.

[0100] Figure 3 An example computing device of a vehicle 100 is shown according to one embodiment. For example, the vehicle 100 may be an autonomous vehicle, a non-autonomous vehicle, an emergency vehicle, a service vehicle, etc.

[0101] The vehicle 100 includes a vehicle computing device 110, such as an onboard computer. The vehicle computing device 110 is Figure 1 In another example, the vehicle computing device 110 is Figure 2 1 is an instance of host system 101, and memory 160 is an instance of memory system 105.

[0102] The vehicle computing device 110 includes a processor 120 coupled to a vehicle communication component 130, such as a reader, writer, and / or other computing device capable of performing the functions described below, which is coupled to (or includes) an antenna 140. The vehicle communication component 130 includes a processor 150 coupled to a memory 160, such as non-volatile flash memory, but embodiments are not limited to this kind of memory device.

[0103] In one instance, memory 160 is adapted to store all information related to the vehicle (e.g., driver, passengers, and cargo carried) in a manner that enables vehicle 100 to provide this information when approaching a checkpoint by using a communication interface (e.g., the so-called DICE-RIoT protocol), as described below.

[0104] In one example, vehicle information (e.g., vehicle ID / license plate number) is already stored in the vehicle memory 160, and the vehicle 100 is able to identify the electronic ID of the passenger and / or the ID of the luggage, cargo, etc. being transported, for example, through the communication component 130 and by using the known DICE-RIoT protocol or similar protocol, and then store this information in the memory 160. In one example, the electronic ID, the luggage, and cargo containers being transported are equipped with wireless transponders, NFC, Bluetooth, RFID, contactless sensors, magnetic strips, etc., and the communication component 130 can use readers and / or electromagnetic fields to obtain the required information from such remote sources.

[0105] In one example, all passenger IDs and / or IDs of carried luggage, cargo, etc. are equipped with electronic devices capable of exchanging data with the communication component. Those electronic devices may be active or passive elements, in the sense that they may be active in that they are powered by electricity, or may be activated and powered by an external power source that provides the required power supply only when the electrical device is in its vicinity.

[0106] The rental vehicle or autonomous vehicle may use readers and / or electromagnetic fields to obtain information inside or near the vehicle, or alternatively, may even receive information from a remote source, such as when the driver of the rental vehicle is already aware of the rental system due to a previous reservation. When the driver arrives to pick up the vehicle, another check may be performed in real time.

[0107] Similarly, all information about the transported luggage and cargo (and about passengers) carried by the vehicle 100 can be maintained to be always up to date. To this end, the electronic ID of the passenger and / or the ID of the carried luggage and cargo are updated in real time due to wireless transponders associated with the luggage and cargo or possessed by the passenger (not shown).

[0108] In one example, communication between the vehicle communication component 130 and nearby sources (eg, cargo transponders, etc.) occurs via the DICE-RIoT protocol.

[0109] In one example, the vehicle computing device 110 can control operating parameters of the vehicle 100, such as steering and speed. For example, a controller (not shown) can be coupled to a steering control system 170 and a speed control system 180. In addition, the vehicle computing device 110 can be coupled to an information system 190. The information system 190 can be configured to display messages, such as route information or checkpoint safety messages, and can display visual warnings and / or output audible warnings. The communication component 130 can receive information from additional computing devices, such as from an external computing device (not shown).

[0110] Figure 4An example system 390 is shown having a host device 350 communicating with an example computing device of a vehicle 300 according to one embodiment. The computing device includes a passive communication component 310, such as a short-range communication device (e.g., an NFC tag). The communication component 310 may be in the vehicle 300, which may be as shown in FIG. Figure 3 1 is configured as shown for vehicle 100, and includes components of vehicle 100 in addition to a communication component 310 that may be configured as vehicle communication component 130. Communication component 310 includes a chip 320 (e.g., a CPU or application controller implementing vehicle 300) having a non-volatile storage component 330 that stores information about vehicle 300 (e.g., vehicle ID, driver / passenger information, cargo information carried, etc.). Communication component 310 may include an antenna 340.

[0111] Host device 350 is, for example, an active communication device (e.g., it includes a power source) that can receive information from and / or transmit information to communication component 310. In some examples, host device 350 can include a reader (e.g., an NFC reader), such as a toll reader, or other components. Host device 350 can be an external device that is arranged (e.g., embedded) near a checkpoint (e.g., at the boundary of a trust zone) or generally near a restricted access area. In some embodiments, host device 350 can also be carried by police for use as a portable device.

[0112] Host device 350 may include a processor 360, a memory 370 such as a non-volatile memory, and an antenna 380. Memory 370 may include an NFC protocol that allows host device 350 to communicate with communication component 310. For example, host device 350 and communication component 310 may communicate using an NFC protocol, such as at approximately 13.56 megahertz and in accordance with the ISO / IEC 18000-3 international standard. Other methods using RFID tags may be used.

[0113] Host device 350 may also communicate with a server or other computing device (e.g., via a wireless network to a central operations center). For example, host device 350 may be wirelessly coupled or hardwired to a server or communications center. In some instances, host device 350 may communicate with the operations center via WIFI or via the Internet. When vehicle 300 brings antenna 340 within communication range of antenna 380, host device 350 may power communication component 310. In some instances, host device 350 may receive real-time information from the operations center and may transmit the information to vehicle 300. In some embodiments, communications component 310 may have its own battery.

[0114] In one embodiment, host device 350 is adapted to read / send information from / to vehicle 300 that is equipped with a communication component 310 (eg, an active device) configured to allow for the exchange of information.

[0115] Reference again Figure 3 , the vehicle communication component 130 of the vehicle 100 can be activated internally to pick up real-time relevant information about passenger IDs, luggage and / or cargo being transported (e.g., when equipped with the above-mentioned Figure 4 The vehicle's computing device can detect information within a spatial range of several meters (e.g., 2 to 3 meters) so that all data corresponding to passengers, luggage, and cargo can be acquired. In one example, this occurs when the vehicle approaches an external communication component (e.g., a server or other computing device acting as a host device) within a certain proximity so that communication can begin and / or become enhanced. The communication distance is, for example, 2 to 3 meters.

[0116] In one embodiment, the vehicle communication component 130 may encrypt data when communicating with external entities and / or with internal entities. In some cases, data about the luggage, cargo, or even passengers being transported may be confidential or contain confidential information (e.g., the normal status of the passengers or confidential documents or hazardous materials). In this case, it is necessary to maintain the information and data stored in the memory portion associated with the vehicle computing device as encrypted data.

[0117] In various embodiments discussed below, methods for encrypting and decrypting communications between an internal vehicle computing device and an external entity (e.g., a server acting as a host device) are discussed. In one example, this method can even be applied between the internal vehicle computing device and electronic components associated with passengers, luggage, and cargo carried on board the vehicle.

[0118] In one example, the vehicle communication component 130 sends the vehicle public key to the external communication component (e.g., acting as a host device 151), and the external communication component sends the external public key to the vehicle communication component 130. These public keys (vehicle and external) can be used to encrypt data sent to each corresponding communication component, and verify the identity of each communication component and exchange confirmation and other information. As an example, as further described below, the vehicle communication component 130 can encrypt data using the received external public key and send the encrypted data to the external communication component. Similarly, the external communication component can encrypt data using the received vehicle public key and send the encrypted data to the vehicle communication component 130. The data sent by the vehicle 100 may include car information, passenger information, cargo information, etc. The information can be optionally sent with a digital signature to verify the identity of the vehicle 100. In addition, the information can be provided to the vehicle 100 and displayed on the dashboard of the vehicle 100, or sent to an email to a computing device associated with the vehicle 100 (e.g., a user device or a central server monitoring the vehicle). The vehicle can be identified based on the vehicle's identification, VIN number, etc. together with the vehicle's digital signature.

[0119] In one example, data exchanged between a vehicle and an external entity may have a freshness level for use by the other party. As an example, data sent by a vehicle to an external entity to indicate the same instruction may change every certain time frame or for a certain amount of data sent. This prevents a hacker from intercepting confidential information contained in previously sent data and sending the same data again to produce the same result. If the data is slightly modified but still indicates the same instruction, a hacker may be able to send the same information at a later point in time, but the same instruction will not be executed because the recipient expects the modified data to execute the same instruction.

[0120] Data exchanged between the vehicle 100 and an external entity (eg, a computing system or device) (not shown) may be performed using a number of encryption and / or decryption methods as described below. The security of the data may ensure that unauthorized activities are prevented from interfering with the operation of the vehicle 100 and the external entity.

[0121] Figure 5A An application board is shown generating a triplet including an identifier, a certificate, and a key that is sent to a host device according to one embodiment. The host device uses the triplet to verify the identity of the application board. The application board is Figure 1 The host device is an instance of a computing device 141. Figure 1 An example of a host device 151.

[0122] In one embodiment, the application board and the host computer include a communication component that performs encryption and / or decryption operations for communications (e.g., regarding information and data) using a Device Identity Combination Engine (DICE)-Robust Internet of Things (RIoT) protocol. In one example, the DICE-RIoT protocol is applied to communications between the vehicle communication component and external communication components, and to communications within the vehicle environment between the vehicle communication component and various wireless electronic devices associated with each of the passenger ID, luggage, cargo, etc.

[0123] Figure 5B An example computing system is shown that is segmented and started using layers according to one embodiment. The system includes an external communication component 430' and a vehicle communication component 430". When the vehicle is close to or in the vicinity of an external entity, the vehicle's associated vehicle communication component 430" can exchange data with the external entity as described above, for example, using a sensor (e.g., a radio frequency identification sensor or RFID, etc.).

[0124] In other embodiments, component 430′ may be an application board located in a vehicle, and component 430″ may be a host device also located in the vehicle that uses the DICE-RIoT protocol to verify the identity of component 430′ (e.g., as described above with respect to Figure 1 discussed).

[0125] In one embodiment, the DICE-RIoT protocol is used by a computing device to boot in stages using layers, where each layer authenticates and loads subsequent layers and provides increasingly complex runtime services at each layer. Thus, one layer can be served by the previous layer and serve the next layer, creating an interconnected network of layers built on lower layers and serving higher layers. Alternatively, other protocols can be used instead of the DICE-RIoT protocol.

[0126] In one example implementation of the communication protocol, the security of the communication protocol is based on a secret value, which is a device secret (e.g., UDS) set during manufacturing (or even later). The device secret UDS exists within the device to which it is provisioned (e.g., as a Figure 1 The device secret 149 is stored).

[0127] The device secrets UDS is accessible at boot time by the first stage ROM based boot loader. The system then provides a mechanism to make the device secrets inaccessible until the next boot cycle, and only the boot loader (e.g., boot layer) can access the device secrets UDS. Thus, in this approach, booting starts with a specific architecture layering of the device secrets UDS.

[0128] like Figure 5BAs illustrated, layer 0L0 and layer 1L1 are within an external communication component 430'. Layer 0L0 may provide a fused derived secret FDS key to layer 1L1. The FDS key may be based on the identity of the code of layer 1L1 and other security-related data. A specific protocol (e.g., a robust Internet of Things (RIoT) core protocol) may use the FDS to verify the code of layer 1L1 that it loads. In an example, the specific protocol may include a device identification combination engine (DICE) and / or a RIoT core protocol. As an example, the FDS may include the layer 1L1 firmware image itself, a list that cryptographically identifies authorized layer 1L1 firmware, a firmware version number of the firmware signed in the context of a secure boot implementation, and / or security-critical configuration settings for the device. The device secret UDS may be used to create the FDS and stored in the memory of the external communication component. Therefore, layer 0L0 never displays the actual device secret UDS, but instead provides a derived key (e.g., an FDS key) to the next layer in the boot chain.

[0129] As indicated by arrow 410', external communication component 430' is adapted to transmit data to vehicle communication component 430". The transmitted data may include a public external identification, certificate (eg, external identification certificate) and / or external public key, such as will be combined with Figure 6 Description. The layer 2L2 of the vehicle communication component 430 ″ may receive the transmitted data and execute the data in the operation of the operating system OS, for example, on the first application App1 and the second application App2 .

[0130] Likewise, as indicated by arrow 410", the vehicle communication component 430" may transmit data including a public vehicle identification, certificate (e.g., a vehicle identification certificate) and / or a vehicle public key. As an example, after verification (e.g., after verifying the certificate), the vehicle communication component 430" may send a vehicle identification number VIN for further authentication, identification and / or verification of the vehicle.

[0131] like Figure 5B and Figure 6 As shown in FIG. 4 , in an example operation, the external communication component 430 ′ may read the device secret DS, hash the identity of layer 1 L1, and perform the following calculations:

[0132] FDS = KDF[UDS, Hash("Immutable Information")]

[0133] Where KDF is a cryptographic one-way key derivation function (e.g., HMAC-SHA256). In the above calculation, the hash can be any cryptographic primitive, such as SHA256, MD5, SHA3, etc.

[0134] In at least one example, the vehicle may communicate using either an anonymous login or an authenticated login. Authenticated login may allow the vehicle to obtain additional information that may not be accessible when communicating in anonymous mode. In at least one example, authentication may include providing a vehicle identification number (VIN) and / or authentication information, such as an exchange of a public key, as will be described below. In either of the anonymous and authenticated modes, an external entity (e.g., a checkpoint police officer at the border of the trust zone) may communicate with the vehicle to provide an external public key associated with the external entity to the vehicle.

[0135] Figure 6 An example computing device is shown that generates identifiers, certificates, and keys using an asymmetric generator according to one embodiment.In one embodiment, the computing device implements a process for determining parameters (eg, within layer L1 of an external device, or in an alternative embodiment, within layer L1 of an internal computing device).

[0136] In one embodiment, the determination includes the information that is then sent (as indicated by arrow 510') to a vehicle communication component (eg, Figure 5B Parameters of the external public identification, external certificate and external public key of layer 2L2 (reference number 430" in FIG. 4 ). Figure 6 The arrows 510' and 510" correspond to Figure 5B Arrows 410' and 410". And, Figure 6 The layers in correspond to Figure 5B layer.

[0137] In another embodiment, a message from a host device ("host message") is merged with an external public key by mode (data) merge 531 to provide merged data for encryption. The merged data is input to encryptor 530. In one example, the host message is concatenated with the external public key. The resulting parameters include a triplet that is sent to the host device and used to verify the identity of the computing device. For example, the external public identification, the external certificate, and the external public key are used by the authentication component of the host device to verify the identity. In one example, the host device is Figure 1 host device 151.

[0138] like Figure 6As shown, the FDS from layer 0L0 is sent to layer 1L1 and is used by asymmetric ID generator 520 to generate a public identity IDlkpublic and a private identity IDlkprivate. In the abbreviation "IDlkpublic", "lk" indicates a common layer k (in this example, layer 1L1), and "public" indicates a publicly shared identity. The public identity IDlkpublic is illustrated as shared by an arrow extending to the right and extending outside of layer 1L1 of an external communication component. The generated private identity IDlkprivate is used as a key input into an encryptor 530. The encryptor 530 can be, for example, any processor, computing device, etc. for encrypting data.

[0139] Layer 1L1 of the external communication component may include an asymmetric key generator 540. In at least one example, a random number generator RND may optionally input a random number into the asymmetric key generator 540. The asymmetric key generator 540 may generate a random number that is consistent with the external communication component (e.g., Figure 5B The public key KLkpublic (referred to as the external public key) and the private key KLkprivate (referred to as the external private key) associated with the external communication component 430' in FIG.

[0140] The external public key KLkpublic may be an input (as "data") into the encryptor 530. As mentioned above, in some embodiments, a host message previously received from the host device as part of the authentication process is combined with KLkpublic to provide the combined data as input data to the encryptor 530.

[0141] The encryptor 530 may use the input of the external private identification IDlkprivate and the external public key KLkpublic to produce a result K'. The external private key KLkprivate and the result K' may be input into the additional encryptor 550, thereby producing an output K". The output K" is an external certificate IDL1certificate that is transmitted to layer 2L2 (or alternatively, to a host device that verifies the identity). The external certificate IDL1certificate may be able to verify and / or authenticate the source of data sent from the device. As an example, data sent from an external communication component may be associated with the identity of the external communication component by verifying the certificate, such as by combining Figure 7 Further description. In addition, the external public key KL1public key can be transmitted to the layer 2L2. Therefore, the public identification ID11public of the external communication component, the certificate IDL1certificate and the external public key KL1public key can be transmitted to the layer 2L2 of the vehicle communication component.

[0142] Figure 7A verification component is shown that verifies the identity of a computing device using a decryption operation according to one embodiment. The verification component includes decryptors 730, 750. The verification component implements a process of verifying a certificate according to an embodiment of the present disclosure.

[0143] exist Figure 7 In the illustrated example, from an external communication component (e.g., from Figure 5B The layer 1L1) of the external communication component 430' in provides a public key KL1public, a certificate IDL1certificate and a public identification IDL1public.

[0144] The data of the certificate IDL1certificate and the external public key KL1public can be used as input into the decryptor 730. The decryptor 730 can be any processor, computing device, etc. used to decrypt data. The result of the decryption of the certificate IDL1certificate and the external public key KL1public can be used as input into the decryptor 750 together with the public identification IDL1public, thereby producing an output. As illustrated at box 760, the external public key KL1public and the output from the decryptor 750 can indicate whether the certificate is verified, thereby producing a yes or no as an output. A private key is associated with a single layer, and a specific certificate can only be produced by a specific layer.

[0145] In response to the certificate being verified (e.g., after authentication), data received from the verified device can be accepted, decrypted, and / or processed. In response to the certificate not being verified, data received from the verified device can be discarded, removed, and / or ignored. In this way, unauthorized devices sending illicit data can be detected and avoided. As an example, a hacker sending data to be processed can be identified and the hacker data not processed.

[0146] In an alternative embodiment, from Figure 1 The computing device 141 or from Figure 2 The memory system 105 of the present invention provides a public key KL1public, a certificate IDL1certificate, and a public identification IDL1public. This triple is generated by the computing device 141 in response to receiving a host message from the host device. Before providing IDL1certificate as input to the decryptor 730, IDL1certificate and the message from the host device ("host message") are merged by mode (data) merge 731. In one example, the merge is a concatenation of the data. The merged data is provided as input to the decryptor 730. The verification process then proceeds in other ways as described above.

[0147] Figure 8A block diagram showing an example process to verify a certificate according to one embodiment. In the case where a device sends data that can be verified to avoid subsequent repudiation, a signature can be generated and sent with the data. For example, a first device may make a request to a second device, and once the second device performs the request, the first device may indicate that the first device never made such a request. Anti-repudiation methods, such as using a signature, can avoid repudiation by the first device and ensure that the second device can perform the requested task without subsequent difficulty.

[0148] Vehicle computing device 810" (e.g., Figure 3 The vehicle computing device 110 or Figure 1 The vehicle computing device 141) may send data Dat" to an external computing device 810' (or generally to any other computing device). The vehicle computing device 810" may generate a signature Sk using the vehicle private key KLkprivate. The signature Sk may be transmitted to the external computing device 810'. The external computing device 810' may verify using the data Dat' and a previously received public key KLkpublic (e.g., a vehicle public key). In this way, signature verification operates by encrypting the signature using a private key and decrypting the signature using a public key. In this way, a unique signature for each device may be kept private to the device sending the signature while allowing the receiving device to be able to decrypt the signature for verification. This is in contrast to encryption / decryption of data, which is encrypted by the sending device using the public key of the receiving device and decrypted by the receiving device using the private key of the receiver. In at least one instance, the vehicle may verify the digital signature using an internal cryptographic process, such as an elliptic curve digital signature (ECDSA) or a similar process.

[0149] Due to the exchange and verification of certificates and public keys, the devices are able to communicate with each other in a secure manner. When the vehicle approaches an external entity (such as a trust zone boundary, a border security entity or a host device that is generally electronically controlled), the corresponding communication device (which has Figure 7The DICE-RIoT protocol described above can be used to encrypt / decrypt the exchanged messages / information. In some embodiments, so-called immutable information (e.g., license plate number / ID, VIN, insurance number) is not encrypted, while other information is encrypted. In other words, in the exchanged messages, there may be unencrypted data as well as encrypted data: therefore, the information may be encrypted or unencrypted or mixed. The correctness of the message is then ensured by verifying that the content of the message is valid using the certificate / public key.

[0150] Fig. 9 A method of verifying the identity of a computing device using an identifier, a certificate, and a key according to one embodiment is shown. For example, Fig. 9 The method can be found in Figures 1 to 7 implemented in the system.

[0151] Fig. 9 The method may be performed by processing logic, which may include hardware (e.g., a processing device, a circuit, dedicated logic, programmable logic, microcode, hardware of a device, an integrated circuit, etc.), software (e.g., instructions running or executed on a processing device), or a combination thereof. In some embodiments, Fig. 9 The method is at least partly based on Figure 1 The identity component 147 and the verification component 153 are executed.

[0152] Although shown in a particular order or sequence, the order of the processes may be modified unless otherwise specified. Therefore, it should be understood that the illustrated embodiments are merely examples, and the illustrated processes may be performed in a different order, and some processes may be performed in parallel. In addition, one or more processes may be omitted in various embodiments. Therefore, not all processes are required in each embodiment. Other process flows are possible.

[0153] At block 921, a message is received from a host device. For example, computing device 141 receives a message from host device 151 (eg, "Host Message" or "Host Message|Freshness").

[0154] At block 923, an identifier, certificate, and key (eg, a public key K L1The ID is associated with the identity of the computing device. The certificate is generated using a message from the host device (e.g., a "host message"). In one embodiment, the message is combined with the public key before encryption. This encryption uses the private identifier ID L1 private as the key. Private identifier ID L1 Private and public identifiers L1 public association (eg, an associated pair generated by asymmetric ID generator 520).

[0155] In one example, the identity component 147 generates an identifier, a certificate, and a key to provide a triple. In one example, the triple is generated based on the DICE-RIoT protocol. In one example, Figure 6 The triple is produced as described in .

[0156] In one example, using the DICE-RIoT protocol, each layer (L k ) provides the key and certificate set to the next layer (L k+1 ), and each certificate can be verified by the receiving layer. The fused derived secret FDS is calculated as follows:

[0157] FDS = HMAC-SHA256[UDS,SHA256("L1's identity")]

[0158] In one example, Layer 1 L1 in the DICE-RIoT architecture generates a certificate using a host message sent by a host device. Layer 1 calculates two associated key pairs as follows:

[0159] (ID lk public ,ID lk private ) and (KL k public ,KL k private )

[0160] Layer 1 also computes two signatures as follows:

[0161] K'=encrypted(ID lk private ,KL k public |Host message)

[0162] K"=encryption(KL k private ,K')

[0163] From the above processing, layer 1 provides triple as follows:

[0164]

[0165] More generally, each layer provides three things as follows:

[0166] K Lk=For each k=1:N, {a set of keys and certificates}

[0167] Using the corresponding triplets, each layer is able to prove its identity to the next layer.

[0168] In one example, layer 2 corresponds to application firmware, and subsequent layers correspond to an operating system and / or applications of the host device.

[0169] At block 925, the generated identifier, certificate, and key are sent to the host device. The host device uses the identifier, certificate, and key to verify the identity of the computing device. In one example, host device 151 receives the identifier, certificate, and key from computing device 141. Host device 151 uses verification component 153 to verify the identity of the computing device.

[0170] In one example, the verification component 153 performs a decryption operation as part of the verification process. Decryption includes combining the message from the host with the certificate before decrypting it using the key received from the computing device 141. In one example, Figure 7 Verification of the identity of the computing device is performed as described in .

[0171] In one example, the decryption operation is performed as follows:

[0172] Using KL 1 public Yes (ID L1 certificate) to provide K'

[0173] Use ID L1 public Decrypt K' to provide the result

[0174] The result is compared with KL 1 public Compare. If the result is equal to KL 1 public , then the identity is verified. In one example, the application board identity is verified.

[0175] In one embodiment, the identity of a person or animal is verified. Verification of the identity of a person is performed similarly to verifying the identity of computing device 141 as described above. In one instance, computing device 141 is integrated into a person's passport. Public administrations of countries that have issued passports may use a UDS specific to a class of documents (e.g., driver's licenses, passports, ID cards, etc.). For example, for Italy, Sicily, Messina, Passport Office, UDS = 0x12234...4444. For Germany, Bavaria, Munich, Passport Office, UDS = 0xaabb...00322

[0176] In an example about a passport, the identity of L1 is the following ASCII string:

[0177] Country | Document Type | Etc. (e.g., "Italy, Sicily, Messina, Passport Office")

[0178] The “granularity” of the allocation can be determined by public administration in each country.

[0179] Fig. 9 Various embodiments of the method provide various advantages. For example, things can be identified and certified as produced by a specific factory without using a third-party key infrastructure (e.g., PKI = Public Key Infrastructure). Malicious or hacker in the middle attacks are prevented due to replay protection. The method can be used for mass production of things.

[0180] In addition, the customer UDS is protected at the hardware level (e.g., inaccessibility of layer 0 outside the component). The UDS cannot be read by anyone, but it can be replaced (e.g., only by the customer through the use of a secure protocol). Examples of secure protocols include secure protocols based on authenticated, replay protected commands and / or based on secret sharing algorithms such as Diffie Hellman (e.g., ECDH Elliptic Curve Diffie Hellman). In one example, the UDS is communicated to the customer (not the end user) using a secure infrastructure. The UDS can be customized by the customer.

[0181] Additionally, component identification can work without the presence of an Internet or other network connection. Also, the method can be used to easily check the identity of things, animals, and people at trust zone boundaries (e.g., national borders, internal checkpoints, etc.).

[0182] In one example, knowledge of the UDS permits the host device to safely replace the UDS. For example, replacement may be done when the host wishes to change the identity of a thing, or the host wishes to keep the identity of the thing unknown to anyone else, including the original manufacturer.

[0183] In another example, a replace command is used by a host device. For example, the host device may send a replace UDS command to the computing device. The replace command includes an existing UDS and a new UDS to be attributed to the computing device. In one example, the replace command has a field including the following hash value: hash(existing UDS|new UDS).

[0184] In another example, an authenticated replay protection command is used with the following fields: Replace_command|Freshness|Signature

[0185] Where signature = MAC[secret_key,Replace_command|freshness|hash(existing UDS|new UDS)]

[0186] The secret key is an additional key and is a key used for authentication commands present on the device. For example, the secret key may be a session key as described below (see, e.g. Fig.12 ).

[0187] In one embodiment, a method includes: receiving, by a computing device (e.g., computing device 141), a message from a host device (e.g., host device 151); generating, by the computing device, an identifier, a certificate, and a key, wherein the identifier is associated with an identity of the computing device and the certificate is generated using the message; and sending, by the computing device, the identifier, the certificate, and the key to the host device, wherein the host device is configured to use the identifier, the certificate, and the key to verify the identity of the computing device.

[0188] In one embodiment, verifying the identity of the computing device includes concatenating the message with the credential to provide the first data.

[0189] In one embodiment, verifying the identity of the computing device further comprises decrypting the first data using the key to provide the second data.

[0190] In one embodiment, verifying the identity of the computing device further comprises decrypting the second data using the identifier to provide a result, and comparing the result to the key.

[0191] In one embodiment, the identifier is a public identifier and the computing device stores a secret key, the method further comprising: using the secret key as an input to a message authentication code to generate a derived secret; wherein the derived secret is used as an input to an asymmetric generator to generate the public identifier.

[0192] In one embodiment, the identifier is a first public identifier, and the computing device stores a first device secret used to generate the first public identifier, the method further comprising: receiving a replacement command from a host device; in response to receiving the replacement command, replacing the first device secret with a second device secret; and sending a second public identifier generated using the second device secret to the host device.

[0193] In one embodiment, the secret key is a public key, and generating the certificate includes concatenating the message with the public key to provide a data input for encryption.

[0194] In one embodiment, the identifier is a public identifier, and a first asymmetric generator generates a public identifier and a private identifier as an associated pair; the key is a public key, and a second asymmetric generator generates a public key and a private key as an associated pair; and generating a certificate includes: concatenating a message with the public key to provide first data; encrypting the first data using the private identifier to provide second data; and encrypting the second data using the private key to provide a certificate.

[0195] In one embodiment, the key is a public key, the method further comprising generating a random number as an input to an asymmetric key generator, wherein the public key and the associated private key are generated using the asymmetric key generator.

[0196] In one embodiment, a physically unclonable function (PUF) is used to generate the random numbers.

[0197] In one embodiment, a system includes: at least one processor; and a memory containing instructions, wherein the instructions are configured to direct the at least one processor to perform the following operations: send a message to a computing device; receive an identifier, a certificate, and a key from the computing device, wherein the identifier is associated with an identity of the computing device and the certificate is generated by the computing device using the message; and verify the identity of the computing device using the identifier, the certificate, and the key.

[0198] In one embodiment, verifying the identity of the computing device includes: concatenating a message with a certificate to provide first data; decrypting the first data using a key to provide second data; decrypting the second data using an identifier to provide a result; and comparing the result to the key.

[0199] In one embodiment, the identifier is a first public identifier, the computing device stores a first device secret used to generate the first public identifier, and the instructions are further configured to direct the at least one processor to perform the following operations: send a replacement command to the computing device, the replacement command causing the computing device to replace the first device secret with a second device secret, and receive from the computing device a second public identifier generated using the second device secret.

[0200] In one embodiment, the computing device is configured to use the second device secret as input to a message authentication code that provides a derived secret, and to generate the second public identifier using the derived secret.

[0201] In one embodiment, the replace command includes a field having a value based on the first device secret.

[0202] In one embodiment, the system further includes a freshness mechanism configured to generate a freshness, wherein the message sent to the computing device includes the freshness.

[0203] In one embodiment, the identity of the computing device comprises an alphanumeric string.

[0204] In one embodiment, a non-transitory computer storage medium stores instructions that, when executed on a computing device, cause the computing device to at least: receive a message from a host device; generate an identifier, a certificate, and a key, wherein the identifier corresponds to an identity of the computing device and the certificate is generated using the message; and send the identifier, the certificate, and the key to the host device for use in verifying the identity of the computing device.

[0205] In one embodiment, the identifier is a public identifier associated with a private identifier, the key is a public key associated with a private key, and generating a certificate includes: concatenating a message with the public key to provide first data; encrypting the first data using the private identifier to provide second data; and encrypting the second data using the private key to provide a certificate.

[0206] In one embodiment, verifying the identity of the computing device includes performing a decryption operation using the identifier to provide a result, and comparing the result to a key.

[0207] Use Physically Unclonable Function (PUF) to generate values

[0208] At least some embodiments disclosed below provide an improved architecture for generating values ​​using a physically unclonable function (PUF). In some embodiments, the PUF value itself can be used as a device secret, or used to generate a device secret. In one example, the PUF value is used as a device secret as described above (e.g., see Figure 5A and Figure 5B ) is used together with the DICE-RIoT protocol of the PUF. In one example, the value generated by the PUF is used as the input to the message authentication code (MAC). The output from the MAC is used as the UDS.

[0209] In some embodiments, the PUF value or a value generated from the PUF value can be used as a random number (e.g., a device-specific random number). In one example, when the PUF value is generated via the asymmetric key generator described above (e.g., see Figure 6 ) uses a random number (e.g., RND) as input when generating associated public and private keys.

[0210] In general, the following architecture generates an output by feeding the input provided from one or more PUFs into a message authentication code (MAC). The output from the MAC provides an improved PUF (such as, for example, the UDS above).

[0211] Generally, semiconductor chip manufacturers face the problem of key injection, which is the programming of a unique secret key for each chip or die provided, for example, from a semiconductor chip. It is desirable to perform key injection in a secure environment to avoid leaking or revealing the secret key injected into the chip. It is also desirable to ensure that the key cannot be attacked or read back after the chip is produced. In some cases, for example, the key injection procedure is certified or performed by a third-party infrastructure.

[0212] Chip manufacturers want to reduce the production cost of chips that include encryption capabilities. Chip manufacturers also want to simplify the production process while maintaining a constant level of security performance for the manufactured chips. However, key injection is one of the more expensive production steps.

[0213] Chip manufacturers also face the problem of improving the uniformity of PUFs when used as pseudo-random number generators. In some cases, this problem can include cross-correlation between dies due to the phenomena on which the seed values ​​provided by the PUF are based.

[0214] PUF is based on unpredictable physical phenomena, such as on-chip parasitics, on-chip path delays, etc., which are unique to each die. These phenomena are used, for example, to provide seed values ​​for pseudo-random number generators.

[0215] Two different chips selected in a production line must have different PUF values. The PUF value generated in each chip must not change during the lifetime of the device. If two chips have similar keys (e.g., there is a low Hamming distance between them), it is possible to use the key of one chip to guess the key of the other chip (e.g., pre-image hacking).

[0216] Using the improved PUF architecture described below can provide a solution to one or more of the above problems by providing an output value suitable for providing the functionality of a PUF on each chip or die. The improved PUF architecture below uses a PUF that enables each chip or die to automatically generate a unique security key at each power-up of the chip or die. The security key does not need to be stored in non-volatile memory, which can be attacked or otherwise cracked.

[0217] The improved PUF architecture further uses a MAC to generate an improved PUF output (eg, a unique key) for use in a cryptographic function or process, such as integrated into a semiconductor chip. The use of a MAC can, for example, increase the Hamming distance between keys generated on different chips.

[0218] In at least some embodiments disclosed herein, an improved PUF architecture is provided that uses the output from the MAC as a way to generate a seed or other value. Thus, the improved PUF architecture provides, for example, a way to perform key injection that reduces manufacturing costs and improves the reliability and / or uniformity of the PUF operation on the final chip.

[0219] In one embodiment, a method includes: providing at least one value by at least one PUF; and generating a first output based on a MAC, wherein the MAC uses the at least one value provided by the at least one PUF as an input for generating the first output.

[0220] In one embodiment, a system includes: at least one PUF device; a message authentication code MAC module configured to receive a first input based on at least one value provided by the at least one PUF device; at least one processor; and a memory containing instructions configured to direct the at least one processor to generate a first output from the MAC module based on the first input. In various embodiments, the MAC module can be implemented using hardware and / or software.

[0221] In one embodiment, the system further includes a selector module for selecting one or more of the PUF devices for providing values ​​to the MAC module. For example, values ​​provided from several PUF devices may be concatenated and provided as input to the MAC module. In various embodiments, the selector module may be implemented using hardware and / or software.

[0222] Fig.10 A system for generating a unique key 125 from the output of a message authentication code (MAC) 123 that receives input from a physical unclonable function (PUF) device 121 according to one embodiment is shown. The system provides a PUF architecture 111 for generating a unique key 125 (or other value) from the output of a message authentication code (MAC) module 123. The MAC module 123 receives an input value obtained from the physical unclonable function (PUF) device 121.

[0223] Fig.10 The PUF device 121 in the embodiment may be, for example, any one of various known types of PUFs. The MAC module 123 provides, for example, a one-way function such as SHA1, SHA2, MD5, CRC, TIGER, and the like.

[0224] The architecture 111 can, for example, improve the Hamming distance of the PUF values ​​or codes generated between chips. The MAC function is unpredictable (e.g., an input sequence with only a single bit difference provided to the MAC function provides two completely different output results). Therefore, the input to the MAC function cannot be discerned or determined when only having knowledge of the output. The architecture 111 can also, for example, improve the uniformity of the PUF as a pseudo-random number generator.

[0225] In one example, the value generated by the PUF architecture 111 (e.g., the unique key 125 or another value) can be a number having N bits, where N depends on the encryption algorithm implemented on the chip (e.g., the memory device 103 or another device) that includes the PUF architecture 111. In one example, the chip implements an encryption function using HMAC-HA256, in which case the output from the MAC module 123 has a size N of 256 bits. The use of the output from the MAC module 123 provides a message length suitable for use as an output value of the key (without the need for further compression or padding).

[0226] The PUF architecture 111 is implemented in a device such as the illustrated memory device 103, or may be implemented in other types of computing devices such as integrated circuits implemented in a plurality of semiconductor chips provided by a wafer fabrication line.

[0227] In one embodiment, MAC module 123 cooperates with and / or is integrated into or is part of encryption module 127, e.g., which may provide encryption functions for memory device 103. For example, the output of MAC module 123 may be suitable for use as a key due to the MAC used by memory device 103 for other encryption purposes.

[0228] The operation of the PUF architecture 111, the encryption module 127, and / or other functions of the memory device 103 may be controlled by the controller 107. The controller 107 may include, for example, one or more microprocessors.

[0229] exist Fig.10 In the embodiment, the host 101 can communicate with the memory device 103 via a communication channel. The host 101 can be a computer having one or more central processing units (CPUs), and computer peripheral devices (e.g., the memory device 103) can be attached to the computer via an interconnect, such as a computer bus (e.g., Peripheral Component Interconnect (PCI), Extended PCI (PCI-X), PCI Express (PCIe)), a communication portion, and / or a computer network.

[0230] In one embodiment, unique key 125 is used as UDS to provide identity for memory device 103. Controller 107 implements layer 0L0 and layer 1L1 in the DICE-RIoT architecture. In response to receiving a host message from host 101 via host interface 105, cryptographic module 127 performs processing to generate a triplet, as described above. Host 101 uses the triplet to verify the identity of memory device 103. Memory device 103 is an example of computing device 141.

[0231] For the purpose of exemplary illustration, it may be noted that there are generally two technical problems. The first problem is to verify the identity of the board to the host. The problem can be handled by using public triples and asymmetric cryptography, as discussed, for example, above for DICE-RIoT. This method is secure and clever, but in some cases, may be too expensive / time-consuming to be used directly by the circuit board itself. The second problem is to verify the identity of the memory on the board to the board (e.g., to avoid unauthorized memory replacement) (this is performed, for example, after each power-up). The public triples and asymmetric cryptography above can be used to solve the second problem. However, a lighter security mechanism based only on a MAC function is generally sufficient to handle the second problem.

[0232] The memory device 103 may be used, for example, to store data of the host 101 in a non-volatile storage medium 109. Typically, examples of memory devices include a hard disk drive (HDD), a solid state drive (SSD), a flash memory, a dynamic random access memory, a tape, a network attached storage device, etc. The memory device 103 has a host interface 105 that implements communication with the host 101 using a communication channel. For example, in one embodiment, the communication channel between the host 101 and the memory device 103 is a peripheral component interconnect express (PCI express or PCIe) bus; and the host 101 and the memory device 103 communicate with each other using the NVMe protocol (Non-Volatile Memory Host Controller Interface Specification (NVMHCl), also referred to as NVM express (NVMe)).

[0233] In some embodiments, the communication channel between the host 101 and the memory device 103 includes a computer network, such as a local area network, a wireless local area network, a wireless personal area network, a cellular communication network, a broadband high-speed always-connected wireless communication connection (e.g., a current or future generation of mobile network link); and the host 101 and the memory device 103 can be configured to communicate with each other using data storage management and using commands similar to those in the NVMe protocol.

[0234] Controller 107 may run firmware 104 to perform operations in response to communications from host 101 and / or other operations. In general, firmware is a type of computer program that provides control, monitoring, and data manipulation for an engineered computing device. Fig.10 , firmware 104 controls the operation of controller 107 in operating memory device 103, such as the operation of PUF architecture 111, as discussed further below.

[0235] The memory device 103 has a non-volatile storage medium 109, such as a magnetic material coated on a hard disk, and / or a memory cell in an integrated circuit. The storage medium 109 is non-volatile because no power is required to maintain the data / information stored in the non-volatile storage medium 109, and the data / information can be retrieved after the non-volatile storage medium 109 is powered off and then powered on again. The memory cells can be implemented using various memory / storage technologies such as NAND gate-based flash memory, phase change memory (PCM), magnetic memory (MRAM), resistive random access memory, and 3D XPoint, so that the storage medium 109 is non-volatile and can retain the data stored therein for days, months, and / or years without power.

[0236] The memory device 103 includes volatile dynamic random access memory (DRAM) 106 for storing runtime data and instructions used by the controller 107 to improve the computing performance of the controller 107 and / or provide a buffer for data transferred between the host 101 and the non-volatile storage medium 109. The DRAM 106 is volatile because it requires power to maintain the data / information stored therein, which is immediately or quickly lost when power is interrupted.

[0237] Volatile DRAM 106 typically has less latency than non-volatile storage media 109, but loses its data quickly when power is lost. Therefore, it is advantageous to use volatile DRAM 106 to temporarily store instructions and data for controller 107 in its current computing task to improve performance. In some cases, volatile DRAM 106 is replaced by volatile static random access memory (SRAM) that uses less power than DRAM in some applications. When non-volatile storage media 109 has similar data access performance (e.g., in terms of latency, read / write speed) as volatile DRAM 106, volatile DRAM 106 can be removed; and controller 107 can perform calculations by operating on non-volatile storage media 109 for instructions and data instead of operating on volatile DRAM 106.

[0238] For example, cross-point storage and memory devices (e.g., 3D XPoint memory) have data access performance comparable to volatile DRAM 106. Cross-point memory devices use transistor-free memory elements, each of which has a memory cell and a selector stacked together in a column. The memory element columns are connected via two vertical wire layers, one of which is above the memory element columns and the other is below the memory element columns. Each memory element can be selected individually at the intersection of one wire on each of the two layers. Cross-point memory devices are fast and non-volatile, and can be used as a universal memory pool for processing and storage.

[0239] In some cases, the controller 107 has an intra-processor cache memory that has better data access performance than the volatile DRAM 106 and / or the non-volatile storage media 109. Therefore, during computing operations of the controller 107, portions of the instructions and data used in the current computing task are cached in the intra-processor cache memory of the controller 107. In some cases, the controller 107 has multiple processors, each with its own intra-processor cache memory.

[0240] Optionally, controller 107 performs data-intensive in-memory processing using the data and / or instructions organized in memory device 103. For example, in response to a request from host 101, controller 107 performs real-time analysis on a data set stored in memory device 103, and communicates a reduced data set to host 101 in response. For example, in some applications, memory device 103 is connected to a real-time sensor to store sensor input; and a processor of controller 107 is configured to perform machine learning and / or pattern recognition based on the sensor input to support an artificial intelligence (AI) system implemented at least in part via memory device 103 and / or host 101.

[0241] In some embodiments, the processor of controller 107 is integrated with memory (e.g., 106 or 109) in computer chip manufacturing to achieve in-memory processing and thus overcome the von Neumann bottleneck, which limits computing performance due to throughput limitations caused by latency in data movement between processors and memories configured separately according to the von Neumann architecture. The integration of processing and memory increases processing speed and memory transfer rate, and reduces latency and power consumption.

[0242] The memory device 103 can be used in various computing systems, such as cloud computing systems, edge computing systems, fog computing systems, and / or stand-alone computers. In cloud computing systems, remote computer servers are connected to a network to store, manage, and process data. Edge computing systems optimize cloud computing by performing data processing at the edge of a computer network close to the data source, thereby reducing data communications with centralized servers and / or data storage. Fog computing systems use one or more end-user devices or near-user edge devices to store data, thereby reducing or eliminating the need to store data in a centralized data warehouse.

[0243] At least some of the embodiments disclosed herein may be implemented using computer instructions executed by controller 107 (e.g., firmware 104). In some cases, hardware circuits may be used to implement at least some of the functionality of firmware 104. Firmware 104 may be initially stored in non-volatile storage media 109 or another non-volatile device and loaded into volatile DRAM 106 and / or intra-processor cache memory for execution by controller 107.

[0244] For example, firmware 104 may be configured to use the techniques discussed below when operating a PUF architecture. However, the techniques discussed below are not limited to Fig.10 The computer system and / or examples discussed above are used.

[0245] In some implementations, the output of MAC module 123 may be used to provide, for example, a root key or seed value. In other implementations, the output may be used to generate one or more session keys.

[0246] In one embodiment, the output from the MAC module 123 may be transmitted to another computing device. For example, the unique key 125 may be transmitted to the host 101 via the host interface 105 .

[0247] Fig.11 A system is shown for generating a unique key 125 from the output of a MAC 123 that receives input from one or more PUF devices selected by a selector module 204, according to one embodiment. According to one embodiment, the system uses Fig.10 A PUF architecture similar to architecture 111 of PUF 120 but including multiple PUF devices 202 and a selector module 204 generates a unique key 125 from the output of MAC module 123. MAC module 123 receives input from one or more PUF devices 202 selected by selector module 204. In one example, PUF device 202 includes PUF device 121.

[0248] The PUF devices 202 may be, for example, the same or different (e.g., based on different random physical phenomena). In one embodiment, the selector module 204 acts as a smart PUF selection block or circuit to select one or more of the PUF devices 202 from which to obtain values ​​to provide as input to the MAC module 123.

[0249] In one embodiment, the selector module 204 selects the PUF device 202 based at least in part on the results from testing the PUF device 202. For example, the selector module 204 may test the repeatability of each PUF device 202. If any PUF device 202 fails the test, the selector module 204 prevents the failed device from providing input values ​​to the MAC module 123. In one example, the failed device may be excluded temporarily or indefinitely.

[0250] In some implementations, the selector module 204 permits testing of the PUF functionality of each chip during production and / or during use in the field (e.g., by checking the repeatability of the values ​​provided by each PUF device 202). If two or more values ​​provided by a given PUF device are different, the PUF device is determined to have failed and is excluded from use as input to the MAC module 123.

[0251] In one embodiment, the selector module 204 is used to simultaneously use multiple PUF devices 202 as sources for computing an improved PUF output from the MAC module 123. For example, the selector module 204 may concatenate a value from a first PUF device with a value from a second PUF device to provide as input to the MAC module 123. In some implementations, this architecture permits obtaining a robust PUF output due to its dependence on several different physical phenomena.

[0252] Fig.12 A system for generating a unique key from the output of a MAC that receives input from one or more PUF devices and input from a monotonic counter 302 (and / or input from another freshness mechanism such as a NONCE, a timestamp, etc.) is shown according to one embodiment. According to one embodiment, the system generates a unique key 125 from the output of a MAC module 123. Fig.12 The PUF architecture described in Fig.11 , except that a monotonic counter 302 is included to provide a value to the selector module 204. In different embodiments, the monotonic counter 302 may be implemented using hardware and / or software.

[0253] The MAC module 123 receives input from one or more PUF devices 202 and input from a monotonic counter 302. In one example, the values ​​obtained from the PUF device 202 and the monotonic counter 302 are concatenated and then provided as input to the MAC module 123. In some implementations, the monotonic counter 302 is a non-volatile counter that increments its value only when requested. In some embodiments, the monotonic counter 302 is incremented after each power-on cycle of the chip.

[0254] In some embodiments, Fig.12 The PUF architecture can be used to provide a way to securely share keys between semiconductor chips and other components in applications, such as a public key mechanism.

[0255] In some embodiments, the monotonic counter 302 is incremented before each calculation of the PUF, which ensures that the input to the MAC module 123 is different at each cycle, and therefore the output (and / or pattern of outputs) provided is different. In some examples, this method can be used to generate session keys, where each session key is different.

[0256] In some embodiments, the selector module 204 may selectively include or reject the monotonic counter 302 (or other freshness mechanisms such as NONCE, timestamp) to provide the counter value as input to the MAC module 123 .

[0257] In some embodiments, the monotonic counter 302 is also used by the encryption module 127. In some embodiments, the PUF architecture including the monotonic counter can be used as a session key generator to ensure a different key at each cycle. In some implementations, the generated session key is protected in this way: Session Key = MAC [One or more PUFs | MTC or other freshness]

[0258] In other embodiments, the mechanism is used as follows:

[0259] Session Key = MAC key_based [Root_Key, MTC or other freshness mechanism]

[0260] Where: Root_Key = the output value provided from the MAC module 123 above, or any other kind of key present on the chip.

[0261] MAC above key_based The function is, for example, a MAC algorithm based on a secret key. For example, there may be two types of MAC algorithms in cryptography:

[0262] 1. Secret key based algorithms, similar to e.g. the HMAC family (HMAC-SHA256 is key based).

[0263] 2. Algorithms that are not based on secret keys, such as SHA256 (standalone SHA is not key-based).

[0264] It should be noted that a key-based MAC can be transformed into a non-key-based MAC by setting the key to a known value (e.g., 0x000...0xFFFF, etc....).

[0265] Fig.13 A method of generating an output from a MAC using one or more input values ​​provided by one or more PUFs is shown according to one embodiment. For example, Fig.13 The method can be found in Fig.10 is implemented in the memory device 103.

[0266] Fig.13 The method includes, at block 411, providing one or more values ​​by at least one PUF (eg, providing values ​​from one or more of the PUF devices 202).

[0267] At block 413, one or more of the PUFs may be tested for repeatability, such as described above. This testing is optional.

[0268] At block 415, if the test has been performed at block 413, and it has been determined that the PUF device has failed the test, then the failed PUF device is denied from providing input to the MAC. This denial may be performed, for example, by the selector module 204, as discussed above.

[0269] At block 417, a value is provided from a monotonic counter, such as monotonic counter 302. The use of a monotonic counter in the PUF architecture is optional.

[0270] At block 419, an output is generated from a MAC that uses one or more values ​​provided by the PUF (and optionally at least one value from a monotonic counter) as input to the MAC.

[0271] Various other embodiments are now described below for a method implemented in a computing device, the method comprising: providing at least one value by at least one physically unclonable function (PUF); and generating a first output based on a message authentication code (MAC), wherein the MAC uses at least one value provided by at least one PUF as an input for generating the first output.

[0272] In one embodiment, the computing device is a first computing device, and the method further includes transmitting the first output to the second computing device, wherein the first output is a unique identifier of the first computing device.

[0273] In one embodiment, providing at least one value includes selecting a first value from a first PUF and selecting a second value from a second PUF.

[0274] In one embodiment, the method further comprises: providing a value from a monotonic counter; wherein generating the first output further comprises using the value from the monotonic counter as an additional to the MAC for generating the first output.

[0275] In one embodiment, the method further includes: generating a plurality of session keys based on respective outputs provided by a MAC, wherein a monotonic counter provides a value used as input to the MAC; and incrementing the monotonic counter after generating each of the session keys.

[0276] In one embodiment, the method further includes: testing a first PUF of the at least one PUF for repeatability; and based on determining that the first PUF failed the test, denying the first PUF from providing any input to the MAC when generating the first output.

[0277] In one embodiment, the testing includes comparing two or more values ​​provided by the first PUF.

[0278] In one embodiment, the computing device is a memory device, and the memory device includes a non-volatile storage medium configured to store output values ​​generated using a MAC.

[0279] In one embodiment, the method further comprises executing, by at least one processor, at least one cryptographic function, wherein executing the at least one cryptographic function comprises using an output value generated using the MAC.

[0280] In one embodiment, a non-transitory computer storage medium stores instructions that, when executed on a memory device (e.g., memory device 103), cause the memory device to perform a method comprising: providing at least one value by at least one physically unclonable function (PUF); and generating a first output based on a message authentication code (MAC), wherein the MAC uses at least one value provided by at least one PUF as an input for generating the first output.

[0281] In various other embodiments described below, the system may be executed Figure 4 The system comprises: at least one physically unclonable function (PUF) device; a message authentication code (MAC) module configured to receive a first input based on at least one value provided by the at least one PUF device; at least one processor; and a memory containing instructions configured to direct the at least one processor to generate a first output from the MAC module based on the first input.

[0282] In one embodiment, the MAC module includes circuitry. In one embodiment, a first output from the MAC module is a key identifying the die. In one embodiment, the first output from the MAC module is a root key, and the instructions are further configured to direct the at least one processor to generate a session key using the output from the MAC module.

[0283] In one embodiment, the system is part of a semiconductor chip (e.g., one of several chips obtained from a semiconductor wafer), the first output from the MAC module is a unique value that identifies the chip, and the instructions are further configured to direct the at least one processor to transmit the unique value to a computing device.

[0284] In one embodiment, the at least one PUF device includes a plurality of PUF devices (eg, PUF device 202), and the system further includes a selector module configured to select the at least one PUF device that provides the at least one value.

[0285] In one embodiment, the selector module is further configured to generate a first input for the MAC module by concatenating a first value from the first PUF device and a second value from the second PUF device.

[0286] In one embodiment, the system further includes a monotonic counter configured to provide a counter value, and the instructions are further configured to direct the at least one processor to generate the first input by concatenating the counter value with at least one value provided by at least one PUF device.

[0287] In one embodiment, the system further comprises a selector module configured to select at least one PUF device providing at least one value, wherein linking the counter value with the at least one value provided by the at least one PUF device is performed by the selector module.

[0288] In one embodiment, the monotonic counter is further configured to increment the counter value after generating the first input to provide an incremented value; and the instructions are further configured to direct the at least one processor to generate a second output from the MAC module based on the incremented value and at least one new value provided by the at least one PUF device.

[0289] Fig.14 A system for generating a root key from the output of a MAC that receives input from one or more PUF devices and input from a monotonic counter (and / or input from another freshness mechanism such as a NONCE, timestamp, etc.), and adds additional MACs to generate session keys, according to one embodiment, is shown.

[0290] In one embodiment, according to one embodiment, the system generates a root key from the output of a MAC that receives input from one or more PUF devices 202 and input from a monotonic counter 302 (and / or input from another freshness mechanism such as a NONCE, a timestamp, etc.), and adds an additional MAC module 504 to generate a session key using the root key input. In this embodiment, the MAC module 123 provides the root key 502 as an output from the MAC module 123. The root key 502 is an input to the MAC module 504, which may use a MAC function such as session key=MAC key_based [Root_Key, MTC or other freshness mechanism], which is described above. The root key input in this key-based function may be the root key 502, as illustrated.

[0291] Furthermore, in one embodiment, the monotonic counter 302 may provide input to the MAC module 504. In other embodiments, a different monotonic counter or other value from the chip may be provided as an input to the MAC module 504 instead of using the monotonic counter 302. In some cases, the monotonic counter 302 provides the counter value to the MAC module 504 but not to the selector module 204. In other cases, the counter value may be provided to both MAC modules, or excluded from both modules.

[0292] Key generation and secure storage

[0293] As mentioned above, PUFs can be used for secure key generation. Various embodiments discussed below relate to generating an initial key using at least one PUF, applying processing to increase obfuscation of the initial key, and storing the final obfuscated key in non-volatile memory. The final obfuscated key and / or the intermediate key used to generate the final obfuscated key can be shared with another computing device and used for secure communication with the other computing device (e.g., messaging using symmetric cryptography based on the shared key). In some embodiments, secure key generation is performed for a computing device to be used in an automotive application (e.g., a controller in an autonomous vehicle).

[0294] In alternative embodiments, the initial key is generated in other ways that do not require the use of the above at least one PUF. In one embodiment, the initial key can be generated by using an injected key. For example, the initial key exists in the chip due to injection in a factory or other secure environment. In this case, the application process is performed by applying the obfuscation process to the injected key to increase the obfuscation of the initial key.

[0295] The automotive environment presents a technical problem of introducing "noise" during the key generation phase. The various embodiments below provide a technical solution to this problem by using a method that attenuates or avoids key changes due to this induced noise by storing obfuscated keys inside a non-volatile memory area.

[0296] The automotive environment can affect key generation in various ways. For example, powering up the engine can cause a drop in applied power to the computing device, causing keys to be generated in an erroneous manner. Temperature extremes can also affect the circuits that generate the keys. Other sources, such as magnetic fields from power lines, can cause inter-symbol interference or crosstalk, making the host unable to recognize the device.

[0297] In contrast, if the keys are generated in a secure environment and stored in memory, they will not be affected by noise. Depending on the strategy used to propagate the keys among end users / customers of the computing device product, the secure environment may be installed directly in the car, in a test environment, or in a factory (e.g., where the computing device generating the keys is manufactured), for example.

[0298] In one example, an ADAS or other computing system such as that used in a vehicle is subject to a power supply change. This may occur, for example, during opening the vehicle, braking, powering the engine, etc.

[0299] Various embodiments of generating and storing keys as discussed below provide the advantage of being substantially independent of external factors (e.g., power supply variations, temperature, and other external sources of noise). For example, another advantage in some embodiments is that the generation of the key vector is the same for each cycle.

[0300] When storing a key, another advantage provided in some embodiments is that the key is substantially immune to hardware attacks (e.g., that a hacker might have in place). For example, one such attack is monitoring the power-on current of the device in order to correlate current changes with bits associated with the key. Other attacks may use, for example, voltage measurements (e.g., Vdd supply voltage). Some attacks may use, for example, temperature changes to interfere with the operation of the device.

[0301] In some embodiments, the above-described Figures 10 to 14 The described methods and / or architectures generate an initial key. For example, a PUF is used to generate a key for each power-on cycle of a computing device storing a key. In alternative embodiments, other methods may be used to generate an initial key.

[0302] In one exemplary method, as discussed above, key injection uses at least one PUF and a MAC algorithm (e.g., SHA256) to generate a key for a device that is significantly different from other devices (e.g., from an adjacent die located on a wafer). The MAC cryptography algorithm provides the benefit of increasing the entropy of the bits generated by the PUF.

[0303] In one embodiment, the generated key (e.g., an initial key as provided from a PUF and then a MAC algorithm) is stored in a non-volatile area of ​​the device after preprocessing the key to mitigate or avoid hacker attacks and also improve the reliability of the stored key. In one embodiment, after the key is stored, the circuit that generates the key can be disabled. Preprocessing is generally referred to as obfuscation in this article. In one example, circuits and / or other logic are used to implement obfuscation on the device. In one instance, the stored key can be read by the device because the key is independent of external noise sources. An internal mechanism is used to read any data of the device.

[0304] In various embodiments, storing the key as described herein increases tolerance to noise. Also, it makes it difficult for a hacker to read the stored key, for example using power monitoring or other hacking methods.

[0305] At least some embodiments herein use PUFs and encryption algorithms (e.g., HMAC-HA256) to make key generation independent of external factors (e.g., temperature or voltage) that may otherwise cause the keys to differ from one power-on of the device to the next. If this occurs, it may be a problem for the host to be able to exchange messages with the device. Various embodiments make key generation more robust by placing the stored keys in memory so that they are not affected by external factors.

[0306] In one embodiment, the key is generated once on the device and stored in the non-volatile memory of the device. In one example, the key can be generated using the contents of the SRAM before a reset is applied to the SRAM. The key as a function of the PUF is generated using the pseudo-random value output from the PUF. The contents of the SRAM are read before a reset of the appliance or other device. The key can also be regenerated at other times via a command sequence as may be needed. In one example, the generated key is used as a UDS in the DICE-RIoT protocol as described above. In one example, the command sequence uses a replace command to replace a previously generated UDS with a new UDS, as described above.

[0307] In one embodiment, the key generation is independent of the cryptography implemented by the device. The generated key is shared with the host. This embodiment stores the key and / or reads the key in the device in a way that prevents an attacker from guessing the key and using the key internally, for example by analyzing the shape of the current drawn by the device during key use.

[0308] In addition, for example, in asymmetric cryptography, the generated key becomes a variable password for the system's secret key. The key is not shared with others. For public key cryptography, the key is used to generate a corresponding public key.

[0309] In various embodiments, an initial key is generated using an injected key or using one or more PUFs (e.g., to provide an initial key PUF0). The initial key then undergoes one or more steps of an obfuscation process to provide an intermediate key (e.g., PUF1, PUF2, ... PUF5), as described below. The output from this process (e.g., PUF5) is an obfuscated key stored in a non-volatile memory of the device. When an injected key is used, an obfuscation process is applied to the injected key similarly as described below for the non-limiting example of PUF0.

[0310] In one embodiment, as mentioned above, for the case of initial injection of keys, the following mechanism is used:

[0311] Session Key = MAC key_based [Root_Key, MTC or other freshness mechanism]

[0312] Where: Root_Key = any other kind of key present on the chip (for example, the key may be the initial key injected into the chip at the factory or other secure environment)

[0313] In one embodiment, at the first power-up of the device, a special sequence wakes up at least one circuit of the device (e.g., a read circuit) and verifies that the circuit is performing properly. The device then generates an initial key PUF0, as mentioned above. This key can be stored or further processed to make it more robust for secure storage, as described below.

[0314] The intermediate key PUF1 is generated by concatenating PUF0 with a predetermined bit sequence (eg, a sequence known to others) to generate PUF1. In one embodiment, PUF1 is used to verify the ability of the device to correctly read the key and ensure that noise (eg, fluctuations in the power supply) does not affect the generated key.

[0315] The next intermediate key PUF2 is generated. PUF1 is interleaved with the reverse bit pattern (eg, formed by reversing the bits of PUF1 and sometimes referred to herein as PUF1 strips) to generate PUF2.

[0316] In one embodiment, PUF2 has the same number of bits 0 and 1. This makes the shape of the device current essentially the same for any key (e.g., any key stored on the device). This reduces the possibility of an attacker guessing the key value by looking at the shape of the device current when the device reads the key.

[0317] The next intermediate key PUF3 is generated. The bits of PUF2 are interleaved with pseudo-random bits to form PUF3. This further helps to obfuscate the key. In one embodiment, pseudo-random bits are derived from PUF1 or PUF2 by using a hash function. For example, these derived bits are added to PUF2 to form PUF3.

[0318] The next intermediate key PUF4 is generated. An error correction code (ECC) is generated by the internal circuitry of the device (e.g., during programming). The bits of the ECC are added to PUF3 to generate PUF4. In one embodiment, the ECC bits help guard against the effects of non-volatile memory (e.g., NVRAM) aging that can be caused by, for example, device endurance limits, X-rays, and particles. Non-volatile memory aging can also be caused, for example, by an increase in the number of electrons located in the NV cell that can cause a bit flip.

[0319] The next intermediate key PUF5 is generated. PUF5 is a concatenation of several copies of PUF4. The redundancy of having multiple copies of PUF4 present in PUF5 further increases robustness by increasing the likelihood that the key can be correctly read at a later time. In one embodiment, several copies of PUF5 are stored in various non-volatile memory storage areas to further increase robustness. For example, even if PUF5 is damaged in one of the areas, PUF5 can be read from other areas of the area and the correct key can be extracted.

[0320] In one embodiment, PUF1 or PUF3 is a key shared with the host for symmetric encryption, or used to generate public keys for asymmetric cryptography. In one embodiment, PUF4 and PUF5 are not shared with the end user or the host.

[0321] The above method is modular because PUF2, PUF3, PUF4 and / or PUF5 are not required to generate the obfuscated key. Alternatively, in various embodiments, one or more of the aforementioned obfuscation steps may be applied to the initial key, and the order may further be changed. For example, for systems known not to have a Vdd voltage supply drop, the number of obfuscation steps may be reduced.

[0322] In one embodiment, when the obfuscation key is stored, the bit pattern will be physically spread around the non-volatile storage medium (e.g., in different rows and columns). For example, a device is able to read bits simultaneously and prevent multi-bit errors from occurring.

[0323] Fig.15 A computing device 603 is shown for storing an obfuscated key 635 in a non-volatile memory (e.g., non-volatile storage medium 109) according to one embodiment. The computing device 603 is Figure 1141 of a computing device. In one example, an obfuscated key is used as UDS. (Note) For example, obfuscation adds entropy to the bits of the key to avoid possible attempts by a hacker to learn the value of the key. The device is always able to extract the key by removing the added bits used for obfuscation. In one example, a common hacking attack consists of guessing a secret key generated / detailed inside the device by processing the distribution of current drawn by the device within a certain time frame with statistical tools. Obfuscation mitigates this problem in a manner that is worth considering.

[0324] The initial key 625 is generated based on a value provided by the at least one physical unclonable function device 121. The obfuscated key 635 is generated based on the initial key 625. After being generated, the obfuscated key 635 is stored in the non-volatile storage medium 109.

[0325] In one embodiment, the message authentication code (MAC) 123 uses the value from the PUF device 121 as input and provides as output an initial key 625. In one embodiment, the obfuscation processing module 630 is used to perform processing on the initial key 625 to provide an obfuscated key 635 (e.g., PUF5), such as discussed above.

[0326] In one embodiment, the obfuscated key 635 is securely distributed to another computing device, as described in related U.S. non-provisional application Ser. No. 15 / 965,731, filed Apr. 27, 2018, by Mondello et al., entitled "SECURE DISTRIBUTION OF SECRET KEY USING A MONOTONIC COUNTER," the entire contents of which are incorporated by reference as if fully set forth herein. In other embodiments, the initial key 625 and / or any one or more of the intermediate keys from the obfuscation process described herein may be securely distributed in the same or similar manner. Optionally, the end user / customer reads the value of the initial key (e.g., PUF0), the intermediate key, and / or the final obfuscated key (e.g., PUF5) using the aforementioned methods. For example, the end user may verify the proper execution of the internal generation of the key through the device, and / or monitor the statistical quality of the key generation.

[0327] Fig.16 An example of an intermediate key (PUF2) generated by the obfuscation processing module 630 during the obfuscation process according to one embodiment is shown. As mentioned above, the bits of PUF1 are reversed to provide reversed bits 702. Bits 702 are interleaved with the bits of PUF1 as illustrated. For example, every second bit in the illustrated key is an interleaved reversed bit 702.

[0328] Fig.17Demonstrating an embodiment of Fig.16 802. As mentioned above, the bits of PUF2 are further interleaved with pseudo-random bits 802. As illustrated, bits 802 are interleaved with PUF2. For example, every third bit in the illustrated key is an interleaved pseudo-random bit 802.

[0329] Fig.18 A method for generating and storing an obfuscation key (e.g., obfuscation key 635) in a non-volatile memory (e.g., non-volatile storage medium 109) according to one embodiment is shown. In one example, Figure 2 The memory system 105 stores the obfuscated key in the non-volatile memory 121.

[0330] In block 911, an initial key is generated based on a value provided by at least one physically unclonable function (PUF).

[0331] In other embodiments, an initial key is generated by key injection in block 911. For example, the initial key may simply be a value injected into the chip during manufacturing.

[0332] In block 913, an obfuscation key is generated based on the initial key. For example, the generated obfuscation key is PUF3 or PUF5.

[0333] In block 915, the obfuscation key is stored in a non-volatile memory of the computing device. For example, the obfuscation key is stored in NAND flash memory or EEPROM.

[0334] In one embodiment, the method includes: generating an initial key using key injection; generating an obfuscated key based on the initial key; and storing the obfuscated key in a non-volatile memory. For example, the initial key may be a key injected during a key injection process at manufacturing time.

[0335] In one embodiment, the method includes: generating an initial key provided by key injection or generating an initial key based on a value provided by at least one physically unclonable function (PUF); generating an obfuscated key based on the initial key; and storing the obfuscated key in a non-volatile memory of a computing device.

[0336] In one embodiment, generating the initial key includes using a value from the PUF (or, for example, another value on the chip) as an input to a message authentication code (MAC) to generate the initial key.

[0337] In one embodiment, the obfuscation key is stored in non-volatile memory outside of the user addressable memory space.

[0338] In one embodiment, generating the obfuscation key includes concatenating the initial key with a predetermined bit pattern.

[0339] In one embodiment, concatenating the initial key with a predetermined bit pattern provides a first key (eg, PUF1); and generating the obfuscated key further includes interleaving the first key with an inverted bit pattern, wherein the inverted bit pattern is provided by inverting bits of the first key.

[0340] In one embodiment, interleaving the first key with an inverted bit pattern provides a second key (eg, PUF2); generating the obfuscated key further includes interleaving the second key with pseudo-random bits.

[0341] In one embodiment, the method further comprises deriving pseudo-random bits from the first key or the second key using a hash function.

[0342] In one embodiment, interleaving the second key with pseudo-random bits provides a third key (eg, PUF3); and generating the obfuscated key further includes concatenating the third key with error correction code bits.

[0343] In one embodiment, the computing device is a first computing device, and the method further includes sharing at least one of the initial key, the first key, or the third key with a second computing device, and receiving a message encrypted using the shared at least one of the initial key, the first key, or the third key from the second computing device.

[0344] In one embodiment, concatenating the third key with the error correction code bits provides a fourth key (eg, PUF4); and generating the obfuscated key further includes concatenating the fourth key with one or more copies of the fourth key.

[0345] In one embodiment, concatenating a fourth key with one or more copies of the fourth key provides a fifth key (e.g., PUF5); and storing the obfuscated key includes storing a first copy of the fifth key on at least one of rows or blocks of a non-volatile memory that is different from a row or block in which a second copy of the fifth key is stored.

[0346] In one embodiment, a system includes: at least one physical unclonable function (PUF) device (e.g., PUF device 121) configured to provide a first value; a non-volatile memory (e.g., non-volatile storage medium 109) configured to store an obfuscation key (e.g., key 635); at least one processor; and a memory containing instructions configured to direct the at least one processor to: generate an initial key based on the first value provided by the at least one PUF device; generate an obfuscation key based on the initial key; and store the obfuscation key in the non-volatile memory.

[0347] In one embodiment, the system further includes a message authentication code (MAC) module (e.g., MAC 123) configured to receive a value provided by at least one PUF device, wherein generating the initial key includes using the first value as an input to the MAC module to generate the initial key.

[0348] In one embodiment, generating the obfuscated key includes at least one of: concatenating the key with a predetermined bit pattern; interleaving the first key with the inverse bit pattern of the first key; interleaving the key with pseudo-random bits; concatenating the key with error correction code bits; or concatenating the second key with one or more copies of the second key.

[0349] In one embodiment, the stored obfuscation key has an equal number of zero bits and one bits.

[0350] In one embodiment, generating the obfuscated key includes concatenating the initial key with the first bit pattern.

[0351] In one embodiment, concatenating the initial key with the first bit pattern provides the first key; and generating the obfuscated key further includes interleaving the first key with the second bit pattern.

[0352] In one embodiment, generating the obfuscation key further comprises interleaving the key with pseudo-random bits.

[0353] In one embodiment, generating the obfuscation key further comprises concatenating the key with error correction code bits.

[0354] In one embodiment, a non-transitory computer storage medium stores instructions that, when executed on a computing device, cause the computing device to perform a method comprising: generating an initial key using at least one physically unclonable function (PUF); generating an obfuscated key based on the initial key; and storing the obfuscated key in a non-volatile memory.

[0355] Fig.19 A computing device 1003 is shown for generating an initial key 625 based on key injection 1010, obfuscating the initial key, and storing the obfuscated key in a non-volatile memory, according to one embodiment.

[0356] In one embodiment, the initial key 625 is generated by using the injection key 1010. For example, the initial key 625 is present in the chip by being injected into a factory or other secure environment during manufacturing or other assembly or testing. In one example, the initial key 625 is used as the initial UDS for the computing device 1003. Obfuscation can also be applied to the UDS. The UDS is a securely generated secret that DICE-RIoT begins using to generate keys and certificates. Application processing is performed to increase the obfuscation of the initial key by applying obfuscation processing (via module 630) to the injected key (such as the value from key injection 1010). In other embodiments, obfuscation processing can be applied to any other value that may be stored or otherwise present on the chip or die.

[0357] Changes in key generation and secure storage

[0358] Various additional non-limiting embodiments are now described below. In one embodiment, after a first power-up of the system board (or during said first power-up), a special sequence is activated to turn on a device containing a cryptographic engine (e.g., cryptographic module 127). The sequence further wakes up the internal PUF and verifies its functionality, after which the PUF generates an initial value PUFO, e.g., as described above. The PUFO value is processed by an on-chip algorithm (e.g., by obfuscation processing module 630) and written in a special area of ​​the non-volatile array (outside of the user addressable space). In an alternative embodiment, instead of the PUFO value, the injected key is processed by an on-chip algorithm similar to that described below to provide an obfuscated key for storage.

[0359] In one embodiment, an obfuscation process is performed to prevent Vdd (voltage) and / or temperature fault hacking attacks. This process includes concatenating PUF0 with a well-known pattern (e.g., it contains a fixed amount of 0 / 1 bits). During the life of the device (e.g., chip), when the PUF value is read internally, these bits permit determination of whether the read circuitry can properly discern 0 / 1 bits. For example, PUF1 = PUF0||010101...01

[0360] Next, the result of the above processing (e.g., PUF1) is further embodied with virtual bits (e.g., to avoid Icc hacker analysis). Specifically, for example, the bits of PUF1 are interleaved with the inverted version of PUF1 (i.e., PUF1 strips, which are formed by inverting each bit of PUF1). For example, PUF2 = PUF1 interleaved PUF1 strips.

[0361] In one embodiment, the rules for interleaving depend on the kind of column decoder (e.g., NV non-volatile array) present on the chip / device. The device ensures that at each read of the PUF value (from the non-volatile array), the read circuit processes (in a single shot) the same number of bits from PUF1 and PUF2 strips. This ensures that the same number of bits are read at values ​​of 0 and 1, which provides a regular shape in the supply current (Idd).

[0362] Next, the bits of PUF2 are further interleaved with pseudo-random bits. In one example, the interleaving depends on the non-volatile array column decoder structure. In one embodiment, the output has the same number of PUF2 bits filled with a specific number of pseudo-random bits (e.g., in order to obfuscate the final residual correlation that may exist in the PUF2 pattern).

[0363] In one embodiment, pseudo-random bits may be derived from PUF1 or PUF2 by using a hash function.Other alternative methods may also be used.

[0364] In one embodiment, optionally, to reduce or prevent bit loss due to non-volatile aging, the bits of PUF3 are concatenated with error correction code (ECC) bits. In one embodiment, the bits of PUF4 are optionally replicated one or more times (which also extends ECC capabilities). For example, the foregoing can be implemented on NAND memory. In one example, PUF5 = PUF4 || PUF4 || ... || PUF4

[0365] In one embodiment, the value of PUF5 may be written two or more times on different rows and or blocks of the non-volatile memory array.

[0366] As a result of the obfuscation process described above, for example, once the final PUF value is written to a non-volatile array block, it may be used with diminished key reliability or no concern for key reliability (e.g., due to noise or charge loss), or any attempt to infer its value through Idd analysis or force its value through Vdd fault attacks.

[0367] In one embodiment, once the obfuscation process is completed, the PUF circuit may be disabled. In one embodiment, after deactivation, the PUF device may provide values ​​used internally on the device for other purposes (eg, using standard read operations internal to the non-volatile array).

[0368] In one embodiment, the key bits are different from the random bits when the key is extracted from PUF3. For example, the internal logic of the device storing the key knows the location and method required to return from PUF5 to the previous or original PUF (eg, PUF3).

[0369] In one embodiment, the bit positions of the key bits are known to the device extracting the key. For example, depending on design choice, the internal logic of the device may receive one of the intermediate PUF or the final key PUF5. Subsequently, applying one or more operations in reverse order will obtain the original PUF. For example, the processing steps from PUF1 to PUF5 are performed to store the obfuscated PUF in a way that a hacker would have to both: read the content (e.g., key bits), and also know the operations applied in order to go back and determine the original key.

[0370] Using PUF to generate identity of computing device

[0371] Various embodiments related to generating an identity of a computing device using a Physically Unclonable Function (PUF) are now described below. The generality of the following description is not limited by the various embodiments described above.

[0372] In previous approaches, it was necessary for the manufacturer of a computing device to share one or more secret keys with the customer who purchased the computing device in order to establish the identity of the computing device. However, the sharing of secret keys poses technical problems due to the cumbersome, complex, and expensive secure channels and infrastructure required for sharing the keys with customers. Furthermore, human services are required to implement the key sharing. Furthermore, the aforementioned security requirements may increase the risk that the security measures may be broken by hackers or other unauthorized persons.

[0373] Various embodiments of the present disclosure discussed below provide technical solutions to the above technical problems. In various embodiments, a computing device generates an identity using one or more PUFs. In one example, the identity is a UDS.

[0374] In one embodiment, generation of the computing device's identity is assigned in an automatic manner (e.g., based on a scheduled time or occurrence of a predetermined event, in response to which the computing device will self-generate a UDS using the PUF). By using PUF to assign identities, the complexity and expense of identity assignment can be reduced.

[0375] After the computing device generates an identity, it can be used to generate a triple of an identifier, a certificate, and a key. In one embodiment, the triple is generated in response to receiving a message from a host device. The host device can use the generated identifier, certificate, and key to verify the identity of the computing device. After verifying the identity, further secure communications between the host device and the computing device can be performed using the key.

[0376] In some embodiments, the computing device generates an identity in response to receiving a command from a host device. For example, the command may be a secure replacement command authenticated by the computing device. After generating the identity, the computing device sends a confirmation message to the host device to confirm the generation of the replacement identity. In one example, the replacement identity is a new UDS stored in non-volatile memory and replaces a previously stored UDS (e.g., a UDS assigned by the original manufacturer of the computing device).

[0377] In one embodiment, the identity is a device secret stored in a memory of the computing device (e.g., a UDS as used in the DICE-RIoT protocol, such as discussed above). At least one value is provided by one or more PUFs of the computing device. The computing device generates the device secret using a key derivation function (KDF). The values ​​provided by the one or more PUFs are inputs to the KDF. The output of the KDF provides the device secret. The output of the KDF is stored as the device secret in the memory of the computing device.

[0378] In one example, the KDF is a hash. In one example, the KDF is a message authentication code.

[0379] In one embodiment, the computing device stores a secret key for communicating with a host device and the KDF is a message authentication code (MAC). At least one value provided by the one or more PUFs is a first input to the MAC and the secret key is used as a second input to the MAC.

[0380] In some examples, the computing device may be a flash memory device. For example, a serial NOR may be used.

[0381] Fig. 20 Computing device 141 is shown as being used to generate an identity (e.g., a UDS of computing device 141) using a physically unclonable function (PUF) 2005, according to one embodiment. More specifically, a value is provided by PUF 2005. This value is provided as input to a key derivation function (KDF) 2007. The output from the KDF is stored as device secret 149.

[0382] In one embodiment, device secret 149 is a UDS as used by the DICE-RIoT protocol. Similar to as described above, UDS can be used as a basis for generating a triplet for sending to host device 151. This triplet contains a public key that can be used by host device 151 to communicate securely with computing device 141.

[0383] In one embodiment, the generation of the device secret is performed in response to receiving a command from the host device 151 via the host interface 2009. In one example, the command is a replace command. In one example, the command is accompanied by a signature signed by the host device 151 using a secret key. After the device secret is generated, a confirmation message is sent to the host device 151 via the host interface 2009.

[0384] In one embodiment, computing device 141 stores secret key 2013. For example, secret key 2013 may be shared with host device 151. In one example, host device 151 uses secret key 2013 to sign a signature sent with a replace command.

[0385] In one embodiment, KDF 2007 is a message authentication code. Secret key 2013 is used as a key input to KDF 2007 when generating the device secret. The value from PUF 2005 is used as a data input to KDF 2007.

[0386] A freshness mechanism is implemented in computing device 141 using monotonic counter 2003. Monotonic counter 2003 may provide a value used as freshness in secure communications with host device 151.

[0387] A unique identifier (UID) 2001 is stored in the memory of the computing device 141. For example, the UID 2001 is injected at the factory.

[0388] In one embodiment, computing device 141 is delivered to the customer with a well-known UDS (e.g., common UDS=0x00000...000 or similar). The customer uses host device 151 in its factory to request computing device 141 to self-generate a new UDS (e.g., UDS_puf). This step may be accomplished using an authentication command. Only a customer or host device that knows secret key 2013 is able to do this (e.g., secret key 2013 may be more generally used to manage the set of authentication commands supported by computing device 141).

[0389] Once the UDS_puf is generated, it is used to replace the original (normal) UDS. The replacement occurs by using an authentication command. An external host device (client) can read the UDS_puf.

[0390] The generated UDS_puf may be used to implement the DICE-RIoT protocol. For example, the FDS may be calculated using the UDS_puf, similarly as described above for identity component 147 and identity component 107. In one example, FDS = HMAC-SHA256 [UDS, SHA256 ("identity of L1")].

[0391] In addition, UDS_puf can be used to generate a triplet containing an identifier, a certificate, and a key (e.g., K L1 ), similar to the above for Figure 1 The host device sends a key (eg, K L1 public) for trusted communications with computing device 141.

[0392] In one embodiment, the above identity generation mechanism may be automatically executed by a computing device (e.g., an application board including a processor) upon first use of the application board, or executed in the field upon the occurrence of a scheduled or predetermined event (e.g., as scheduled / determined by a customer and stored in memory 145 as a configuration of the computing device (e.g., update, etc.)).

[0393] In one embodiment, self-identity generation is performed as follows: A configurator host device (e.g., a laptop computer with software) is connected to a computing device coupled to an autonomous vehicle bus (e.g., using a secure air interface, etc.). The host device uses an authentication command to request the computing device to self-generate a UDS (e.g., a UDS PUF ). Authentication is based on a secret key 2013 (eg, the secret key may be injected by the manufacturer and provided to the customer using a secure infrastructure).

[0394] Authentication command execution is performed through an authenticated response (e.g. Fig. 20 In UDS PUF When the generation is complete, the host device is notified about the UDS generated by using a secure protocol (e.g., by sending via a secure wired and / or wireless network using the freshness provided by the monotonic counter 2003). PUF .

[0395] In one embodiment, host interface 2009 is a command interface that supports authentication and replay of protected commands. In one embodiment, authentication is based on a secret key (e.g., secret key 2013) and uses a MAC algorithm (e.g., HMAC). In one example, an identity generation command is received from host device 151 that includes a signature based on the command opcode, command parameters, and freshness. In one example, signature = MAC (opcode | parameters | freshness, secret key).

[0396] In one embodiment, computing device 141 provides an identity generation confirmation including a signature. In one example, signature = MAC (command result | freshness, secret key).

[0397] In one embodiment, the secret key 2013 is injected at the factory. In one example, the secret key can be symmetric (e.g., based on HMAC-SHA256). In another example, the secret key can use an asymmetric scheme (e.g., ECDSA).

[0398] Various options may be used to generate the device secret 149 (e.g., UDS PUF For example, once the generation process is activated by using the appropriate authentication and replay protected commands, the UDS is generated according to the command options selected as follows PUF :

[0399] Option #1: Read at least one value from a PUF pattern (sometimes referred to as "PUF RAW") (sometimes also denoted as UDS RAW ) and feeds it into a KDF 2007 (key derivation function) (e.g., SHA256). The result of this process is the final key. Therefore: UDS PUF =KDF(UDS RAW )

[0400] Option #2: Read PUF RAW Mode (UDS RAW ) and UDS PUF Calculated as: UDS PUF =KDF(UDS RAW |UID), where UID is a public unique ID assigned to all devices (in some cases, UIDs may also be assigned to non-secure devices). In one embodiment, UID 2001 is used as input to the KDF.

[0401] Option #3: Read PUF RAW Mode (UDS RAW ), and UDS PUF Calculated as: UDS PUF =KDF(UDS RAW |UID|HASH(user-pattern)), where the host provides the user-pattern in the command payload by requesting self-identity.

[0402] Option #4: Read PUF RAW Mode (UDS RAW ), and UDS PUF Calculated as: UDS PUF =KDF(UDS RAW |UID|HASH(usermode)|freshness), where freshness is, for example, the freshness provided in the command payload.

[0403] More generally, UDS is calculated as UDS PUF=KDF[(information provided by the host device|information present in the device)] The KDF function may be used as a simple HASH function (eg, SHA256) using a secret key, or a MAC function (eg, HMAC-SHA256).

[0404] In one example, when the MAC function is used as a KDF, the secret key used is the same key used to provide the authentication command. The device secret (e.g., UDS) may be generated using one of the following options:

[0405] Option #5: UDS PUF =MAC[Secret_Key,(UDS RAW )]

[0406] Option #6: UDS PUF =MAC[Secret_Key,(UDS RAW |UID)]

[0407] Option #7: UDS PUF =MAC[Secret_Key,(UDS RAW |UID|HASH(user mode))]

[0408] Option #8: UDS PUF =MAC[Secret_Key,(UDS RAW |UID|HASH(user mode)|freshness)]

[0409] More generally, UDS PUF =MAC[Secret_Key,(information provided by the host|information existing in the device)]

[0410] As mentioned above, UDS PUF After being generated, it can be communicated to the host device 151. In one embodiment, the host device 151 can directly read the UDS PUF .UDS PUF Only a predetermined number of times (e.g., only once or several times) may be read. In one example, the following may be used for Fig.21 The process described. Reading UDS at a predetermined number of times PUF Thereafter, the read mechanism is permanently disabled.For example, this method may be used in a secure environment (eg, in a customer's factory when using computing device 141 to assemble a computing system or other product).

[0411] In one example, the UDS is encrypted by computing device 141 using a public key received from host device 151 PUF , and then UDS PUFis sent to the host device 151. After this procedure, the UDS is permanently disabled PUF Reading mechanism.

[0412] In another embodiment, the computing device 141 includes an obfuscation processing module 2011. The obfuscation processing module 2011 is Fig.15 The obfuscation processing module 630 or Fig.19 An instance of the obfuscation processing module 630.

[0413] In one embodiment, the UDS PUF The host public key is encrypted and communicated to the host device; the host device decrypts it using its corresponding secret key. The host public key is communicated to the computing device 141 during a specific communication setup phase. In one embodiment, the UDS can be completed by using the authentication command through a direct read operation. PUF The computing device 141 stores the UDS key in a tamper-proof area. In each case where the UDS key is also used for internal (computing device) use, an obfuscation mechanism is used to avoid information leakage (e.g., by avoiding any chance of a hacker guessing the stored UDS key, such as by a hacker analyzing the current or voltage waveform). For example, the obfuscation mechanism used may be as described above for Fig.15 , 18 , as described in 19.

[0414] In one embodiment, computing device 141 obfuscates the device secret before sending the encrypted device secret to host device 151 .

[0415] Fig.21 A system is shown that sends an initial value provided by a monotonic counter of the system for use in determining whether tampering with the system has occurred according to one embodiment. In one example, the system is Fig. 20 The computing device 141.

[0416] In some embodiments, the system uses a secret key for secure communications with other devices (e.g., host device 151). The secret key is generated and stored on the system (e.g., by key injection at the factory after initial assembly of the system board). The system's monotonic counter is used to provide an initial value.

[0417] In another embodiment, the secret key is a UDS used with the DICE-RIoT protocol. The secret key is generated in response to a command from a host device (eg, host device 151).

[0418] In one embodiment, the initial value is sent to an intended recipient of the physical system by electronic communication (e.g., the recipient will receive the physical system after the physical system is physically transported to the recipient's location). After receiving the system, the recipient reads the output value from the monotonic counter. The recipient (e.g., using the recipient's computing device or server that earlier received the initial value) compares the initial value to the output value to determine whether tampering with the system has occurred. In one example, tampering is an unauthorized attempt by an intruder to access a secret key of the system during its physical transport.

[0419] In one embodiment, the secret key may be generated, for example, using a real RNG (Random Number Generator) or PUF, or pre-injected into the system (memory) in a factory-like secure environment.

[0420] In one embodiment, the generated key is associated with an initial value of a monotonic counter. The initial value is used by a recipient of the system to determine whether an unauthorized attempt to access the stored key has been made. In one embodiment, the key injection process may use the output from a Physical Unclonable Function (PUF) to generate a key for each power-on cycle of a computing device storing the key.

[0421] More precisely, Fig.21 A system 351 is shown that sends an initial value provided by a monotonic counter 355 for use in determining whether tampering has occurred with the system 351, according to one embodiment. For example, it may be determined whether the system 351 has been tampered with by a hacker seeking unauthorized access to stored keys during physical transportation of the system 351. In one example, the system 351 is a system board including a non-volatile memory 306, a processor 304, a monotonic counter 355, and a power supply 318.

[0422] One or more keys 314 are generated under the control of the processor 304. The non-volatile memory 306 is used to store the generated keys. The non-volatile memory 306 is, for example, a non-volatile memory device (eg, a 3D cross point storage area).

[0423] The monotonic counter 355 is initialized to provide an initial value. This initial value is associated with the stored key 314. The initial value is sent by the processor 304 to another computing device via the external interface 312. For example, the initial value can be sent to a server of a receiver to which the system 351 will be shipped after manufacturing and key injection is completed.

[0424] When the system 351 is received by the receiver, the computing device of the receiver determines the initial value. For example, the computing device may store the initial value received when sending as described above in a memory. The computing device reads the output value from the monotonic counter 355. This output value is compared with the initial value to determine whether the system 351 has been tampered with.

[0425] In one embodiment, the computing device of the receiver determines the number of events that have occurred that incremented the monotonic counter 355. For example, the output value from the monotonic counter 355 may be configured to increment on each power-on operation of the system 351 (e.g., as detected by monitoring the power supply 318). The output value from the monotonic counter 355 may additionally and / or alternatively be configured to increment upon each attempt to perform a read access to the stored key 314.

[0426] The initial value received from the transmitter can be adjusted based on this number of known events, for example by tracking the number of known events that cause the monotonic counter to increment. The adjusted initial value is then compared to the output value read from the monotonic counter 355. If the values ​​match, then no tampering has occurred. If the values ​​do not match, then the computing device determines that tampering has been detected.

[0427] In response to determining that tampering has been detected, one or more functions of the system 351 may be disabled. In one embodiment, the processor 304 receives a communication from the receiver's computing device that includes an indication that tampering has been detected. In response to receiving the communication, the processor 304 disables at least one function of the system 351. In one example, the disabled function is read access to the stored key 314.

[0428] In one embodiment, the system 351 may be configured such that the counter value output from the monotonic counter 355 cannot exceed a predetermined maximum value. For example, as each counter value is read from the monotonic counter 355, a determination is made as to whether the counter value exceeds the predetermined maximum value. If the counter value exceeds the predetermined maximum value, read access to the stored key 314 may be permanently disabled (e.g., under the control of the processor 304).

[0429] In one embodiment, system 351 is embodied on a semiconductor die. In another embodiment, system 351 is formed on a system board. Applications are stored in system memory 353 and executed by processor 304. Execution of the applications occurs after power-on of system 351. For example, a receiver of system 351 may execute an application after determining that no tampering with the system has occurred.

[0430] In one embodiment, the key 314 is generated using output values ​​from one or more physically unclonable functions (PUFs). For example, the key is generated for each power-on cycle of the system 351. In another example, the key 314 is a UDS generated in response to a replacement command received via the external interface 312 from the host device.

[0431] In one embodiment, system 351 is a controller that stores key 314. External interface 312 is used to send the initial value from monotonic counter 355 to an external non-volatile memory device (not shown) on the same system board as the controller. The external non-volatile memory device determines whether tampering has occurred by reading the output value from monotonic counter 355 and comparing the output value to the initial value received from system 351.

[0432] In one embodiment, system memory 353 includes volatile memory 308 and / or non-volatile memory 306. Cryptographic module 316 is used to perform cryptographic operations for secure communications via external interface 312 using key 314 (eg, a symmetric key).

[0433] Establishing secret key sharing is now described in various further embodiments. A secure communication channel is set up through key sharing between actors participating in the communication. Components used in a Trusted Platform Module board typically do not have sufficient processing power to implement schemes such as public key cryptography.

[0434] In one embodiment, one or more secret keys are shared between the device / board manufacturer and the OEM / end customer. Also, keys can be shared between different devices in the same board in the field as needed. As discussed above, one or more monotonic counters are used inside the security device.

[0435] In one example, before a device leaves the factory or manufacturer's test line, one or more secret keys (e.g., one or more secret keys as follows: Secret_Key with k=1...N) are injected into the device depending on the device capabilities and user needs. k ).

[0436] Depending on the device capabilities and user needs, initialize one or more monotonic counters (N (N ≥ 1) different MTC → MTC k =0). The monotonic counter (MTC) is configured to increment the output value whenever a power-on of the system / device occurs, and to increment the output value whenever a stored secret key is (attempted) to be read.

[0437] Each MTC k The value of may be public and shared with clients (e.g., as an MTC kA command sequence (eg, as arbitrarily determined by a system designer) is used to read the key from the device, and the command sequence may be public (eg, the strength of the method is not based on the security of the read protocol).

[0438] In one embodiment, a dedicated MTC need not be used for each secret key. This may vary depending on the type of security service used. For example, some keys may be read together and only a single MTC is required.

[0439] In one embodiment, after the system arrives at the receiver (eg, client / user) location, several operations are performed. The receiver's computing device first determines the initial value of the monotonic counter 355. For example, retrieve the MTC k =(MTC0) k For each MTC or associated key, the MTC k The value can be different.

[0440] The receiver client / user then powers up the system 351. The first operation addressed is MTC k If each MTC k value and the expected (received) initial value [e.g. MTC k =(MTC0+1) k ] matches, then the device is determined to have not been powered on during physical shipping or other transfer, and the system 351 is considered authentic (and not tampered with) by the end user / customer.

[0441] If one or more values ​​do not match, tampering has occurred. For example, an unauthorized person powered on the device and attempted to read the stored key. If tampering has occurred, the device is discarded and an indication of the tampering is communicated to the sender (e.g., the delivery company) to avoid additional technical security issues.

[0442] In another embodiment, multiple read operations are performed by the receiver of system 351. For example, a device leaves the factory and arrives at a customer / OEM. The customer reads one or more keys and the associated MTC k Increment by one. If MTC j Incremental value beyond the silicon factory (transmitter) preset value (MTC j MAX ), then permanently deactivate the key (Secret_Key j ) is read.

[0443] In this example, MTC j MAXis a predetermined maximum value agreed by the sender and the customer for each MTC. This method allows one or more attempts to read the key after / before reading the MTC value to be checked or performed. This allows, for example, to discover any unauthorized access while also ensuring that the OEM / customer performs a read of the key several times before deactivation.

[0444] If during the read process one or more MTCs have unexpected values ​​(e.g., the initial value and the read value do not match after adjustment for a known number of read and power-on operations), then the device is discarded. In one embodiment, once all keys are read, the associated MTCs are cleared. k And it can be reused for other purposes (e.g., cryptographic processing functions as configured and desired by the system designer, etc.).

[0445] In another example, the monitoring device / board uses. This example uses a monotonic counter that already exists on the security device (e.g., exists for use by the cryptographic module 316). Before the device leaves the silicon factory, the following operations are performed on the test production line: Initialize M (M≥1) different MTC → MTC k =0.

[0446] Such MTC counters are incremented at each power-up of the device.Each counter may be initialized to zero (or to another initial value as desired).

[0447] The receiver end user can use the received MTC k Values ​​are used to obtain various information about the system 351, such as detecting unauthorized use of components / boards, determining that a component has been de-soldered and used outside the board to hack into the component, power cycle counting based on customer service implementation fee mechanism, device board warranty policy and power loss.

[0448] In one embodiment, implementing the fee mechanism involves associating a value with a specific usage. An application can be provided to a customer, and the customer must pay a fee to unlock a specific usage.

[0449] In one embodiment, the secret key is shared between different devices on the same system board (eg, shared in the field when used by an end user). The key is shared by exchanging between components on the same board (eg, between processors).

[0450] In one embodiment, each of the one or more monotonic counters (MTCk) is initialized by setting the initial value of the one or more monotonic counters (MTCk) to zero. In one example, when MTCk=0 is set, MTCk is a counter associated with a specific key indicated by the number k. MTCk=0 indicates that the initial value of the counter is zero. Each k indicates one of the numbers of counters, which corresponds to the number of internal keys stored in the device. The value of each counter is stored in a non-volatile memory area of ​​the device.

[0451] In one embodiment, the power-on (sometimes also referred to as power-up) of the device is detected. The device contains internal circuitry that measures the value of a power supply. When the power supply exceeds a certain threshold, the circuitry triggers an internal signal to indicate the presence (detection) of a power-on event. This signal may cause a monotonic counter to increment.

[0452] In one embodiment, an attempt to read a secret key is detected. A counter (MTCk) is incremented each time power is detected (as described above), and additionally each time the device interface recognizes a command sequence to read the key. Knowing the initial MTCk value permits the final receiver of the device (e.g., end user / customer) to know the device (and counter) state when shipping of the device occurs. Thus, if during shipping there is an attempt to power and / or read the device, this results in a change in the counter value stored in each MTCk.

[0453] In one embodiment, a command sequence is used to read a key from a device. For example, the device has an external interface that accepts commands from other computing components or devices. The key is available for reading until a maximum counter value is reached. In one example, the device has a sequence at the interface: command (e.g., 0x9b) + variable parameters (0x34) + signature. The device understands that the key will be read and provides the key at the output.

[0454] In one embodiment, after testing of the device (e.g., at the initial factory), the MTCk of each key will have an initial value from initialization. So, for example, if k=3, then: MTC0=20d, MTC1=25d, MTC2=10. The values ​​MTC0, MTC1, MTC2 are sent / transmitted to the client as a set of initial values. When the device is received by the client, the client uses a command sequence to read the values. The client (e.g., using a computing device) then determines whether the device was compromised during shipping. In one example, the monotonic counter value is read using the command sequence described above.

[0455] In one embodiment, a predetermined maximum counter value is preset. For example, the preset maximum value may be selected based on a customer procedure. For example, assume that a customer wants to read a key 10 times. Also, assume that MTCk is incremented only when power is on, and assume that there are two counters MTC0 (associated only with power on) and MTC1 (associated only with a key read command procedure). After factory testing, based on monitoring the MTC0 and MTC1 values, it is found that: MTC0=30, MTC1=25. In one example, an internal threshold is set to 40 for MTC0 and 35 for MTC1.

[0456] In one embodiment, the MTC associated with the key is incremented with power-up and attempts to read the key. The MTC is incremented whenever one of two events occurs.

[0457] In one embodiment, the monotonic counter is cleared (and optionally re-used for other processing of the device) after the final read of the stored key. The system designer can configure this as desired. The final read is for example when the purpose of the MTC count is no longer needed as a detector of malicious events. This frees up the counter as a resource, which can be used for other purposes, or the counter can keep counting.

[0458] In one embodiment, the monotonic counter can be used to obtain various types of information, such as desoldered components, enforce fee mechanisms, enforce warranties, etc. The counter can be used to monitor different types of occurrences because incrementing the counter value can be externally triggered.

[0459] In one embodiment, multiple key read options are provided. In one example, for initialization of the MTC value (for the MTCk associated with the key), a maximum threshold is set to allow each key to be read by the component and / or the end user's receiver. Multiple read options allow the threshold to be changed (which may be different for each key) based on the maximum number of attempts to read a particular key.

[0460] In one embodiment, different types of devices can use the above method. For example, a CPU or MCU typically has an internal hardware security module that is not accessible from the outside. In one example, it is necessary to have the same key stored in the device or component (for example, when using a symmetric key method) in order to operate correctly. In this case, the CPU / MCU benefits from the distribution / sharing of the key to authorized entities, devices or components. This sharing allows the key to be read (for example, this sharing corresponds to the programming of the value in the MTCk threshold).

[0461] In one embodiment, the above method is used for customer firmware injection. For example, the above method allows storing critical content internally in a device (e.g., application firmware) and implementing movement of the device in an unsecured environment (without compromising the integrity of the firmware / device). In one example, for firmware injection, unexpected changes in the MTCk counter value during transportation are used as an index indicating that firmware compromise has occurred.

[0462] Fig. 22 A method for generating an identity of a computing device using a physically unclonable function (PUF) according to one embodiment is shown. For example, Fig. 22 The method can be found in Fig. 20 implemented in the system.

[0463] Fig. 22 The method may be performed by processing logic, which may include hardware (e.g., a processing device, a circuit, dedicated logic, programmable logic, microcode, hardware of a device, an integrated circuit, etc.), software (e.g., instructions running or executed on a processing device), or a combination thereof. In some embodiments, at least in part by Fig. 20 The processor 143 executes Fig. 22 method.

[0464] Although shown in a particular order or sequence, the order of the processes may be modified unless otherwise specified. Therefore, it should be understood that the illustrated embodiments are merely examples, and the illustrated processes may be performed in a different order, and some processes may be performed in parallel. In addition, one or more processes may be omitted in various embodiments. Therefore, not all processes are required in every embodiment. Other process flows are possible.

[0465] At block 2211, at least one value is provided by at least one physically unclonable function (PUF). For example, PUF 2005 provides a value as an output.

[0466] At block 2213, a device secret is generated using a key derivation function (KDF). At least one value provided by at least one PUF is used as an input to the KDF. For example, the output from PUF 2005 is used as an input to KDF 2007.

[0467] At block 2215, the generated device secret is stored. For example, KDF 2007 generates an output that is stored as device secret 149.

[0468] In one embodiment, the method includes: generating a device secret (e.g., a UDS stored as device secret 149) by a computing device, the generation including: providing at least one value by at least one physically unclonable function (PUF); and generating the device secret using a key derivation function (e.g., KDF 2007), wherein at least one value provided by at least one PUF is an input to the KDF; and storing the generated device secret in a memory of the computing device.

[0469] In one embodiment, storing the generated device secret includes replacing a device secret previously stored in memory with the generated device secret.

[0470] In one embodiment, the KDF is a hash or a message authentication code.

[0471] In one embodiment, the method further includes storing a secret key for communicating with the host device, wherein the KDF is a message authentication code (MAC), at least one value provided by the at least one PUF is a first input to the MAC, and the secret key is a second input to the MAC.

[0472] In one embodiment, the device secret is generated in response to an event, and the event is receipt of a command by the computing device from the host device.

[0473] In one embodiment, the method further includes receiving a host public key from the host device, encrypting the generated device secret using the host public key, and sending the encrypted device secret to the host device.

[0474] In one embodiment, the method further comprises permanently disabling read access to the device secret in the memory after sending the encrypted device secret to the host device.

[0475] In one embodiment, the method further includes obfuscating the device secret prior to sending the encrypted device secret to the host device.

[0476] In one embodiment, the method further includes storing a secret key (e.g., secret key 2013) in a memory by the computing device, wherein the command is authenticated by the computing device using a message authentication code (MAC) and the secret key is used as an input to the MAC.

[0477] In one embodiment, the host device sends a signature for authenticating the command, the signature is generated using a MAC, and the freshness generated by the host device is used as an input to the MAC.

[0478] In one embodiment, freshness is an additional input to the KDF.

[0479] In one embodiment, the host device provides commands to user mode, and wherein a hash of the user mode is an additional input to the KDF.

[0480] In one embodiment, the method further comprises authenticating, by the computing device, the command prior to generating the device secret.

[0481] In one embodiment, the method further includes storing a unique identifier of the computing device, wherein the unique identifier is an additional input to the KDF.

[0482] In one embodiment, the device secret is generated in response to an event, and the event is the detection of usage of the computing system by the computing device.

[0483] In one embodiment, the use is the execution of an application by a computing system. In one embodiment, the use is the initialization of a boot loading process.

[0484] In one embodiment, according to the method of technical solution 1, the device secret is generated in response to an event, and the event is a time-scheduled event.

[0485] In one embodiment, a system includes: at least one processor; and

[0486] A memory containing instructions configured to direct the at least one processor to: generate a device secret, the generation comprising: providing at least one value by at least one physically unclonable function (PUF); and generating the device secret using a key derivation function (KDF), wherein the at least one value provided by the at least one PUF is an input to the KDF; and storing the generated device secret in a memory of the computing device.

[0487] In one embodiment, the instructions are further configured to direct the at least one processor to: receive a replace command from a host device; and send a public identifier generated using a generated device secret to the host device; wherein the device secret is generated in response to receiving the replace command; wherein storing the generated device secret includes replacing a previously stored device secret with the generated device secret.

[0488] In one embodiment, a non-transitory computer storage medium stores instructions that, when executed on a computing device, cause the computing device to at least perform the following operations: provide at least one value through at least one physically unclonable function (PUF); generate a device secret using a key derivation function (KDF), wherein at least one value provided by the at least one PUF is an input to the KDF; and store the generated device secret in a memory.

[0489] Conclusion

[0490] Non-transitory computer storage media may be used to store instructions for firmware 104, or for storing instructions for processor 143 or processing device 111. When the instructions are executed by controller 107, such as memory device 103 or computing device 603, the instructions cause controller 107 to perform any of the methods discussed above.

[0491] In this specification, various functions and operations may be described as being performed by computer instructions or caused by computer instructions to simplify the description. However, those skilled in the art will recognize that the intention of such expressions is that the functions are derived from one or more controllers or processors (such as microprocessors) executing computer instructions. Alternatively or in combination, the functions and operations can be implemented using dedicated circuits with or without software instructions, such as using application specific integrated circuits (ASICs) or field programmable gate arrays (FPGAs). Embodiments can be implemented using hard-wired circuit systems without software instructions or in combination with software instructions. Therefore, the technology is not limited to any particular combination of hardware circuits and software, nor is it limited to any particular source of instructions executed by a data processing system.

[0492] While some embodiments may be implemented on fully functional computers and computer systems, various embodiments are capable of being distributed in various forms of computing products and are applicable regardless of the particular type of machine or computer readable media actually used to achieve the distribution.

[0493] At least some aspects disclosed may be embodied at least in part in software. That is, the techniques may be implemented in a computer system or other data processing system in response to its processor (e.g., a microprocessor or microcontroller) executing sequences of instructions contained in a memory (e.g., ROM, volatile RAM, non-volatile memory, cache, or remote storage device).

[0494] The routines executed to implement the embodiments may be implemented as part of an operating system or a specific application, component, program, object, module, or sequence of instructions referred to as a “computer program.” A computer program typically includes one or more sets of instructions in various memories and storage devices in a computer at various times, and when the instruction sets are read and executed by one or more processors in a computer, causes the computer to perform the operations necessary to perform the elements involved in the various aspects.

[0495] Tangible non-transitory computer storage media can be used to store software and data, which, when executed by a data processing system, causes the system to perform various methods. Executable software and data can be stored in various locations including, for example, ROM, volatile RAM, non-volatile memory, and / or cache. Portions of this software and / or data can be stored in any of these storage devices. In addition, data and instructions can be obtained from a centralized server or a peer-to-peer network. Different portions of data and instructions can be obtained from different centralized servers and / or peer-to-peer networks at different times and in different communication sessions or the same communication session. All data and instructions can be obtained before executing the application. Alternatively, portions of data and instructions can be obtained dynamically and timely when execution is required. Therefore, it is not required that data and instructions are all on machine-readable media at a specific time.

[0496] Examples of computer-readable storage media include, but are not limited to, recordable and non-recordable types of media, such as volatile and nonvolatile memory devices, read-only memory (ROM), random access memory (RAM), flash memory devices, floppy and other removable disks, magnetic disk storage media, and optical storage media (e.g., compact disk read-only memory (CDROM), digital versatile disk (DVD), etc.), and other media. Instructions may be embodied in transitory media, such as electrical, optical, acoustic or other forms of propagated signals, such as carrier waves, infrared signals, digital signals, etc. Transitory media are generally used to transmit instructions, but are not considered capable of storing the instructions.

[0497] In various embodiments, hard-wired circuitry may be used in combination with software instructions to implement the technology.Thus, the technology is neither limited to any specific combination of hardware circuitry and software nor to any specific source for the instructions executed by the data processing system.

[0498] Although some of the figures illustrate several operations in a particular order, operations that are not dependent on order may be reordered, and other operations may be combined or decomposed. Although some reorderings or other groupings are specifically mentioned, other reorderings or groupings are obvious to those skilled in the art, so an exhaustive list of alternatives is not provided. In addition, it should be recognized that the stages may be implemented in hardware, firmware, software, or any combination thereof.

[0499] The above description and drawings are illustrative and should not be construed as limiting. Many specific details are described to provide a thorough understanding. However, in some cases, well-known or conventional details are not described to avoid obscuring the description. References to one or an embodiment in this disclosure do not necessarily refer to the same embodiment; and such references mean at least one.

[0500] In the foregoing description, the present disclosure has been described with reference to specific exemplary embodiments of the present disclosure. It will be apparent that various modifications may be made thereto without departing from the broader spirit and scope set forth in the appended claims. Therefore, the description and drawings should be regarded as illustrative rather than restrictive.

Claims

1. A method performed by a computing device, the method comprising: Generating, by a processor of the computing device, a device secret, the generating comprising: providing at least one value via at least one physically unclonable function (PUF); and generating the device secret using a key derivation function (KDF), wherein the at least one value provided by the at least one PUF is an input to the KDF; storing the generated device secret in a memory of the computing device; and Read access to the device secret stored in the memory is disabled in response to determining that a predetermined number of read operations have been performed. 2 . The method of claim 1 , wherein storing the generated device secret comprises replacing a device secret previously stored in the memory with the generated device secret.

3. The method of claim 1, wherein the KDF is a hash or a message authentication code.

4. The method of claim 1 , further comprising storing a secret key for communicating with a host device, wherein the KDF is a message authentication code (MAC), the at least one value provided by the at least one PUF is a first input to the MAC, and the secret key is a second input to the MAC.

5. The method of claim 1, wherein the device secret is generated in response to an event, and the event is receiving a command from a host device by the computing device.

6. The method of claim 5, further comprising receiving a host public key from the host device, encrypting the generated device secret using the host public key, and sending the encrypted device secret to the host device.

7. The method of claim 6, further comprising permanently disabling read access to the device secret in the memory after sending the encrypted device secret to the host device.

8. The method of claim 6, further comprising obfuscating the device secret when reading the device secret from the memory of the computing device.

9. The method of claim 5, further comprising storing, by the computing device, a secret key in a memory, wherein the command is authenticated by the computing device using a message authentication code (MAC) and the secret key is used as an input to the MAC.

10. The method of claim 9, wherein the host device sends a signature for authenticating the command, the signature generated using the MAC, and a freshness generated by the host device is used as an input to the MAC. The method of claim 10 , wherein the freshness is an additional input to the KDF.

12. The method of claim 5, wherein the host device provides the command to a user mode, and wherein a hash of the user mode is an additional input to the KDF.

13. The method of claim 5, further comprising authenticating the command by the computing device prior to generating the device secret.

14. The method of claim 1, further comprising storing a unique identifier for the computing device, wherein the unique identifier is an additional input to the KDF.

15. The method of claim 1, wherein the device secret is generated in response to an event, and the event is detection of usage of a computing system by the computing device. The method of claim 15 , wherein the using is executing an application program via the computing system.

17. The method of claim 1, wherein the device secret is generated in response to an event, and the event is a time-scheduled event.

18. A computing device comprising: at least one processor; and A memory containing instructions configured to direct the at least one processor to: Generating a device secret, the generating comprising: providing at least one value via at least one physically unclonable function (PUF); and generating the device secret using a key derivation function (KDF), wherein the at least one value provided by the at least one PUF is an input to the KDF; storing the generated device secret in the memory of the computing device; and Read access to the device secret stored in the memory is disabled in response to determining that a predetermined number of read operations have been performed.

19. The computing device of claim 18, wherein the instructions are further configured to direct the at least one processor to: receiving a replace command from a host device; and sending to said host device a public identifier secretly generated using said generating device; wherein said device secret is generated in response to receiving said replace command; Wherein storing the generated device secret comprises replacing a previously stored device secret with the generated device secret.

20. The computing device of claim 18, further comprising an identity component configured to generate an identifier, a certificate, and a key, the identifier being secretly associated with the device and unique to the computing device, and the certificate being generated using a message received from a host device, wherein the identifier, the certificate, and the key are sent to the host device via a host interface for authenticating the computing device.

21. The computing device of claim 20, wherein the identity component comprises an asymmetric identity generator that outputs a public identifier and a private identifier based on the device secret, the public identifier being the identifier sent to the host device.

22. The computing device of claim 21, wherein the identity component comprises an asymmetric key generator that uses a random number as input and outputs a public key and a private key, the public key being the key sent to the host device.

23. A computing device according to claim 22, wherein the identity component includes a first encryptor and a second encryptor, wherein the first encryptor uses the private identifier, the public key and the message from the host device as input and outputs a result key, and the second encryptor uses the private key and the result key as input and outputs the certificate sent to the host device.

24. A non-transitory computer storage medium storing instructions that, when executed on a computing device, cause the computing device to at least: providing at least one value via at least one physically unclonable function (PUF); generating a device secret using a key derivation function (KDF), wherein the at least one value provided by the at least one PUF is an input to the KDF; storing the generated device secret in a memory of the computing device; and Read access to the device secret stored in the memory is disabled in response to determining that a predetermined number of read operations have been performed.

Images