Remote Attestation Procedures Architecture

3.1. Network Endpoint Assessment

Network operators want a trustworthy report of identity and version of information of the hardware and software on the machines attached to their network, for purposes such as inventory, auditing, and/or logging. The network operator may also want a policy by which full access is only granted to devices that meet some definition of health, and so wants to get claims about such information and verify their validity. Remote attestation is desired to prevent vulnerable or compromised devices from getting access to the network and potentially harming others.¶

Typically, solutions start with a specific component (called a "Root of Trust") that provides device identity and protected storage for measurements. These components perform a series of measurements, and express this with Evidence as to the hardware and firmware/software that is running.¶

    FIXME from Henk: Measurements at early stages of
    Layered Attestation are NOT evidence yet.
    This text does not cover that yet

Attester:

A device desiring access to a network¶

Relying Party:

A network infrastructure device such as a router, switch, or access point¶

3.2. Confidential Machine Learning (ML) Model Protection

A device manufacturer wants to protect its intellectual property in terms of the ML model it developed and that runs in the devices that its customers purchased, and it wants to prevent attackers, potentially including the customer themselves, from seeing the details of the model.¶

This typically works by having some protected environment in the device attest to some manufacturer service. If remote attestation succeeds, then the manufacturer service releases either the model, or a key to decrypt a model the Attester already has in encrypted form, to the requester.¶

Attester:

A device desiring to run an ML model to do inferencing¶

Relying Party:

A server or service holding ML models it desires to protect¶

3.3. Confidential Data Retrieval

This is a generalization of the ML model use case above, where the data can be any highly confidential data, such as health data about customers, payroll data about employees, future business plans, etc. Attestation is desired to prevent leaking data to compromised devices.¶

Attester:

An entity desiring to retrieve confidential data¶

Relying Party:

An entity that holds confidential data for retrieval by other entities¶

3.4. Critical Infrastructure Control

In this use case, potentially dangerous physical equipment (e.g., power grid, traffic control, hazardous chemical processing, etc.) is connected to a network. The organization managing such infrastructure needs to ensure that only authorized code and users can control such processes, and they are protected from malware or other adversaries. When a protocol operation can affect some critical system, the device attached to the critical equipment thus wants some assurance that the requester has not been compromised. As such, remote attestation can be used to only accept commands from requesters that are within policy.¶

Attester:

A device or application wishing to control physical equipment¶

Relying Party:

A device or application connected to potentially dangerous physical equipment (hazardous chemical processing, traffic control, power grid, etc.)¶

3.5. Trusted Execution Environment (TEE) Provisioning

A "Trusted Application Manager (TAM)" server is responsible for managing the applications running in the TEE of a client device. To do this, the TAM wants to assess the state of a TEE, or of applications in the TEE, of a client device. The TEE attests to the TAM, which can then decide whether the TEE is already in compliance with the TAM's latest policy, or if the TAM needs to uninstall, update, or install approved applications in the TEE to bring it back into compliance with the TAM's policy.¶

Attester:

A device with a trusted execution environment capable of running trusted applications that can be updated¶

Relying Party:

A Trusted Application Manager¶

3.6. Hardware Watchdog

One significant problem is malware that holds a device hostage and does not allow it to reboot to prevent updates to be applied. This is a significant problem, because it allows a fleet of devices to be held hostage for ransom.¶

A hardware watchdog can be implemented by forcing a reboot unless remote attestation to a server succeeds within a periodic interval, and having the reboot do remediation by bringing a device into compliance, including installation of patches as needed.¶

Attester:

The device that is desired to keep from being held hostage for a long period of time¶

Relying Party:

A remote server that will securely grant the Attester permission to continue operating (i.e., not reboot) for a period of time¶

4.1. Two Types of Environments of an Attester

An Attester consists of at least one Attesting Environment and at least one Target Environment. In some implementations, the Attesting and Target Environments might be combined. Other implementations might have multiple Attesting and Target Environments, such as in the examples described in more detail in Section 4.2 and Section 4.3. Other examples may exist, and the examples discussed could even be combined into even more complex implementations.¶

Claims are collected from Target Environments. That is, Attesting Environments collect the raw values and the information to be represented in claims, such as by doing some measurement of a Target Environment's code, memory, and/or registers. Attesting Environments then format the claims appropriately, and typically use key material and cryptographic functions, such as signing or cipher algorithms, to create Evidence. Examples of environments that can be used as Attesting Environments include Trusted Execution Environments (TEE), embedded Secure Elements (eSE), or Hardware Security Modules (HSM).¶

4.2. Layered Attestation Procedures

By definition, the Attester role takes on the duty to create Evidence. The fact that an Attester role is composed of several types of environments that can be nested or staged adds complexity to the architectural layout of how an Attester - in itself - is composed and therefore has to conduct the Claims collection in order to create believable Attestation Evidence. The following example is intended to illustrate this composition:¶

A very common example is elaborated on to illustrate Layered Attestation.¶

    .----------.                    .----------.
    |          |                    |          |
    | Endorser |------------------->| Verifier |
    |          |    Endorsements    |          |
    '----------'  for A, B, and C   '----------'
                                          ^
.------------------------------------.    |
|                                    |    |
|   .---------------------------.    |    |
|   | Target                    |    |    | Layered
|   | Environment               |    |    | Evidence
|   | C                         |    |    |   for
|   '---------------------------'    |    | B and C
|           Collect |                |    |
|           claims  |                |    |
|   .---------------|-----------.    |    |
|   | Target        v           |    |    |
|   | Environment .-----------. |    |    |
|   | B           | Attesting | |    |    |
|   |             |Environment|-----------'
|   |             |     B     | |    |
|   |             '-----------' |    |
|   |                     ^     |    |
|   '---------------------|-----'    |
|           Collect |     | Evidence |
|           claims  v     |  for B   |
|                 .-----------.      |
|                 | Attesting |      |
|                 |Environment|      |
|                 |     A     |      |
|                 '-----------'      |
|                                    |
'------------------------------------'

The very first Attesting Environment has to ensure the integrity of the (U)EFI / BIOS / Firmware that initially boots up a composite device (e.g., a cell phone).¶

    Henk: we are looking for a better term than UEFI/BIOS/Firmware

These Claims have to be measured securely. At this stage of the boot-cycle of a composite device, the Claims collected typically cannot be composed into Evidence.¶

The very first Attesting Environment in this example can be a hardware component that is a Static Code Root of Trust. As in any other scenario, this hardware component is the first Attesting Environment. It collects a rather concise number of Claims about the Target Environment. The Target Environment in this example is the (U)EFI / BIOS / Firmware After the boot sequence started, the Target Environment conducts the most important and defining feature of Layered Attestation: The successfully measured environment that is the (U)EFI / BIOS / Firmware now becomes the Attesting Environment. Analogously, the Attesting Environment hands off its duty to one of its Target Environments. This procedure in Layered Attestation is called Staging.¶

Now, the duties have been transferred and Layered Attestation takes place. The initial Attesting Environment relinquishes its duties to the Target Environment. It is important to note that the new Attesting Environment cannot alter the content about its own measurements. If the Attesting Environment would be able to do that, Layered Attestation would become unfeasible.¶

In this example the duty of being the Attesting Environment is now taken over by the (U)EFI / BIOS / Firmware that was the Attested Environment before. This transfer of duty is the essential part of Layered Attestation. The (U)EFI / BIOS / Firmware now is the Attesting Environment. The next Target Environment is, in this example, a bootloader. There are potentially multiple kernels to boot, the decision is up to the bootloader. Only a bootloader with intact integrity will make an appropriate decision. Therefore, Claims about the integrity of a bootloader are now collected by the freshly appointed Attesting Environment that is the (U)EFI / BIOS / Firmware. Collected Claims have to be stored by the current Attesting Environment in a similar shielded and secured manner, so that the next Attesting Environment is not capable of altering the collection of claims stored.¶

Continuing with this example, the bootloader is now in charge of collecting Claims about the next execution environment. The next execution environment in this example is the kernel to be booted up. Analogously, the next transfer of duties in this Layered Attestation example occurs: The duty of being an Attesting Environment is transferred to a successfully measured kernel. In this sequence, the kernel is now collecting additional Claims and is storing them in a secure and shielded manner.¶

    [Henk: we might have to define what successful
    means in this example and beyond]

The essence of this example is a cascade of staged boot environments. Each environment (after the initial one that is a root-of-trust) has the duty of measuring its next environment before it is started. Therefore, creating a layered boot sequence and correspondingly enabling Layered Attestation.¶

4.3. Composite Device

A Composite Device is an entity composed of multiple sub-entities such that its trustworthiness has to be determined by the appraisal of all these sub-entities.¶

Each sub-entity has at least one Attesting Environment collecting the claims from at least one Target Environment, then this sub-entity generates Evidence about its trustworthiness. Therefore each sub-entity can be called an Attester. Among all the Attesters, there may be only some which have the ability to communicate with the Verifier while others do not.¶

For example, a carrier-grade router is consists of a chassis and multiple slots. The trustworthiness of the router depends on all its slots' trustworthiness. Each slot has an Attesting Environment such as a TEE collecting the claims of its boot process, after which it generates Evidence from the claims. Among these slots, only a main slot can communicate with the Verifier while other slots cannot. But other slots can communicate with the main slot by the links between them inside the router. So the main slot collects the Evidence of other slots, produces the final Evidence of the whole router and conveys the final Evidence to the Verifier. Therefore the router is a Composite Device, each slot is an Attester, and the main slot is the transiting Attester.¶

Another example is a multi-chassis router composed of multiple single carrier-grade routers. The multi-chassis router provides higher throughput by interconnecting multiple routers and can be logically treated as one router for simpler management. Among these routers, there is only one main router that connects to the Verifier. Other routers are only connected to the main router by the network cables, and therefore they are managed and appraised via this main router's help. So, in this case, the multi-chassis router is the Composite Device, each router is an Attester and the main router is the lead Attester.¶

Figure 3 depicts the conceptual data flow for a Composite Device.¶

                   .-----------------------------.
                   |           Verifier          |
                   '-----------------------------'
                                   ^
                                   |
                                   | Evidence of
                                   | Composite Device
                                   |
.----------------------------------|-------------------------------.
| .--------------------------------|-----.      .------------.     |
| |  Collect             .------------.  |      |            |     |
| |  Claims   .--------->|  Attesting |<--------| Attester B |-.   |
| |           |          |Environment |  |      '------------. |   |
| |  .----------------.  |            |<----------| Attester C |-. |
| |  |     Target     |  |            |  |        '------------' | |
| |  | Environment(s) |  |            |<------------| ...        | |
| |  |                |  '------------'  | Evidence '------------' |
| |  '----------------'                  |    of                   |
| |                                      | Attesters               |
| | lead Attester A                      | (via Internal Links or  |
| '--------------------------------------' Network Connections)    |
|                                                                  |
|                       Composite Device                           |
'------------------------------------------------------------------'

In the Composite Device, each Attester generates its own Evidence by its Attesting Environment(s) collecting the claims from its Target Environment(s). The lead Attester collects the Evidence of all other Attesters and then generates the Evidence of the whole Composite Attester.¶

An entity can take on multiple RATS roles (e.g., Attester, Verifier, Relying Party, etc.) at the same time. The combination of roles can be arbitrary. For example, in this Composite Device scenario, the entity inside the lead Attester can also take on the role of a Verifier, and the outside entity of Verifier can take on the role of a Relying Party. After collecting the Evidence of other Attesters, this inside Verifier verifies them using Endorsements and Appraisal Policies (obtained the same way as any other Verifier), to generate Attestation Results. The inside Verifier then sends the Attestation Results of other Attesters, whether in the same conveyance protocol as the Evidence or not, to the outside Verifier.¶

In this situation, the trust model described in Section 6 is also suitable for this inside Verifier.¶

5.1. Passport Model

In this model, an Attester sends Evidence to a Verifier, which compares the Evidence against its Appraisal Policy. The Verifier then gives back an Attestation Result. If the Attestation Result was a successful one, the Attester can then present the Attestation Result to a Relying Party, which then compares the Attestation Result against its own Appraisal Policy.¶

There are three ways in which the process may fail. First, the Verifier may refuse to issue the Attestation Result due to some error in processing, or some missing input to the Verifier. The second way in which the process may fail is when the resulting Result is examined by the Relying Party, and based upon the Appraisal Policy, the result does not pass the policy. The third way is when the Verifier is unreachable.¶

Since the resource access protocol between the Attester and Relying Party includes an Attestation Result, in this model the details of that protocol constrain the serialization format of the Attestation Result. The format of the Evidence on the other hand is only constrained by the Attester-Verifier remote attestation protocol.¶

      +-------------+
      |             | Compare Evidence
      |   Verifier  | against Appraisal Policy
      |             |
      +-------------+
           ^    |
   Evidence|    |Attestation
           |    |  Result
           |    v
      +----------+              +---------+
      |          |------------->|         |Compare Attestation
      | Attester | Attestation  | Relying | Result against
      |          |    Result    |  Party  | Appraisal
      +----------+              +---------+  Policy

The passport model is so named because of its resemblance to how nations issue passports to their citizens. The nature of the Evidence that an individual needs to provide to its local authority is specific to the country involved. The citizen retains control of the resulting passport document and presents it to other entities when it needs to assert a citizenship or identity claim, such as an airport immigration desk. The passport is considered sufficient because it vouches for the citizenship and identity claims, and it is issued by a trusted authority. Thus, in this immigration desk analogy, the passport issuing agency is a Verifier, the passport is an Attestation Result, and the immigration desk is a Relying Party.¶

5.2. Background-Check Model

In this model, an Attester sends Evidence to a Relying Party, which simply passes it on to a Verifier. The Verifier then compares the Evidence against its Appraisal Policy, and returns an Attestation Result to the Relying Party. The Relying Party then compares the Attestation Result against its own appraisal policy.¶

The resource access protocol between the Attester and Relying Party includes Evidence rather than an Attestation Result, but that Evidence is not processed by the Relying Party. Since the Evidence is merely forwarded on to a trusted Verifier, any serialization format can be used for Evidence because the Relying Party does not need a parser for it. The only requirement is that the Evidence can be encapsulated in the format required by the resource access protocol between the Attester and Relying Party.¶

However, like in the Passport model, an Attestation Result is still consumed by the Relying Party and so the serialization format of the Attestation Result is still important. If the Relying Party is a constrained node whose purpose is to serve a given type resource using a standard resource access protocol, it already needs the parser(s) required by that existing protocol. Hence, the ability to let the Relying Party obtain an Attestation Result in the same serialization format allows minimizing the code footprint and attack surface area of the Relying Party, especially if the Relying Party is a constrained node.¶

                               +-------------+
                               |             | Compare Evidence
                               |   Verifier  | against Appraisal
                               |             | Policy
                               +-------------+
                                    ^    |
                            Evidence|    |Attestation
                                    |    |  Result
                                    |    v
   +------------+               +-------------+
   |            |-------------->|             | Compare Attestation
   |   Attester |   Evidence    |   Relying   | Result against
   |            |               |    Party    | Appraisal Policy
   +------------+               +-------------+

The background-check model is so named because of the resemblance of how employers and volunteer organizations perform background checks. When a prospective employee provides claims about education or previous experience, the employer will contact the respective institutions or former employers to validate the claim. Volunteer organizations often perform police background checks on volunteers in order to determine the volunteer's trustworthiness. Thus, in this analogy, a prospective volunteer is an Attester, the organization is the Relying Party, and a former employer or government agency that issues a report is a Verifier.¶

5.3. Combinations

One variation of the background-check model is where the Relying Party and the Verifier on the same machine, and so there is no need for a protocol between the two.¶

It is also worth pointing out that the choice of model is generally up to the Relying Party, and the same device may need to create Evidence for different Relying Parties and different use cases (e.g., a network infrastructure device to gain access to the network, and then a server holding confidential data to get access to that data). As such, both models may simultaneously be in use by the same device.¶

Figure 6 shows another example of a combination where Relying Party 1 uses the passport model, whereas Relying Party 2 uses an extension of the background-check model. Specifically, in addition to the basic functionality shown in Figure 5, Relying Party 2 actually provides the Attestation Result back to the Attester, allowing the Attester to use it with other Relying Parties. This is the model that the Trusted Application Manager plans to support in the TEEP architecture [I-D.ietf-teep-architecture].¶

      +-------------+
      |             | Compare Evidence
      |   Verifier  | against Appraisal Policy
      |             |
      +-------------+
           ^    |
   Evidence|    |Attestation
           |    |  Result
           |    v
      +-------------+
      |             | Compare
      |   Relying   | Attestation Result
      |   Party 2   | against Appraisal Policy
      +-------------+
           ^    |
   Evidence|    |Attestation
           |    |  Result
           |    v
      +----------+               +----------+
      |          |-------------->|          | Compare Attestation
      | Attester |  Attestation  |  Relying | Result against
      |          |     Result    |  Party 1 | Appraisal Policy
      +----------+               +----------+

7.1. Evidence

Today, Evidence tends to be highly device-specific, since the information in the Evidence often includes vendor-specific information that is necessary to fully describe the manufacturer and model of the device including its security properties, the health of the device, and the level of confidence in the correctness of the information. Evidence is typically signed by the device (whether by hardware, firmware, or software on the device), and its appraisal in isolation would require Appraisal Policy to be based on device-specific details (e.g., a device public key).¶

7.2. Endorsements

An Endorsement is a secure statement that some entity (e.g., a manufacturer) vouches for the integrity of the device's signing capability. For example, if the signing capability is in hardware, then an Endorsement might be a manufacturer certificate that signs a public key whose corresponding private key is only known inside the device's hardware. Thus, when Evidence and such an Endorsement are used together, an appraisal procedure can be conducted based on Appraisal Policies that may not be specific to the device instance, but merely specific to the manufacturer providing the Endorsement. For example, an Appraisal Policy might simply check that devices from a given manufacturer have information matching a set of known-good reference values, or an Appraisal Policy might have a set of more complex logic on how to appraise the validity of information.¶

However, while an Appraisal Policy that treats all devices from a given manufacturer the same may be appropriate for some use cases, it would be inappropriate to use such an Appraisal Policy as the sole means of authorization for use cases that wish to constrain which compliant devices are considered authorized for some purpose. For example, an enterprise using remote attestation for Network Endpoint Assessment may not wish to let every healthy laptop from the same manufacturer onto the network, but instead only want to let devices that it legally owns onto the network. Thus, an Endorsement may be helpful information in authenticating information about a device, but is not necessarily sufficient to authorize access to resources which may need device-specific information such as a public key for the device or component or user on the device.¶

7.3. Attestation Results

Attestation Results may indicate compliance or non-compliance with a Verifier's Appraisal Policy. A result that indicates non-compliance can be used by an Attester (in the passport model) or a Relying Party (in the background-check model) to indicate that the Attester should not be treated as authorized and may be in need of remediation. In some cases, it may even indicate that the Evidence itself cannot be authenticated as being correct.¶

An Attestation Result that indicates compliance can be used by a Relying Party to make authorization decisions based on the Relying Party's Appraisal Policy. The simplest such policy might be to simply authorize any party supplying a compliant Attestation Result signed by a trusted Verifier. A more complex policy might also entail comparing information provided in the result against known-good reference values, or applying more complex logic on such information.¶

Thus, Attestation Results often need to include detailed information about the Attester, for use by Relying Parties, much like physical passports and drivers licenses include personal information such as name and date of birth. Unlike Evidence, which is often very device- and vendor-specific, Attestation Results can be vendor-neutral if the Verifier has a way to generate vendor-agnostic information based on the appraisal of vendor-specific information in Evidence. This allows a Relying Party's Appraisal Policy to be simpler, potentially based on standard ways of expressing the information, while still allowing interoperability with heterogeneous devices.¶

Finally, whereas Evidence is signed by the device (or indirectly by a manufacturer, if Endorsements are used), Attestation Results are signed by a Verifier, allowing a Relying Party to only need a trust relationship with one entity, rather than a larger set of entities, for purposes of its Appraisal Policy.¶