Microsoft Technology Licensing, LLC

Sebastian Lange

Gregory Darrell Fee

Adam G Poulos

Victor Tan

Christopher Wellington Brumme

Brian Irl Pratt

Loren Kohnfelder

Mahesh Prakriya

Aaron Ross Goldfeder

Michael Gashler

Darko Kirovski

Vance Palmer Morrison

Barry Clayton Bond

Dinarte R Morais

Gerald Reuben DeJean

John M Hawkins

Ivan Medvedev

Simon Jeremy Hall

Dario Russi

Jamie Lee Cool

Sergey A Khorun

Brian F Sullivan

Various techniques are described to protect secrets held by closed computing devices. In an ecosystem where devices operate and are offered a wide range of services from a service provider, the service provider may want to prevent users from sharing services between devices. In order to guarantee that services are not shared between devices, each device can be manufactured with a different set of secrets such as per device identifiers. Unscrupulous individuals may try to gain access to the secrets and transfer secrets from one device to another. In order to prevent this type of attack, each closed computing system can be manufactured to include a protected memory location that is tied to the device.

Tamper evidence per device protected identity

Various mechanisms are disclosed for protecting the security of memory in a computing environment. A security layer can have an encryption layer and a hashing layer that can dynamically encrypt and then dynamically hash sensitive information, as it is being loaded to dynamic memory of a computing device. For example, a memory unit that can correspond to a memory page can be processed by the security layer, and header data, code, and protect-worthy data can be secured, while other non-sensitive data can be left alone. Once such information is secured and stored in dynamic memory, it can be accessed at a later time by a processor and unencrypted and hash checked. Then, it can be loaded back onto the dynamic memory, thereby preventing direct memory access attacks.

Tamper resistant memory protection

An individualized per device initialization of a computing device is unique relative to the initialization of other computing devices. A common initialization program, common to all computing devices of a particular type such as a game console, may be modified to be unique for each computing device. Modification may comprise the application of at least one individualized per device secret, e.g., key, to at least a portion of the common initialization program such as at least one initialization stage. Initialization is tied to one or more device specific identities. In this way, initialization vulnerabilities discovered on a particular device cannot be exploited en masse on other computing devices because each initialization program stored in each computing device is unique. The device specific nature of the initialization program may be extended to other information input to the computing device in order to prevent unauthorized sharing of information with other computing devices.

Individualized per device initialization of computing devices in avoidance of mass exploitation of vulnerabilities

A one-time-settable tamper resistant software repository may be used in any computing system to store system information such as security violations and policies for responding to them. A one-time-settable tamper resistant software repository may be cryptographically signed, encrypted with a per device key and accessible by only the most privileged software executed by a computing device, e.g., hypervisor or operating system kernel. A one-time-settable tamper resistant software repository may be mirrored in RAM for performance. Recordable event fields in a software repository may be one-time-settable without the ability to reset them in a field operation mode whereas they may be resettable in a different mode such as a manufacturing mode. Memory allocated to a one-time-settable tamper resistant software repository may be reset, reclaimed, reassigned, scaled and otherwise flexibly adapted to changing conditions and priorities in the lifespan of a computing device, which may be particularly useful for service-backed consumer devices.

One time settable tamper resistant software repository

The subject innovation relates to systems and/or methodologies for using randomly positioned electromagnetic scatterers deposited into a predetermined region of an optical storage medium (e.g., Blue Ray DVD, HD DVD, CD, etc.) as a distinct three dimensional (3D) hard-to-copy digital rights management feature. A scatterers' topology is scanned using a matrix of antennas that sense the scatterers' electromagnetic response as the optical disc revolves at near-constant angular speed. The response, f, called an RF fingerprint, is then concatenated with an arbitrary text, t, that defines the digital rights management information imposed by the disk publisher.

Protecting optical media using random, moving radio frequency scatterers

An evidence-based policy manager generates a permission grant set for a code assembly received from a resource location. The policy manager executes in a computer system (e.g., a Web client or server) in combination with the verification module and class loader of the run-time environment. The permission grant set generated for a code assembly is applied in the run-time call stack to help the system determine whether a given system operation by the code assembly is authorized. The policy manager may determine a subset of the permission grant set based on a subset of the received code assembly's evidence, in order to expedite processing of the code assembly. When the evidence subset does not yield the desired permission subset, the policy manager may then perform an evaluation of all evidence received.

Partial grant set evaluation from partial evidence in an evidence-based security policy manager

A host intercepts calls between two executables and determines whether the calls are permissible according to the host's security model which can be identify based, such as user identity based&#x2014;for instance, mapping access rights within a specific data base user context to database object access. Such an identity security model differs from a common language runtime security model where managed code uses Code Access Security to prevent managed assemblies from performing certain operations. Managed assemblies registered with the host are host objects from the host's perspective for which access rights can be defined via security rules, such as are defined for individual user identities. A host can decide access between managed executables based on the host's identity based access rules by trapping any cross assembly calls and deciding whether such calls should proceed or be blocked from taking place based on the corresponding identity security settings.

Cross assembly call interception

All execution paths of one or more assemblies in managed code are simulated to find the permissions for each execution path. The managed code can correspond to a managed shared library or a managed application. Each call in each execution path has a corresponding permissions set. When the library or application has permissions to execute that are not less than the required permission sets for the execution paths, any dynamic execution of the library or application will not trigger a security exception The simulated execution provides a tool that can be used to ensure that code being written will not exceed a maximum security permission for the code. A permission set can be determined by the tool for each assembly corresponding to an application and for each entry point corresponding to a shared library.

Security requirement determination

Evidence-based application security may be implemented at the application and/or application group levels. A manifest may be provided defining at least one trust condition for the application or application group. A policy manager evaluates application evidence (e.g., an XrML license) for an application or group of applications relative to the manifest. The application is only granted permissions on the computer system if the application evidence indicates that the application is trusted. Similarly, a group of applications are only granted permissions on the computer system if the evidence indicates that the group of applications is trusted. If the application evidence satisfies the at least one trust condition defined by the manifest, the policy manager generates a permission grant set for each code assembly that is a member of the at least one application. Evidence may be further evaluated for code assemblies that are members of the trusted application or application group.

Evidence-based application security

A host operating in a managed environment intercepts a call from a managed caller to a particular callee and determines whether the call is permissible according to the host's prior configuration of a plurality of callees. The particular callee, which provides access to a resource that the host can be protecting, can have been previously configured by the host to always allow the call to be made, to never allow the call to be made, or to allow the call to be made based upon the degree to which the host trusts the managed caller.

Growing community of inventors

Seattle, WA, United States of America

Sebastian Lange