Network Working Group
Request for Comments: 4745
Category: Standards Track
H. Schulzrinne
Columbia U.
H. Tschofenig
Siemens Networks GmbH & Co KG
J. Morris
CDT
J. Cuellar
Siemens
J. Polk
J. Rosenberg
Cisco
February 2007

Common Policy: A Document Format for Expressing Privacy Preferences

Status of This Memo

This document specifies an Internet standards track protocol for the Internet community, and requests discussion and suggestions for improvements. Please refer to the current edition of the "Internet Official Protocol Standards" (STD 1) for the standardization state and status of this protocol. Distribution of this memo is unlimited.

Copyright Notice

Copyright © The IETF Trust (2007).

Abstract

This document defines a framework for authorization policies controlling access to application-specific data. This framework combines common location- and presence-specific authorization aspects. An XML schema specifies the language in which common policy rules are represented. The common policy framework can be extended to other application domains.

Table of Contents

   1. Introduction ....................................................3
   2. Terminology .....................................................4
   3. Modes of Operation ..............................................4
      3.1. Passive Request-Response - PS as Server (Responder) ........5
      3.2. Active Request-Response - PS as Client (Initiator) .........5
      3.3. Event Notification .........................................5
   4. Goals and Assumptions ...........................................6
   5. Non-Goals .......................................................7
   6. Basic Data Model and Processing .................................8
      6.1. Identification of Rules ....................................9
      6.2. Extensions .................................................9
   7. Conditions .....................................................10
      7.1. Identity Condition ........................................10
           7.1.1. Overview ...........................................10
           7.1.2. Matching One Entity ................................11
           7.1.3. Matching Multiple Entities .........................11
      7.2. Single Entity .............................................14
      7.3. Sphere ....................................................15
      7.4. Validity ..................................................16
   8. Actions ........................................................17
   9. Transformations ................................................18
   10. Procedure for Combining Permissions ...........................18
      10.1. Introduction .............................................18
      10.2. Combining Rules (CRs) ....................................18
      10.3. Example ..................................................19
   11. Meta Policies .................................................21
   12. Example .......................................................21
   13. XML Schema Definition .........................................22
   14. Security Considerations .......................................25
   15. IANA Considerations ...........................................25
      15.1. Common Policy Namespace Registration .....................25
      15.2. Content-type Registration for
            'application/auth-policy+xml' ............................26
      15.3. Common Policy Schema Registration ........................27
   16. References ....................................................27
      16.1. Normative References .....................................27
      16.2. Informative References ...................................28
   Appendix A. Contributors ..........................................29
   Appendix B. Acknowledgments .......................................29

1. Introduction

This document defines a framework for creating authorization policies for access to application-specific data. This framework is the result of combining the common aspects of single authorization systems that more specifically control access to presence and location information and that previously had been developed separately. The benefit of combining these two authorization systems is two-fold. First, it allows building a system that enhances the value of presence with location information in a natural way and reuses the same underlying authorization mechanism. Second, it encourages a more generic authorization framework with mechanisms for extensibility. The applicability of the framework specified in this document is not limited to policies controlling access to presence and location information data, but can be extended to other application domains.

The general framework defined in this document is intended to be accompanied and enhanced by application-specific policies specified elsewhere. The common policy framework described here is enhanced by domain-specific policy documents, including presence [7] and location [8]. This relationship is shown in Figure 1.

                           +-----------------+
                           |                 |
                           |     Common      |
                           |     Policy      |
                           |                 |
                           +---+---------+---+
                              /|\       /|\
                               |         |
      +-------------------+    |         |    +-------------------+
      |                   |    | enhance |    |                   |
      | Location-specific |    |         |    | Presence-specific |
      |      Policy       |----+         +----|      Policy       |
      |                   |                   |                   |
      +-------------------+                   +-------------------+

Figure 1: Common Policy Enhancements

This document starts with an introduction to the terminology in Section 2, an illustration of basic modes of operation in Section 3, a description of goals (see Section 4) and non-goals (see Section 5) of the policy framework, followed by the data model in Section 6. The structure of a rule, namely, conditions, actions, and transformations, is described in Sections 7, 8, and 9. The procedure for combining permissions is explained in Section 10 and used when conditions for more than one rule are satisfied. A short description of meta policies is given in Section 11. An example is provided in Section 12. The XML schema will be discussed in Section 13. IANA considerations in Section 15 follow security considerations in Section 14.

2. Terminology

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT","RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [1].

This document introduces the following terms:

   PT - Presentity / Target:  The PT is the entity about whom
   
        information has been requested.
   
   RM - Rule Maker:  The RM is an entity that creates the authorization
        rules that restrict access to data items.
   
   PS - (Authorization) Policy Server:  This entity has access to both
        the authorization policies and the data items.  In location-
        specific applications, the entity PS is labeled as location
        server (LS).
   
   WR - Watcher / Recipient:  This entity requests access to data items
        of the PT.  An access operation might be a read, a write, or any
        other operation.

A policy is given by a 'rule set' that contains an unordered list of 'rules'. A 'rule' has a 'conditions', an 'actions', and a 'transformations' part.

The term 'permission' indicates the action and transformation components of a 'rule'.

The term 'using protocol' is defined in [9]. It refers to the protocol used to request access to and to return privacy-sensitive data items.

3. Modes of Operation

The abstract sequence of operations can roughly be described as follows. The PS receives a query for data items for a particular PT, via the using protocol. The using protocol (or more precisely, the authentication protocol) provides the identity of the requestor, either at the time of the query or at the subscription time. The authenticated identity of the WR, together with other information provided by the using protocol or generally available to the server, is then used for searching through the rule set. All matching rules are combined according to a permission combining algorithm described in Section 10. The combined rules are applied to the application data, resulting in the application of privacy based on the transformation policies. The resulting application data is returned to the WR.

Three different modes of operation can be distinguished:

3.1. Passive Request-Response - PS as Server (Responder)

In a passive request-response mode, the WR queries the PS for data items about the PT. Examples of protocols following this mode of operation include HTTP, FTP, LDAP, finger, and various remote procedure call (RPC) protocols, including Sun RPC, Distributed Computing Environment (DCE), Distributed Component Object Model (DCOM), common object request broker architecture (Corba), and Simple Object Access Protocol (SOAP). The PS uses the rule set to determine whether the WR is authorized to access the PT's information, refusing the request if necessary. Furthermore, the PS might filter information by removing elements or by reducing the resolution of elements.

3.2. Active Request-Response - PS as Client (Initiator)

Alternatively, the PS may contact the WR and convey data items. Examples include HTTP, SIP session setup (INVITE request), H.323 session setup or SMTP.

3.3. Event Notification

Event notification adds a subscription phase to the "Active Request- Response - PS as Client (Initiator)" mode of operation. A watcher or subscriber asks to be added to the notification list for a particular presentity or event. When the presentity changes state or the event occurs, the PS sends a message to the WR containing the updated state. (Presence is a special case of event notification; thus, we often use the term interchangeably.)

In addition, the subscriber may itself add a filter to the subscription, limiting the rate or content of the notifications. If an event, after filtering by the rule-maker-provided rules and by the subscriber-provided rules, only produces the same notification content that was sent previously, no event notification is sent.

A single PS may authorize access to data items in more than one mode. Rather than having different rule sets for different modes all three modes are supported with a one rule set schema. Specific instances of the rule set can omit elements that are only applicable to the subscription model.

4. Goals and Assumptions

Below, we summarize our design goals and constraints.

Table representation:

Each rule must be representable as a row in a relational database. This design goal should allow efficient policy implementation by utilizing standard database optimization techniques.

Permit only:

Rules only provide permissions rather than denying them. Removing a rule can never increase permissions. Depending on the interpretation of 'deny' and 'permit' rules, the ordering of rules might matter, making updating rule sets more complicated since such update mechanisms would have to support insertion at specific locations in the rule set. Additionally, it would make distributed rule sets more complicated. Hence, only 'permit' actions are allowed that result in more efficient rule processing. This also implies that rule ordering is not important. Consequently, to make a policy decision requires processing all rules.

Additive permissions:

A query for access to data items is matched against the rules in the rule database. If several rules match, then the overall permissions granted to the WR are the union of those permissions. A more detailed discussion is provided in Section 10.

Upgradeable:

It should be possible to add additional rules later, without breaking PSs that have not been upgraded. Any such upgrades must not degrade privacy constraints, but PSs not yet upgraded may reveal less information than the rule maker would have chosen.

Capability support:

In addition to the previous goal, a RM should be able to determine which extensions are supported by the PS. The mechanism used to determine the capability of a PS is outside the scope of this specification.

Protocol-independent:

The rule set supports constraints on both notifications or queries as well as subscriptions for event-based systems such as presence systems.

No false assurance:

It appears more dangerous to give the user the impression that the system will prevent disclosure automatically, but fail to do so with a significant probability of operator error or misunderstanding, than to force the user to explicitly invoke simpler rules. For example, rules based on weekday and time-of- day ranges seem particularly subject to misinterpretation and false assumptions on part of the RM. (For example, a non- technical RM would probably assume that the rules are based on the time zone of his current location, which may not be known to other components of the system.)

5. Non-Goals

We explicitly decided that a number of possibly worthwhile capabilities are beyond the scope of this first version. Future versions may include these capabilities, using the extension mechanism described in this document. Non-goals include:

No external references:

Attributes within specific rules cannot refer to external rule sets, databases, directories, or other network elements. Any such external reference would make simple database implementation difficult and hence they are not supported in this version.

No regular expressions:

Conditions are matched on equality or 'greater-than'-style comparisons, not regular expressions, partial matches such as the SQL LIKE operator (e.g., LIKE "%foo%"), or glob-style matches ("*@example.com"). Most of these are better expressed as explicit elements.

No repeat times:

Repeat times (e.g., every day from 9 am to 4 pm) are difficult to make work correctly, due to the different time zones that PT, WR, PS, and RM may occupy. It appears that suggestions for including time intervals are often based on supporting work/non-work distinctions, which unfortunately are difficult to capture by time alone. Note that this feature must not be confused with the 'Validity' element that provides a mechanism to restrict the lifetime of a rule.

6. Basic Data Model and Processing

A rule set (or synonymously, a policy) consists of zero or more rules. The ordering of these rules is irrelevant. The rule set can be stored at the PS and conveyed from RM to PS as a single document, in subsets or as individual rules. A rule consists of three parts: conditions (see Section 7), actions (see Section 8), and transformations (see Section 9).

The conditions part is a set of expressions, each of which evaluates to either TRUE or FALSE. When a WR asks for information about a PT, the PS goes through each rule in the rule set. For each rule, it evaluates the expressions in the conditions part. If all of the expressions evaluate to TRUE, then the rule is applicable to this request. Generally, each expression specifies a condition based on some variable that is associated with the context of the request. These variables can include the identity of the WR, the domain of the WR, the time of day, or even external variables, such as the temperature or the mood of the PT.

Assuming that the rule is applicable to the request, the actions and transformations (commonly referred to as permissions) in the rule specify how the PS is supposed to handle this request. If the request is to view the location of the PT, or to view its presence, the typical action is "permit", which allows the request to proceed.

Assuming the action allows the request to proceed, the transformations part of the rule specifies how the information about the PT -- their location information, their presence, etc. -- is modified before being presented to the WR. These transformations are in the form of positive permissions. That is, they always specify a piece of information that is allowed to be seen by the WR. When a PS processes a request, it takes the transformations specified across all rules that match, and creates the union of them. For computing this union, the data type, such as Integer, Boolean, Set, or the Undef data type, plays a role. The details of the algorithm for combining permissions is described in Section 10. The resulting union effectively represents a "mask" -- it defines what information is exposed to the WR. This mask is applied to the actual location or presence data for the PT, and the data that is permitted by the mask is shown to the WR. If the WR requests a subset of information only (such as city-level civic location data only, instead of the full civic location information), the information delivered to the WR MUST be the intersection of the permissions granted to the WR and the data requested by the WR.

Rules are encoded in XML. To this end, Section 13 contains an XML schema defining the Common Policy Markup Language. This, however, is purely an exchange format between RM and PS. The format does not imply that the RM or the PS use this format internally, e.g., in matching a query with the policy rules. The rules are designed so that a PS can translate the rules into a relational database table, with each rule represented by one row in the database. The database representation is by no means mandatory; we will use it as a convenient and widely-understood example of an internal representation. The database model has the advantage that operations on rows have tightly defined meanings. In addition, it appears plausible that larger-scale implementations will employ a backend database to store and query rules, as they can then benefit from existing optimized indexing, access control, scaling, and integrity constraint mechanisms. Smaller-scale implementations may well choose different implementations, e.g., a simple traversal of the set of rules.

6.1. Identification of Rules

Each rule is equipped with a parameter that identifies the rule. This rule identifier is an opaque token chosen by the RM. A RM MUST NOT use the same identifier for two rules that are available to the PS at the same time for a given PT. If more than one RM modifies the same rule set, then it needs to be ensured that a unique identifier is chosen for each rule. A RM can accomplish this goal by retrieving the already specified rule set and choosing a new identifier for a rule that is different from the existing rule set.

6.2. Extensions

The policy framework defined in this document is meant to be extensible towards specific application domains. Such an extension is accomplished by defining conditions, actions, and transformations that are specific to the desired application domain. Each extension MUST define its own namespace.

Extensions cannot change the schema defined in this document, and this schema is not expected to change except via revision to this specification. Therefore, no versioning procedures for this schema or namespace are provided.

7. Conditions

The access to data items needs to be matched with the rule set stored at the PS. Each instance of a request has different attributes (e.g., the identity of the requestor) that are used for authorization. A rule in a rule set might have a number of conditions that need to be met before executing the remaining parts of a rule (i.e., actions and transformations). Details about rule matching are described in Section 10. This document specifies only a few conditions (i.e., identity, sphere, and validity). Further condition elements can be added via extensions to this document. If a child element of the <conditions> element is in a namespace that is not known or not supported, then this child element evaluates to FALSE.

As noted in Section 5, conditions are matched on equality or "greater than" style comparisons, rather than regular expressions. Equality is determined according to the rules for the data type associated with the element in the schema given in Section 13, unless explicit comparison steps are included in this document. For xs:anyURI types, readers may wish to consult [2] for its discussion xs:anyURI, as well as the text in Section 13.

7.1. Identity Condition

7.1.1. Overview

The identity condition restricts matching of a rule either to a single entity or a group of entities. Only authenticated entities can be matched; acceptable means of authentication are defined in protocol-specific documents. If the <identity> element is absent, identities are not considered, and thus, other conditions in the rule apply to any user, authenticated or not.

The <identity> condition is considered TRUE if any of its child elements (e.g., the <one/> and the <many/> elements defined in this document) evaluate to TRUE, i.e., the results of the individual child element are combined using a logical OR.

If a child element of the <identity> element is in a namespace that is not known or not supported, then this child element evaluates to FALSE.

7.1.2. Matching One Entity

The <one> element matches the authenticated identity (as contained in the 'id' attribute) of exactly one entity or user. For considerations regarding the 'id' attribute, refer to Section 7.2.

An example is shown below:

<?xml version="1.0" encoding="UTF-8"?> <ruleset xmlns="urn:ietf:params:xml:ns:common-policy">

       <rule id="f3g44r1">
           <conditions>
               <identity>
                   <one id="sip:alice@example.com"/>
                   <one id="tel:+1-212-555-1234" />
                   <one id="mailto:bob@example.net" />
               </identity>
           </conditions>
           <actions/>
           <transformations/>
       </rule>
   
   </ruleset>
   
   This example matches if the authenticated identity of the WR is
   either sip:alice@example.com, tel:+1-212-555-1234, or
   mailto:bob@example.net.

7.1.3. Matching Multiple Entities

The <many> element is a mechanism to perform authorization decisions based on the domain part of the authenticated identity. As such, it allows matching a large and possibly unknown number of users within a domain.

Furthermore, it is possible to include one or multiple <except> elements to exclude either individual users or users belonging to a specific domain. Excluding individual entities is implemented using a <except id="..."/> statement. The semantic of the 'id' attribute of the <except> element has the same meaning as the 'id' attribute of the <one> element (see Section 7.2). Excluding users belonging to a specific domain is implemented using the <except domain="..."/> element that excludes any user from the indicated domain.

If multiple <except> elements are listed as child elements of the <many> element, then the result of each <except> element is combined using a logical OR.

Common policy MUST either use UTF-8 or UTF-16 to store domain names in the 'domain' attribute. For non-IDNs (Internationalized Domain Names), lowercase ASCII SHOULD be used. For the comparison operation between the value stored in the 'domain' attribute and the domain value provided via the using protocol (referred to as "protocol domain identifier"), the following rules are applicable:

  1. Translate percent-encoding for either string.
  1. Convert both domain strings using the ToASCII operation described in RFC 3490 [3].
  1. Compare the two domain strings for ASCII equality, for each label. If the string comparison for each label indicates equality, the comparison succeeds. Otherwise, the domains are not equal.

If the conversion fails in step (2), the domains are not equal.

7.1.3.1. Matching Any Authenticated Identity

The <many/> element without any child elements or attributes matches any authenticated user.

The following example shows such a rule that matches any authenticated user:

<?xml version="1.0" encoding="UTF-8"?> <ruleset xmlns="urn:ietf:params:xml:ns:common-policy">

       <rule id="f3g44r5">
           <conditions>
               <identity>
                 <many/>
               </identity>
           </conditions>
           <actions/>
           <transformations/>
       </rule>
   
   </ruleset>
7.1.3.2. Matching Any Authenticated Identity Except Enumerated
Domains/Identities

The <many> element enclosing one or more <except domain="..."/> elements matches any user from any domain except those enumerated. The <except id="..."/> element excludes particular users. The semantics of the 'id' attribute of the <except> element is described in Section 7.2. The results of the child elements of the <many> element are combined using a logical OR.

An example is shown below:

<?xml version="1.0" encoding="UTF-8"?> <ruleset xmlns="urn:ietf:params:xml:ns:common-policy">

       <rule id="f3g44r1">
           <conditions>
               <sphere value="work"/>
               <identity>
                   <many>
                       <except domain="example.com"/>
                       <except domain="example.org"/>
                       <except id="sip:alice@bad.example.net"/>
                       <except id="sip:bob@good.example.net"/>
                       <except id="tel:+1-212-555-1234" />
                       <except id="sip:alice@example.com"/>
                   </many>
               </identity>
               <validity>
                   <from>2003-12-24T17:00:00+01:00</from>
                   <until>2003-12-24T19:00:00+01:00</until>
               </validity>
           </conditions>
           <actions/>
           <transformations/>
       </rule>
   
   </ruleset>

This example matches all users except any user in example.com, or any user in example.org or the particular users alice@bad.example.net, bob@good.example.net, and the user with the telephone number 'tel:+1-212-555-1234'. The last 'except' element is redundant since alice@example.com is already excluded through the first line.

7.1.3.3. Matching Any Authenticated Identity within a Domain Except
Enumerated Identities

The <many> element with a 'domain' attribute and zero or more <except id="..."/> elements matches any authenticated user from the indicated domain except those explicitly enumerated. The semantics of the 'id' attribute of the <except> element is described in Section 7.2.

It is nonsensical to have domains in the 'id' attribute that do not match the value of the 'domain' attribute in the enclosing <many> element.

An example is shown below:

<?xml version="1.0" encoding="UTF-8"?> <ruleset xmlns="urn:ietf:params:xml:ns:common-policy">

       <rule id="f3g44r1">
           <conditions>
               <identity>
                   <many domain="example.com">
                       <except id="sip:alice@example.com"/>
                       <except id="sip:bob@example.com"/>
                   </many>
               </identity>
           </conditions>
           <actions/>
           <transformations/>
       </rule>
   
   </ruleset>
   
   This example matches any user within example.com (such as
   carol@example.com) except alice@example.com and bob@example.com.

7.2. Single Entity

The 'id' attribute used in the <one> and in the <except> element refers to a single entity. In the subsequent text, we use the term 'single-user entity' as a placeholder for the <one> and the <except> element. The <except> element fulfills the purpose of excluding elements from the solution set.

A single-user entity matches the authenticated identity (as contained in the 'id' attribute) of exactly one entity or user. If there is a match, the single-user entity is considered TRUE. The single-user entity MUST NOT contain a 'domain' attribute.

The 'id' attribute contains an identity that MUST first be expressed as a URI. Applications using this framework must describe how the identities they are using can be expressed as URIs.

7.3. Sphere

The <sphere> element belongs to the group of condition elements. It can be used to indicate a state (e.g., 'work', 'home', 'meeting', 'travel') the PT is currently in. A sphere condition matches only if the PT is currently in the state indicated. The state may be conveyed by manual configuration or by some protocol. For example, RPID [10] provides the ability to inform the PS of its current sphere. The application domain needs to describe in more detail how the sphere state is determined. Switching from one sphere to another causes a switch between different modes of visibility. As a result, different subsets of rules might be applicable.

The content of the 'value' attribute of the <sphere> element MAY contain more than one token. The individual tokens MUST be separated by a blank character. A logical OR is used for the matching the tokens against the sphere settings of the PT. As an example, if the content of the 'value' attribute in the sphere attribute contains two tokens 'work' and 'home' then this part of the rule matches if the sphere for a particular PT is either 'work' OR 'home'. To compare the content of the 'value' attribute in the <sphere> element with the stored state information about the PT's sphere setting a case-insensitive string comparison MUST be used for each individual token. There is neither a registry for these values nor a language- specific indication of the sphere content. As such, the tokens are treated as opaque strings.

<?xml version="1.0" encoding="UTF-8"?> <ruleset xmlns="urn:ietf:params:xml:ns:common-policy">

     <rule id="f3g44r2">
       <conditions>
         <sphere value="work"/>
         <identity>
           <one id="sip:andrew@example.com"/>
         </identity>
       </conditions>
       <actions/>
       <transformations/>
     </rule>
     <rule id="y6y55r2">
       <conditions>
         <sphere value="home"/>
         <identity>
           <one id="sip:allison@example.com"/>
         </identity>
       </conditions>
       <actions/>
       <transformations/>
     </rule>
   
     <rule id="z6y55r2">
       <conditions>
         <identity>
              <one id="sip:john@doe.example.com"/>
         </identity>
         <sphere value="home work"/>
       </conditions>
       <actions/>
       <transformations/>
     </rule>
   </ruleset>

The rule example above illustrates that the rule with the entity andrew@example.com matches if the sphere is been set to 'work'. In the second rule, the entity allison@example.com matches if the sphere is set to 'home'. The third rule also matches since the value in the sphere element also contains the token 'home'.

7.4. Validity

The <validity> element is the third condition element specified in this document. It expresses the rule validity period by two attributes, a starting and an ending time. The validity condition is TRUE if the current time is greater than or equal to at least one <from> child, but less than the <until> child after it. This represents a logical OR operation across each <from> and <until> pair. Times are expressed in XML dateTime format. A rule maker might not always have access to the PS to invalidate some rules that grant permissions. Hence, this mechanism allows invalidating granted permissions automatically without further interaction between the rule maker and the PS. The PS does not remove the rules; instead the rule maker has to clean them up.

An example of a rule fragment is shown below:

<?xml version="1.0" encoding="UTF-8"?> <ruleset xmlns="urn:ietf:params:xml:ns:common-policy">

     <rule id="f3g44r3">
       <conditions>
           <validity>
               <from>2003-08-15T10:20:00.000-05:00</from>
               <until>2003-09-15T10:20:00.000-05:00</until>
           </validity>
       </conditions>
       <actions/>
       <transformations/>
     </rule>
   </ruleset>

The <validity> element MUST have the <from> and <until> subelements in pairs. Multiple <from> and <until> elements might appear in pairs (i.e., without nesting of <from> and <until> elements). Using multiple <validity> elements as subelements of the <conditions> element is not useful since all subelements of the <conditions> element are combined as a logical AND.

8. Actions

While conditions are the 'if'-part of rules, actions and transformations form their 'then'-part. The actions and transformations parts of a rule determine which operations the PS MUST execute after having received from a WR a data access request that matches all conditions of this rule. Actions and transformations only permit certain operations; there is no 'deny' functionality. Transformations exclusively specify PS-side operations that lead to a modification of the data items requested by the WR. Regarding location data items, for instance, a transformation could force the PS to lower the precision of the location information that is returned to the WR.

Actions, on the other hand, specify all remaining types of operations the PS is obliged to execute, i.e., all operations that are not of transformation type. Actions are defined by application-specific usages of this framework. The reader is referred to the corresponding extensions to see examples of such elements.

9. Transformations

Two sub-parts follow the conditions part of a rule: transformations and actions. As defined in Section 8, transformations specify operations that the PS MUST execute and that modify the result that is returned to the WR. This functionality is particularly helpful in reducing the granularity of information provided to the WR, as, for example, required for location privacy. Transformations are defined by application-specific usages of this framework.

A simple transformation example is provided in Section 10.

10. Procedure for Combining Permissions

10.1. Introduction

This section describes how rules are selected and how actions and permissions are determined. When a PS receives a request for access to privacy-sensitive data, the request is matched against the rule set. A rule matches if all conditions contained as child elements in the <conditions> element of a rule evaluate to TRUE. Each type of condition defines when it is TRUE. All rules where the conditions match the request form the matching rule set. The permissions in the matching rule set are combined using a set of combining rules (CRs) described in Section 10.2.

10.2. Combining Rules (CRs)

Each type of permission is combined across all matching rules. Each type of action or transformation is combined separately and independently. The combining rules generate a combined permission. The combining rules depend only on the data type of permission. If a particular permission type has no value in a rule, it assumes the lowest possible value for that permission for the purpose of computing the combined permission. That value is given by the data type for booleans (FALSE) and sets (empty set), and MUST be defined by any extension to the Common Policy for other data types.

For boolean permissions, the resulting permission is TRUE if and only if at least one permission in the matching rule set has a value of TRUE and FALSE otherwise. For integer, real-valued and date-time permissions, the resulting permission is the maximum value across the permission values in the matching set of rules. For sets, it is the union of values across the permissions in the matching rule set.

10.3. Example

In the following example we illustrate the process of combining permissions. We will consider three conditions for our purpose, namely those of name identity (WR-ID), sphere, and validity (from,until). The ID column is used as a rule identifier. For editorial reasons we omit the domain part of the WR's identity.

We use two actions in our example, namely X and Y. The values of X and Y are of data types Boolean and Integer, respectively.

The transformation, referred to as Z, uses values that can be set either to '+' (or 3), 'o' (or 2) or '-' (or 1). Permission Z allows us to show the granularity reduction whereby a value of '+' shows the corresponding information unrestricted, and '-' shows nothing. This permission might be related to location information or other presence attributes like mood. Internally, we use the data type Integer for computing the permission of this attribute.

The label 'NULL' in the table indicates that no value is available for a particular cell.

         Conditions                  Actions/Transformations
     +---------------------------------+---------------------+
     | Id  WR-ID    sphere  from until |  X       Y     Z    |
     +---------------------------------+---------------------+
     |  1   bob      home    A1    A2  |  TRUE    10    o    |
     |  2   alice    work    A1    A2  |  FALSE   5     +    |
     |  3   bob      work    A1    A2  |  TRUE    3     -    |
     |  4   tom      work    A1    A2  |  TRUE    5     +    |
     |  5   bob      work    A1    A3  |  NULL    12    o    |
     |  6   bob      work    B1    B2  |  FALSE   10    -    |
     +---------------------------------+---------------------+

Again for editorial reasons, we use the following abbreviations for the two <validity> attributes 'from' and 'until':

     A1=2003-12-24T17:00:00+01:00
     A2=2003-12-24T21:00:00+01:00
     A3=2003-12-24T23:30:00+01:00
     B1=2003-12-22T17:00:00+01:00
     B2=2003-12-23T17:00:00+01:00

Note that B1 < B2 < A1 < A2 < A3.

The entity 'bob' acts as a WR and requests data items. The rule set consists of the six rules shown in the table and identified by the values 1 to 6 in the 'Id' column. The PS receives the query at

2003-12-24T17:15:00+01:00, which falls between A1 and A2. In our example, we assume that the sphere value of the PT is currently set to 'work'.

As a first step, it is necessary to determine which rules fire by evaluating the conditions part of each of them.

Rule 1 does not match since the sphere condition does not match. Rule 2 does not match as the identity of the WR (here 'alice') does not equal 'bob'. Rule 3 matches since all conditions evaluate to TRUE. Rule 4 does not match as the identity of the WR (here 'tom') does not equal 'bob'. Rule 5 matches. Rule 6 does not match since the rule is not valid anymore.

Only rules 3 and 5 fire. We use the actions and transformations part of these two rules to determine the combined permission, as shown below.

             Actions/Transformations
     +-----+-----------------------+
     | Id  |  X       Y      Z     |
     +-----+-----------------------+
     |  3  |  TRUE     3     -     |
     |  5  |  NULL    12     o     |
     +-----+-----------------------+

Each column is treated independently. The combined value of X is set to TRUE since the NULL value equals FALSE according to the description in Section 10.2. For the column with the name Y, we apply the maximum of 3 and 12, so that the combined value of Y is 12. For column Z, we again compute the maximum of 'o' and '-' (i.e., 2 and 1) which is 'o' (2).

The combined permission for all three columns is therefore:

             Actions/Transformations
           +-----------------------+
           |  X       Y      Z     |
           +-----------------------+
           |  TRUE    12     o     |
           +-----------------------+

11. Meta Policies

Meta policies authorize a rule maker to insert, update, or delete a particular rule or an entire rule set. Some authorization policies are required to prevent unauthorized modification of rule sets. Meta policies are outside the scope of this document.

A simple implementation could restrict access to the rule set only to the PT but more sophisticated mechanisms could be useful. As an example of such policies, one could think of parents configuring the policies for their children.

12. Example

This section gives an example of an XML document valid with respect to the XML schema defined in Section 13. Semantically richer examples can be found in documents that extend this schema with application-domain-specific data (e.g., location or presence information).

Below a rule is shown with a condition that matches for a given authenticated identity (bob@example.com) and within a given time period. Additionally, the rule matches only if the target has set its sphere to 'work'.

<?xml version="1.0" encoding="UTF-8"?> <ruleset xmlns="urn:ietf:params:xml:ns:common-policy">

       <rule id="f3g44r1">
           <conditions>
               <identity>
                   <one id="sip:bob@example.com"/>
               </identity>
               <sphere value="work"/>
               <validity>
                   <from>2003-12-24T17:00:00+01:00</from>
                   <until>2003-12-24T19:00:00+01:00</until>
               </validity>
           </conditions>
           <actions/>
           <transformations/>
       </rule>
   </ruleset>

13. XML Schema Definition

This section provides the XML schema definition for the common policy markup language described in this document.

<?xml version="1.0" encoding="UTF-8"?> <xs:schema targetNamespace="urn:ietf:params:xml:ns:common-policy"

    xmlns:cp="urn:ietf:params:xml:ns:common-policy"
    xmlns:xs="http://www.w3.org/2001/XMLSchema"
    elementFormDefault="qualified" attributeFormDefault="unqualified">
    <!-- /ruleset -->
    <xs:element name="ruleset">
        <xs:complexType>
            <xs:complexContent>
                <xs:restriction base="xs:anyType">
                    <xs:sequence>
                        <xs:element name="rule" type="cp:ruleType"
                        minOccurs="0" maxOccurs="unbounded"/>
                    </xs:sequence>
                </xs:restriction>
            </xs:complexContent>
        </xs:complexType>
    </xs:element>
    <!-- /ruleset/rule -->
    <xs:complexType name="ruleType">
        <xs:complexContent>
            <xs:restriction base="xs:anyType">
                <xs:sequence>
                    <xs:element name="conditions"
                    type="cp:conditionsType" minOccurs="0"/>
                    <xs:element name="actions"
                    type="cp:extensibleType" minOccurs="0"/>
                    <xs:element name="transformations"
                    type="cp:extensibleType" minOccurs="0"/>
                </xs:sequence>
                <xs:attribute name="id" type="xs:ID" use="required"/>
            </xs:restriction>
        </xs:complexContent>
    </xs:complexType>
    <!-- //rule/conditions -->
    <xs:complexType name="conditionsType">
        <xs:complexContent>
            <xs:restriction base="xs:anyType">
                <xs:choice maxOccurs="unbounded">
                    <xs:element name="identity"
                    type="cp:identityType" minOccurs="0"/>
                    <xs:element name="sphere"
                    type="cp:sphereType" minOccurs="0"/>
                    <xs:element name="validity"
                    type="cp:validityType" minOccurs="0"/>
                    <xs:any namespace="##other" processContents="lax"
                    minOccurs="0" maxOccurs="unbounded"/>
                </xs:choice>
            </xs:restriction>
        </xs:complexContent>
    </xs:complexType>
    <!-- //conditions/identity -->
    <xs:complexType name="identityType">
        <xs:complexContent>
            <xs:restriction base="xs:anyType">
                <xs:choice  minOccurs="1" maxOccurs="unbounded">
                    <xs:element name="one" type="cp:oneType"/>
                    <xs:element name="many" type="cp:manyType"/>
                    <xs:any namespace="##other" processContents="lax"/>
                </xs:choice>
            </xs:restriction>
        </xs:complexContent>
    </xs:complexType>
    <!-- //identity/one -->
    <xs:complexType name="oneType">
        <xs:complexContent>
            <xs:restriction base="xs:anyType">
                <xs:sequence>
                    <xs:any namespace="##other"
                    minOccurs="0" processContents="lax"/>
                </xs:sequence>
                <xs:attribute name="id"
                type="xs:anyURI" use="required"/>
            </xs:restriction>
        </xs:complexContent>
    </xs:complexType>
    <!-- //identity/many -->
    <xs:complexType name="manyType">
        <xs:complexContent>
            <xs:restriction base="xs:anyType">
                <xs:choice minOccurs="0" maxOccurs="unbounded">
                    <xs:element name="except" type="cp:exceptType"/>
                    <xs:any namespace="##other"
                    minOccurs="0" processContents="lax"/>
                </xs:choice>
                <xs:attribute name="domain"
                use="optional" type="xs:string"/>
            </xs:restriction>
        </xs:complexContent>
    </xs:complexType>
    <!-- //many/except -->

    <xs:complexType name="exceptType">
        <xs:attribute name="domain" type="xs:string" use="optional"/>
        <xs:attribute name="id" type="xs:anyURI" use="optional"/>
    </xs:complexType>
    <!-- //conditions/sphere -->
    <xs:complexType name="sphereType">
        <xs:complexContent>
            <xs:restriction base="xs:anyType">
                <xs:attribute name="value"
                type="xs:string" use="required"/>
            </xs:restriction>
        </xs:complexContent>
    </xs:complexType>
    <!-- //conditions/validity -->
    <xs:complexType name="validityType">
        <xs:complexContent>
            <xs:restriction base="xs:anyType">
                <xs:sequence minOccurs="1" maxOccurs="unbounded">
                    <xs:element name="from" type="xs:dateTime"/>
                    <xs:element name="until" type="xs:dateTime"/>
                </xs:sequence>
            </xs:restriction>
        </xs:complexContent>
    </xs:complexType>
    <!-- //rule/actions or //rule/transformations -->
    <xs:complexType name="extensibleType">
        <xs:complexContent>
            <xs:restriction base="xs:anyType">
                <xs:sequence>
                    <xs:any namespace="##other" processContents="lax"
                    minOccurs="0" maxOccurs="unbounded"/>
                </xs:sequence>
            </xs:restriction>
        </xs:complexContent>
    </xs:complexType>
</xs:schema>

14. Security Considerations

This document describes a framework for policies. This framework is intended to be enhanced elsewhere by application-domain-specific data. Security considerations are to a great extent application-data dependent, and therefore need to be covered by documents that extend the framework defined in this specification. However, new action and transformation permissions along with their allowed values must be defined in a way so that the usage of the permissions combining rules of Section 10 does not lower the level of privacy protection. See Section 10 for more details on this privacy issue.

15. IANA Considerations

This section registers a new XML namespace, a new XML schema, and a new MIME type. This section registers a new XML namespace per the procedures in [4].

15.1. Common Policy Namespace Registration

   URI:  urn:ietf:params:xml:ns:common-policy
   
   Registrant Contact:  IETF GEOPRIV working group, Henning Schulzrinne
      (hgs+geopriv@cs.columbia.edu).

XML:

   BEGIN
   <?xml version="1.0"?>
   <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML Basic 1.0//EN"
     "http://www.w3.org/TR/xhtml-basic/xhtml-basic10.dtd">
   <html xmlns="http://www.w3.org/1999/xhtml">
   <head>
     <meta http-equiv="content-type"
           content="text/html;charset=iso-8859-1"/>
     <title>Common Policy Namespace</title>
   </head>
   <body>
     <h1>Namespace for Common Authorization Policies</h1>
     <h2>urn:ietf:params:xml:ns:common-policy</h2>
   <p>See <a href="ftp://ftp.rfc-editor.org/in-notes/rfc4745.txt">
      RFC 4745</a>.</p>
   </body>
   </html>
   END

15.2. Content-type Registration for 'application/auth-policy+xml'

This specification requests the registration of a new MIME type according to the procedures of RFC 4288 [5] and guidelines in RFC 3023 [6].

   MIME media type name:  application
   
   MIME subtype name:  auth-policy+xml
   
   Mandatory parameters:  none
   
   Optional parameters:  charset

Indicates the character encoding of enclosed XML.

Encoding considerations:

Uses XML, which can employ 8-bit characters, depending on the character encoding used. See RFC 3023 [6], Section 3.2.

Security considerations:

This content type is designed to carry authorization policies. Appropriate precautions should be adopted to limit disclosure of this information. Please refer to Section 14 of RFC 4745 and to the security considerations described in Section 10 of RFC 3023 [6] for more information.

   Interoperability considerations:  None
   
   Published specification:  RFC 4745

Applications which use this media type:

Presence- and location-based systems

Additional information:

      Magic Number:  None
      
      File Extension:  .apxml
      
      Macintosh file type code:  'TEXT'

Personal and email address for further information:

      Hannes Tschofenig, Hannes.Tschofenig@siemens.com
   
   Intended usage:  LIMITED USE

Author:

This specification is a work item of the IETF GEOPRIV working group, with mailing list address <geopriv@ietf.org>.

Change controller:

      The IESG <iesg@ietf.org>

15.3. Common Policy Schema Registration

   URI:  urn:ietf:params:xml:schema:common-policy
   
   Registrant Contact:  IETF GEOPRIV working group, Henning Schulzrinne
      (hgs+geopriv@cs.columbia.edu).
   
   XML:  The XML schema to be registered is contained in Section 13.
      Its first line is
   
   <?xml version="1.0" encoding="UTF-8"?>

and its last line is

   </xs:schema>

16. References

16.1. Normative References

   [1]  Bradner, S., "Key words for use in RFCs to Indicate Requirement
        Levels", BCP 14, RFC 2119, March 1997.
   
   [2]  Duerst, M. and M. Suignard, "Internationalized Resource
        Identifiers (IRIs)", RFC 3987, January 2005.
   
   [3]  Faltstrom, P., Hoffman, P., and A. Costello, "Internationalizing
        Domain Names in Applications (IDNA)", RFC 3490, March 2003.
   
   [4]  Mealling, M., "The IETF XML Registry", BCP 81, RFC 3688,
        January 2004.
   
   [5]  Freed, N. and J. Klensin, "Media Type Specifications and
        Registration Procedures", BCP 13, RFC 4288, December 2005.
   
   [6]  Murata, M., St. Laurent, S., and D. Kohn, "XML Media Types",
        RFC 3023, January 2001.

16.2. Informative References

   [7]  Rosenberg, J., "Presence Authorization Rules", Work in Progress,
        June 2006.
   
   [8]  Schulzrinne, H., Tschofenig, H., Morris, J., Cuellar, J., and J.
        Polk, "A Document Format for Expressing Privacy Preferences for
        Location Information", Work in Progress, February 2006.
   
   [9]  Cuellar, J., Morris, J., Mulligan, D., Peterson, J., and J.
        Polk, "Geopriv Requirements", RFC 3693, February 2004.
   
   [10] Schulzrinne, H., Gurbani, V., Kyzivat, P., and J. Rosenberg,
        "RPID: Rich Presence Extensions to the Presence Information Data
        Format (PIDF)", RFC 4480, July 2006.

Appendix A. Contributors

We would like to thank Christian Guenther for his help with initial versions of this document.

Appendix B. Acknowledgments

This document is partially based on the discussions within the IETF GEOPRIV working group. Discussions at the Geopriv Interim Meeting 2003 in Washington, D.C., helped the working group to make progress on the authorization policies based on the discussions among the participants.

We particularly want to thank Allison Mankin <mankin@psg.com>, Randall Gellens <rg+ietf@qualcomm.com>, Andrew Newton <anewton@ecotroph.net>, Ted Hardie <hardie@qualcomm.com>, and Jon Peterson <jon.peterson@neustar.biz> for discussing a number of details with us. They helped us to improve the quality of this document. Allison, Ted, and Andrew also helped us to make good progress with the internationalization support of the identifier/ domain attributes.

Furthermore, we would like to thank the IETF SIMPLE working group for their discussions of J. Rosenberg's draft on presence authorization policies. We would also like to thank Stefan Berg, Murugaraj Shanmugam, Christian Schmidt, Martin Thomson, Markus Isomaki, Aki Niemi, Eva Maria Leppanen, Josip Matanovic, and Mark Baker for their comments. Martin Thomson helped us with the XML schema. Mark Baker provided a review of the media type. Scott Brim provided a review on behalf of the General Area Review Team.

Authors' Addresses

   Henning Schulzrinne
   Columbia University
   Department of Computer Science
   450 Computer Science Building
   New York, NY  10027
   USA
   
   Phone: +1 212 939 7042
   EMail: schulzrinne@cs.columbia.edu
   URI:   http://www.cs.columbia.edu/~hgs
   
   Hannes Tschofenig
   Siemens Networks GmbH & Co KG
   Otto-Hahn-Ring 6
   Munich, Bavaria  81739
   Germany
   
   EMail: Hannes.Tschofenig@siemens.com
   URI:   http://www.tschofenig.com
   
   John B. Morris, Jr.
   Center for Democracy and Technology
   1634 I Street NW, Suite 1100
   Washington, DC  20006
   USA
   
   EMail: jmorris@cdt.org
   URI:   http://www.cdt.org
   
   Jorge R. Cuellar
   Siemens
   Otto-Hahn-Ring 6
   Munich, Bavaria  81739
   Germany
   
   EMail: Jorge.Cuellar@siemens.com
   James Polk
   Cisco
   2200 East President George Bush Turnpike
   Richardson, Texas  75082
   USA

EMail:

          jmpolk@cisco.com
   
   Jonathan Rosenberg
   Cisco Systems
   600 Lanidex Plaza
   Parsippany, New York  07054
   USA
   
   EMail: jdrosen@cisco.com
   URI:   http://www.jdrosen.net

Full Copyright Statement

Copyright © The IETF Trust (2007).

This document is subject to the rights, licenses and restrictions contained in BCP 78, and except as set forth therein, the authors retain all their rights.

This document and the information contained herein are provided on an "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Intellectual Property

The IETF takes no position regarding the validity or scope of any Intellectual Property Rights or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; nor does it represent that it has made any independent effort to identify any such rights. Information on the procedures with respect to rights in RFC documents can be found in BCP 78 and BCP 79.

Copies of IPR disclosures made to the IETF Secretariat and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementers or users of this specification can be obtained from the IETF on-line IPR repository at http://www.ietf.org/ipr.

The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights that may cover technology that may be required to implement this standard. Please address the information to the IETF at ietf-ipr@ietf.org.

Acknowledgement

Funding for the RFC Editor function is currently provided by the Internet Society.