<?xml version='1.0' encoding='utf-8'?>
<!DOCTYPE rfc [
  <!ENTITY nbsp    "&#160;">
  <!ENTITY zwsp   "&#8203;">
  <!ENTITY nbhy   "&#8209;">
  <!ENTITY wj     "&#8288;">
]>
<?xml-stylesheet type="text/xsl" href="rfc2629.xslt" ?>
<!-- generated by https://github.com/cabo/kramdown-rfc version 1.7.39 (Ruby 3.4.9) -->
<rfc xmlns:xi="http://www.w3.org/2001/XInclude" ipr="trust200902" docName="draft-dsmullen-ppd-architecture-09" category="info" tocInclude="true" sortRefs="true" symRefs="true" version="3">
  <!-- xml2rfc v2v3 conversion 3.33.0 -->
  <front>
    <title abbrev="PPDArch">Privacy Preference Declaration for Home Networks</title>
    <seriesInfo name="Internet-Draft" value="draft-dsmullen-ppd-architecture-09"/>
    <author fullname="Daniel Smullen">
      <organization>CableLabs</organization>
      <address>
        <email>d.smullen@cablelabs.com</email>
      </address>
    </author>
    <author fullname="Brian Scriber">
      <organization>CableLabs</organization>
      <address>
        <email>brian.scriber@computer.org</email>
      </address>
    </author>
    <date year="2026" month="May" day="22"/>
    <keyword>privacy preferences</keyword>
    <keyword>home networks</keyword>
    <keyword>internet of things</keyword>
    <keyword>device transparency</keyword>
    <abstract>
      <?line 37?>

<t>This document describes an architecture for signaling household privacy preferences to devices in home networks through Privacy Preference Declarations (PPDs).  The architecture enables a PPD participant to discover a PPD service endpoint, establish trust in that endpoint through the applicable protocol and security profile, retrieve the applicable household policy instance, and acknowledge receipt of that policy instance.  The acknowledgment establishes that a specific policy instance was made available to the participant; it does not, by itself, assert anything about the participant's subsequent behavior.</t>
    </abstract>
    <note removeInRFC="true">
      <name>About This Document</name>
      <t>
        The latest revision of this draft can be found at <eref target="https://drspangle.github.io/draft-dsmullen-ppd-architecture/draft-dsmullen-ppd-architecture.html"/>.
        Status information for this document may be found at <eref target="https://datatracker.ietf.org/doc/draft-dsmullen-ppd-architecture/"/>.
      </t>
      <t>Source for this draft and an issue tracker can be found at
        <eref target="https://github.com/drspangle/draft-dsmullen-ppd-architecture"/>.</t>
    </note>
  </front>
  <middle>
    <?line 41?>

<section anchor="introduction">
      <name>Introduction</name>
      <t>Internet-connected home devices commonly collect sensitive data, yet the tools
available to households for understanding or controlling that collection are
often fragmented, confusing, or absent. When privacy settings do exist, they
frequently vary widely in semantics and presentation across devices and
services.</t>
      <t>The result is a fragmented operational model. Households must manage privacy
through device-specific controls, while vendors and service providers have no
common way to receive household privacy preferences across devices. That lack
of a shared signaling model makes it harder for households to understand which
participants have been presented with which privacy expectations, and harder
for implementers to support interoperable behavior.</t>
      <t><xref target="RFC7258"/> frames mass data collection as a technical threat, urging protocol designers to limit exposure through encryption and data minimization.
While this principle is crucial in adversarial, internet-scale contexts, the model proposed in this document takes a different approach: rather than hiding data flows, it seeks to govern them.
Privacy here is not achieved by making devices blind, but by making user-defined preferences visible to devices and associated services.</t>
      <t>This use of privacy is related to, but distinct from, the privacy guidance in <xref target="RFC6973"/>, which emphasizes reduced observability, linkability, and identifiability in protocol design.
Those properties remain important, but PPD focuses on a different home-network problem: a user needs a consistent way to express household privacy preferences and to know that those preferences were made available to participating devices or associated services.
This document is also aligned with the user-agency goals described in <xref target="RFC8280"/>, but it is narrower and more operational.
It describes an architecture for privacy-preference signaling and recordkeeping, not a general framework for human-rights analysis or for constraining device behavior.
Home networks are a significant and operationally important IoT environment.
They commonly place a local administrative boundary around large numbers of devices, many with limited or no end-user interface, making them a concrete target for a privacy-preference signaling architecture.
In this architecture, discovery identifies candidate participant-facing service endpoints.
Trust in a selected endpoint, and in the policy instances it presents, is established separately through the applicable protocol and security mechanisms rather than by discovery alone.
This also addresses an asymmetry common in current deployments: the household user is often required to acknowledge device- or vendor-defined terms, while the household has no comparable way to record that a participating device or associated service was presented with the household's privacy policy.
PPD introduces a reciprocal signaling path in which presentation and acknowledgment of a household policy instance can be recorded by the household domain.
The objective is to provide a coherent architectural basis for devices and associated services to retrieve, acknowledge, and keep current with household privacy preferences within that administrative domain.</t>
    </section>
    <section anchor="conventions-and-definitions">
      <name>Conventions and Definitions</name>
      <t>The key words "<bcp14>MUST</bcp14>", "<bcp14>MUST NOT</bcp14>", "<bcp14>REQUIRED</bcp14>", "<bcp14>SHALL</bcp14>", "<bcp14>SHALL
NOT</bcp14>", "<bcp14>SHOULD</bcp14>", "<bcp14>SHOULD NOT</bcp14>", "<bcp14>RECOMMENDED</bcp14>", "<bcp14>NOT RECOMMENDED</bcp14>",
"<bcp14>MAY</bcp14>", and "<bcp14>OPTIONAL</bcp14>" in this document are to be interpreted as
described in BCP 14 <xref target="RFC2119"/> <xref target="RFC8174"/> when, and only when, they
appear in all capitals, as shown here.</t>
      <?line -18?>

<t>The terms below define both protocol roles and core concepts used by this architecture.
The definitions of privacy, transparency, and user control are included here because they describe the conceptual scope of PPD rather than separate protocol mechanisms.</t>
      <section anchor="terminology">
        <name>Terminology</name>
        <dl>
          <dt>Privacy:</dt>
          <dd>
            <t>In this document, the ability of users to understand and shape how data about them, their household, or their home environment is collected, used, retained, and shared by devices and associated services.</t>
          </dd>
          <dt>Transparency:</dt>
          <dd>
            <t>The property that data practices are made visible and understandable to the user, including what data is collected, how it is processed, where it is shared, and what policy preferences apply.</t>
          </dd>
          <dt>User control:</dt>
          <dd>
            <t>The ability of a user or household to define privacy preferences and make those preferences visible to devices or associated services in a consistent and actionable way.</t>
          </dd>
          <dt>Privacy Preference Declaration (PPD):</dt>
          <dd>
            <t>A structured expression of household privacy preferences that can be discovered, retrieved, and acknowledged by PPD participants.</t>
          </dd>
          <dt>PPD service endpoint:</dt>
          <dd>
            <t>A participant-facing service, and the baseline discovery target for participants, through which a PPD participant discovers, retrieves, and acknowledges applicable policy instances.</t>
          </dd>
          <dt>Policy authority:</dt>
          <dd>
            <t>The authoritative source of household policy state and of any inputs used to derive an effective policy for a participant.  The policy authority may be local or remote.  Participants are not required to discover or address the policy authority directly in the baseline architecture.</t>
          </dd>
          <dt>Household policy:</dt>
          <dd>
            <t>A policy selected or maintained for a home network that represents the household's privacy preferences.</t>
          </dd>
          <dt>Effective policy derivation:</dt>
          <dd>
            <t>The logical function, performed by or on behalf of the policy authority, that determines the effective policy instance for a participant.</t>
          </dd>
          <dt>Effective policy:</dt>
          <dd>
            <t>The policy instance that applies to a particular PPD participant at a particular time, after effective policy derivation has resolved household policy state and any applicable participant-specific inputs.</t>
          </dd>
          <dt>PPD participant:</dt>
          <dd>
            <t>A device, or a trusted intermediary such as a backend service acting on behalf of a device, that participates in PPD by retrieving and acknowledging an applicable policy instance.</t>
          </dd>
          <dt>Policy instance:</dt>
          <dd>
            <t>A specific version or representation of an effective policy that can be identified for acknowledgment and recordkeeping.</t>
          </dd>
          <dt>Association:</dt>
          <dd>
            <t>The state established when a PPD participant has retrieved the current applicable effective policy and acknowledged receipt of that specific policy instance.</t>
          </dd>
          <dt>Current association:</dt>
          <dd>
            <t>Association state that still corresponds to the latest applicable effective policy for the participant and remains fresh according to the renewal model enforced by the PPD service endpoint.</t>
          </dd>
          <dt>Association freshness:</dt>
          <dd>
            <t>The property that an association remains within the bounded interval, or before the renewal deadline, accepted by the PPD service endpoint for treating that association as current.</t>
          </dd>
          <dt>Stale association:</dt>
          <dd>
            <t>Association state that still refers to the latest applicable effective policy instance, but whose freshness can no longer be confirmed because renewal did not occur within the bounded interval accepted by the PPD service endpoint.</t>
          </dd>
          <dt>Needs reassociation:</dt>
          <dd>
            <t>A state in which current association cannot be confirmed because the applicable effective policy changed, participant state relevant to effective policy derivation changed, or enough state was lost that the existing association no longer applies reliably.</t>
          </dd>
          <dt>Reassociation:</dt>
          <dd>
            <t>The process by which a PPD participant recovers from stale association or a needs-reassociation state and re-establishes current association.</t>
          </dd>
          <dt>Broken association:</dt>
          <dd>
            <t>A state in which stored or reported information is contradictory or incomplete enough that current association cannot be determined reliably.</t>
          </dd>
          <dt>Policy acknowledgment:</dt>
          <dd>
            <t>A signal that a PPD participant has received a specific effective policy instance.  A policy acknowledgment is not a statement that the device is compatible with every policy term or that the device will behave in a particular way.</t>
          </dd>
          <dt>Network-observed device:</dt>
          <dd>
            <t>A device that is visible to the local network through ordinary network observation but that has not established association through PPD.</t>
          </dd>
          <dt>Unmanaged device:</dt>
          <dd>
            <t>A network-observed device that is not known to participate in PPD or is not currently manageable through PPD association.</t>
          </dd>
        </dl>
      </section>
    </section>
    <section anchor="limitations-of-existing-mechanisms">
      <name>Limitations of Existing Mechanisms</name>
      <t>Current mechanisms for managing data privacy within the home environment exhibit limitations.</t>
      <section anchor="device-specific-configurations">
        <name>Device-specific Configurations</name>
        <t>Individual devices often employ unique privacy settings, thereby complicating user management of privacy across the entire network.
This complexity can inadvertently lead to unintended data sharing.</t>
      </section>
      <section anchor="ineffective-and-unusable-user-interfaces">
        <name>Ineffective and Unusable User Interfaces</name>
        <t>Navigating and configuring privacy settings on individual devices can be a time-consuming and frustrating experience for users.
These ineffective interfaces often lead users to habitually agree to relax their privacy preferences without fully understanding the implications of their decisions.
This fosters a general resignation towards privacy management, making it difficult for users to exert meaningful control over their personal data and ultimately compromising their privacy expectations.</t>
      </section>
      <section anchor="relationship-to-existing-work">
        <name>Relationship to Existing Work</name>
        <section anchor="dnt-and-p3p">
          <name>DNT and P3P</name>
          <t>Protocols like Do Not Track (DNT) and Platform for Privacy Preferences Project (P3P) have not achieved widespread adoption and have proven inadequate for addressing nuanced privacy needs.
These protocols do not provide the participant-specific policy signaling, lifecycle handling, or home-network operational posture needed here.
They also do not provide a practical basis for recording that a participating device or associated service was presented with a household policy instance.</t>
        </section>
        <section anchor="mud">
          <name>MUD</name>
          <t>Manufacturer Usage Description (MUD) <xref target="RFC8520"/> is the closest existing precedent for device-to-home-network signaling.
MUD is focused on manufacturer-defined network communication intent presented to local network infrastructure.
PPD addresses a different problem: household-defined privacy preference signaling, participant-specific effective policy presentation, and recordkeeping about whether a participant was presented with a current household policy instance.
The two approaches may complement each other in a deployment, but MUD does not provide the privacy-policy lifecycle or recordkeeping model described here.</t>
        </section>
        <section anchor="privacy-vocabularies-and-policy-models">
          <name>Privacy Vocabularies and Policy Models</name>
          <t>Vocabulary and policy-expression efforts such as the Data Privacy Vocabulary (DPV) and ODRL are closer to the content layer than to the signaling layer.
PPD does not attempt to replace such work with a new general-purpose ontology or rights language.
Instead, PPD separates concerns: this architecture defines roles and
lifecycle, the companion taxonomy work defines the core fields and shared
computable floor that can map to richer vocabularies, and the companion
protocol work defines the participant-facing signaling path by which an
effective household policy is presented and acknowledged.</t>
        </section>
      </section>
    </section>
    <section anchor="operational-scenarios">
      <name>Operational Scenarios</name>
      <t>This section describes representative operational cases for the architecture in home-network environments.
The scenarios focus on discovery, association, reassociation, and mixed-participant visibility rather than on user-interface details.</t>
      <section anchor="initial-discovery-and-association">
        <name>Initial Discovery and Association</name>
        <t>A PPD participant joins the home network and obtains one or more candidate PPD service endpoints through configuration or local network discovery.
In a common home deployment model, the PPD service endpoint is hosted by a residential gateway or equivalent home-network service.
Discovery identifies reachability, not authority.
The participant establishes a secure connection to a selected endpoint, confirms that endpoint through the applicable trust mechanism, retrieves the applicable effective policy instance, and acknowledges receipt of that policy instance.
The PPD service endpoint may present policy derived from a local or remote policy authority without exposing that internal topology to the participant.
At the end of this process, the participant has established association if the current applicable effective policy has been delivered and acknowledged.
The PPD service endpoint also determines the initial freshness state of that association.</t>
      </section>
      <section anchor="policy-update-and-reassociation">
        <name>Policy Update and Reassociation</name>
        <t>The household policy, or the participant's effective policy, changes.
The PPD service endpoint immediately invalidates current association for the participant.
The participant enters a needs-reassociation state until it retrieves and acknowledges the updated effective policy instance.
This scenario illustrates that association state is tied to a specific policy instance and not to prior acknowledgments alone.
Reassociation re-establishes current association by confirming that the participant has seen the latest applicable policy instance.</t>
      </section>
      <section anchor="association-freshness-expiry-and-renewal">
        <name>Association Freshness Expiry and Renewal</name>
        <t>The applicable effective policy instance is unchanged, but the participant does
not renew within the bounded interval accepted by the PPD service endpoint.
The association becomes stale even though no policy change occurred.
The participant no longer has current association until it completes the
required renewal procedure.
This scenario distinguishes stale association from a needs-reassociation state
caused by a changed policy instance.</t>
      </section>
      <section anchor="participant-state-change">
        <name>Participant State Change</name>
        <t>A participant changes in a way that can affect the applicable effective policy instance, such as a declaration update, capability change, or other state change relevant to effective policy derivation.
The PPD service endpoint determines that current association can no longer be confirmed using existing state alone.
The participant then retrieves and acknowledges the newly applicable effective policy instance.
This scenario keeps the architecture focused on policy signaling and recordkeeping without assuming that every state change requires the same local handling or transport behavior.</t>
      </section>
      <section anchor="mixed-participant-network-visibility">
        <name>Mixed-Participant Network Visibility</name>
        <t>A home network contains both PPD participants and devices that do not participate in PPD.
The household can still use local management functions to distinguish associated participants, participants whose current association cannot be confirmed, and network-observed or unmanaged devices.
This scenario illustrates that non-participating devices are an expected operational reality.
Their presence can inform transparency and local management decisions, but it does not create association or change the baseline signaling role of PPD.</t>
      </section>
    </section>
    <section anchor="goals">
      <name>Goals</name>
      <section anchor="enhance-user-control">
        <name>Enhance User Control</name>
        <ul spacing="normal">
          <li>
            <t>Support a household's ability to define privacy preferences that can be made available consistently across participating devices and associated services.</t>
          </li>
          <li>
            <t>Ensure the household does not have to normalize or mentally reconcile each participant's local privacy vocabulary or interpretation strategy in order to express those preferences.</t>
          </li>
          <li>
            <t>Provide an architectural basis for recording whether the current applicable policy was made available to a participant.</t>
          </li>
          <li>
            <t>Create a reciprocal acknowledgment model in which the household can retain a record that a participant or associated service was presented with, and acknowledged, a specific household policy instance.</t>
          </li>
        </ul>
      </section>
      <section anchor="promote-interoperability">
        <name>Promote Interoperability</name>
        <ul spacing="normal">
          <li>
            <t>Establish a standardized mechanism for devices from diverse manufacturers to discover PPD service endpoints, retrieve applicable privacy policies, and acknowledge policy instances.</t>
          </li>
          <li>
            <t>Support consistent association and reassociation behavior across heterogeneous participants.</t>
          </li>
        </ul>
      </section>
      <section anchor="enable-flexibility">
        <name>Enable Flexibility</name>
        <ul spacing="normal">
          <li>
            <t>Allow deployments to place policy storage and effective-policy derivation locally or remotely without changing the baseline participant-facing contract.</t>
          </li>
          <li>
            <t>Leave room for deployment-specific protocol profiles where constrained environments or different operational models require them, including trusted-intermediary participation for devices that cannot satisfy the minimum authenticated direct-participant bar.</t>
          </li>
        </ul>
      </section>
      <section anchor="facilitate-transparency">
        <name>Facilitate Transparency</name>
        <ul spacing="normal">
          <li>
            <t>Provide a basis for local management functions to distinguish currently associated participants, stale or reassociation-needed participants, and non-participating devices.</t>
          </li>
          <li>
            <t>Improve visibility into which participants have been presented with the current applicable policy instance, without implying enforcement of device behavior.</t>
          </li>
        </ul>
      </section>
    </section>
    <section anchor="scope">
      <name>Scope</name>
      <t>This document defines a high-level architectural framework for Privacy Preference Declaration (PPD) in home-network environments.
It focuses on roles, trust boundaries, lifecycle meaning, and operational assumptions for making household privacy preferences available to devices and associated services.</t>
      <t>This document does not delve into specific implementation details, such as message formats, data structures, security algorithms, or user interface design.
Furthermore, this document does not define mechanisms that modify device behavior, legal and regulatory considerations, or specific security protocols.
Where this document discusses recordkeeping, that recordkeeping is limited to signaling and recording that an applicable household policy was made available to and acknowledged by a PPD participant.
That recordkeeping can provide a basis for later accountability, audit, or dispute analysis, but this document does not define enforcement behavior or prove subsequent compliance.
That boundary is intentional: the architectural problem addressed here is
interoperable preference signaling and recordkeeping across heterogeneous home
deployments, while enforcement depends on deployment-specific control points,
trust models, and participant capabilities that cannot be assumed uniformly at
the architectural layer.</t>
      <t>Specific message formats, transport details, and semantic field definitions
are carried by the companion protocol and taxonomy specifications:
<xref target="I-D.draft-dsmullen-ppd-protocol"/> and
<xref target="I-D.draft-dsmullen-ppd-taxonomy"/>.</t>
    </section>
    <section anchor="architecture-overview">
      <name>Architecture Overview</name>
      <section anchor="assumptions">
        <name>Assumptions</name>
        <t>This document makes the following assumptions:</t>
        <ul spacing="normal">
          <li>
            <t>Resource Constraints: Participants and home-network components may be
constrained in processing power, memory, or bandwidth. The architecture
therefore favors lightweight participant-facing interaction. Where a device
cannot satisfy the minimum authenticated direct-participant bar, this
architecture expects indirect participation through a trusted intermediary
rather than weakening the meaning of direct participation.</t>
          </li>
          <li>
            <t>Single User Policy: Each participant is assumed to be governed by one
effective household policy at a time. Multi-user reconciliation may be
relevant in some deployments, but it is outside the baseline architecture.</t>
          </li>
          <li>
            <t>Endpoint Discovery and Trust: Configuration or local-network mechanisms can
identify candidate PPD service endpoints, but discovery alone does not
establish authority. The applicable protocol profile needs a separate way to
authenticate the selected endpoint and confirm that the policy it presents
is authoritative for the participant's household context.</t>
          </li>
        </ul>
      </section>
      <section anchor="association-state-and-freshness">
        <name>Association State and Freshness</name>
        <t>This architecture treats the PPD service endpoint as the source of truth for
participant association state.
A participant establishes association when it retrieves and acknowledges a
specific applicable effective policy instance.</t>
        <t>Current association exists only when both of the following are true:</t>
        <ul spacing="normal">
          <li>
            <t>the acknowledged policy instance still corresponds to the latest applicable
effective policy for that participant; and</t>
          </li>
          <li>
            <t>the association remains fresh according to the renewal model enforced by the
PPD service endpoint.</t>
          </li>
        </ul>
        <t>If the applicable effective policy instance is unchanged but the freshness
interval expires before renewal, the participant enters stale association.
If the applicable effective policy changes, if participant state relevant to
effective policy derivation changes, or if enough state is lost that the prior
association can no longer be trusted, the participant enters a
needs-reassociation state.
In either case, the participant no longer has current association.</t>
        <t>Participant-initiated exchanges provide the renewal or recovery path, but they
are not the source of truth for whether association is current.
The PPD service endpoint determines whether a participant is current, stale, or
in needs reassociation.</t>
      </section>
      <section anchor="discovery-and-policy-authority-boundary">
        <name>Discovery and Policy-Authority Boundary</name>
        <t>This architecture separates discovery of a participant-facing service endpoint
from trust establishment.
A participant may learn one or more candidate PPD service endpoints through
configuration or local network mechanisms, but discovery alone does not make
any candidate authoritative.
Before treating a policy instance as authoritative, the participant needs the
applicable protocol profile to authenticate the selected endpoint and confirm
that it is authorized to present policy for the household context.</t>
        <t>The participant-facing contract is the PPD service endpoint, not direct access
to the policy authority.
A deployment may place storage, policy combination, and effective policy
derivation behind that service.
When the PPD service endpoint and policy authority are distinct, the deployment
needs to preserve at least:</t>
        <ul spacing="normal">
          <li>
            <t>authenticity of the effective policy instance presented to the participant;</t>
          </li>
          <li>
            <t>integrity of policy-instance identifiers and association-freshness metadata;</t>
          </li>
          <li>
            <t>an unambiguous binding between the selected PPD service endpoint and the
policy authority on whose behalf it presents policy.</t>
          </li>
        </ul>
        <t>These are architectural invariants.
The companion protocol specification document, <xref target="I-D.draft-dsmullen-ppd-protocol"/>, defines the participant-facing transport, metadata confirmation, and security-profile expectations,
while deployment profiles still choose concrete mechanism details.</t>
      </section>
      <section anchor="key-components">
        <name>Key Components</name>
        <t>User Interface: A user-friendly interface (e.g., mobile app, web portal) for creating and managing privacy preferences.</t>
        <t>PPD Service Endpoint: A participant-facing service through which PPD participants discover, retrieve, and acknowledge applicable policy instances.
In a common home deployment model, this service is hosted by a residential gateway or equivalent home-network service.
A participant may learn candidate PPD service endpoints through configuration or local network discovery, but it treats a selected endpoint as authoritative only after the applicable trust mechanism succeeds.</t>
        <t>Policy Authority: The authoritative source of household policy state and any inputs used for effective policy derivation.
The policy authority may be local or remote.
A PPD service endpoint can obtain policy from a policy authority without exposing internal storage or computation topology to participants.
Participants are not required to discover or communicate with the policy authority directly in the baseline architecture.</t>
        <t>Effective Policy Derivation: The logical function, performed by or on behalf of the policy authority, that determines the applicable policy instance for a participant.</t>
        <t>Participant Declarations and Consent Requests: Optional participant inputs that can disclose data-handling declarations or request consent for uses not covered by baseline policy.
These inputs are distinct from the minimal path of policy retrieval and policy acknowledgment.
Where a deployment exposes a coarse comparison result for participant declarations at the protocol boundary, that result belongs on the declaration path rather than in the effective policy or policy acknowledgment objects. That comparison surface is diagnostic only; it is not a baseline negotiation channel, policy-relaxation mechanism, or homeowner-prompt path.</t>
        <t>Recordkeeping and Management Mechanisms: Deployment-specific mechanisms for presenting association state, participant status, effective policy views, and network-observed devices to the household.
Such mechanisms are not device-behavior requirements in the baseline PPD architecture.</t>
      </section>
      <section anchor="data-flows">
        <name>Data Flows</name>
        <t>This section outlines the high-level interactions between users, PPD
participants, the PPD service endpoint, and the policy authority.</t>
        <t>The process begins when a household defines privacy preferences. Those
preferences may express which types of data may be collected, under what
conditions data may be processed or shared, and which retention practices are
acceptable. User-interface design for authoring those preferences is out of
scope, as are the detailed semantic fields and qualifier families used in the
policy representation; those are defined in
<xref target="I-D.draft-dsmullen-ppd-taxonomy"/>.</t>
        <t>Once created, the preferences are maintained by a policy authority, which may
be local or remote and may include storage, effective-policy derivation, or
both. When a new participant joins the home network, it obtains one or more
candidate PPD service endpoints through configuration or local-network
mechanisms. Discovery identifies reachable candidates, but does not by itself
establish authority. The participant authenticates a selected endpoint
according to the applicable protocol profile, retrieves the applicable policy
instance, and acknowledges receipt of that instance. In some deployments, the
participant is a backend service associated with the device rather than the
local device itself.</t>
        <t>The participant-facing contract ends at the PPD service endpoint; any split
between that service and the policy authority is internal to the deployment.
Where those components are distinct, the deployment preserves the authenticity
and integrity of the effective policy instance, policy-instance identifier,
and freshness metadata presented through the service endpoint. Participants
may optionally report declarations at this stage. The service endpoint also
determines the freshness interval or renewal deadline for the resulting
association state.</t>
        <t>If a participant seeks to perform actions not permitted under the baseline
policy, it may initiate a consent request workflow. The design and behavior of
that workflow are out of scope here. Future specifications should ensure that
consent interactions are clear, proportionate, and resistant to manipulative
or fatiguing prompting.</t>
        <t>Current association is not indefinite.
If the participant does not renew before the freshness interval expires, the PPD service endpoint treats the association as stale even if the applicable effective policy instance is unchanged.
Reassociation is required when current association can no longer be confirmed.
This can occur because the applicable effective policy changed, participant
state relevant to effective policy derivation changed, the association became
stale, or enough state was lost that the prior association can no longer be
trusted. Reassociation re-establishes current association by retrieving and
acknowledging the latest applicable effective policy instance, or by
completing the applicable renewal procedure when the policy instance is
unchanged.</t>
        <t>The companion protocol specification document,
<xref target="I-D.draft-dsmullen-ppd-protocol"/>, defines the participant-facing message
formats, baseline machine-readable encoding, and the way association freshness
is conveyed. This architecture remains limited to the signaling and
recordkeeping meaning of those interactions. It does not define how device
behavior is changed by policy, nor how deployments respond when a participant
cannot satisfy a given policy.</t>
      </section>
      <section anchor="non-ppd-and-network-observed-devices">
        <name>Non-PPD and Network-Observed Devices</name>
        <t>Home networks commonly include devices that do not implement PPD, cannot be updated to implement PPD, or are visible only through local network observation.
The architecture treats these devices as expected operational cases rather than exceptional failures.</t>
        <t>A local management function can classify such devices as network-observed or unmanaged based on information available within the home network.
That classification can improve household transparency by showing that a device is present even though it has not established association through PPD.
Network observation does not create association, does not imply that the device has received a household policy, and does not imply anything about the device's behavior.</t>
        <t>Any local response to unmanaged devices, such as notification, inventory display, or other network management action, is a deployment decision outside the baseline PPD signaling architecture.</t>
      </section>
    </section>
    <section anchor="policy-language">
      <name>Policy Language</name>
      <t>The specific details of the privacy policy language are out of scope for this
document. The policy vocabulary and taxonomy of privacy concepts and
attributes are defined in <xref target="I-D.draft-dsmullen-ppd-taxonomy"/>, including the
compact identifier model, the shared computable semantic floor, extension
namespaces, and the mapping expectations used by the companion protocol
specification.</t>
      <section anchor="language-requirements">
        <name>Language Requirements</name>
        <ul spacing="normal">
          <li>
            <t>Human-readable: Policies should be easily understandable by users.</t>
          </li>
          <li>
            <t>Machine-readable: Policies should be machine-processable for automated interpretation and signaling.</t>
          </li>
          <li>
            <t>Extensible: The language should be flexible enough to accommodate evolving privacy needs and technologies.</t>
          </li>
          <li>
            <t>Internationalization-compatible: Policies and identifiers used within them may need to support multilingual environments and non-ASCII characters.</t>
          </li>
        </ul>
        <t>To ensure consistent interpretation and comparison of string-based policy elements, such as device names, labels, or category identifier string handling practices should align with the guidelines defined in <xref target="RFC7564"/>.
This is particularly important when identifiers or user-facing labels are created, stored, or matched across vendors or systems that operate in different locales or character encodings.</t>
      </section>
    </section>
    <section anchor="future-work">
      <name>Future Work</name>
      <t>This document defines the architectural layer for PPD. The companion protocol
and taxonomy specifications define the participant-facing protocol and shared
semantic model. The remaining future work is therefore in adjacent areas that
this architecture intentionally leaves out of scope.</t>
      <section anchor="consent-request-workflows">
        <name>Consent Request Workflows</name>
        <t>The mechanism by which devices request additional user consent for data uses
not covered by the baseline policy is out of scope. Future specifications
should:</t>
        <ul spacing="normal">
          <li>
            <t>Define clear constraints to prevent manipulative or fatiguing consent flows (e.g., dark patterns).</t>
          </li>
          <li>
            <t>Describe consent interactions that are transparent, infrequent, proportionate, and user-respecting.</t>
          </li>
          <li>
            <t>Explore user interface standards or API affordances to preserve meaningful choice.</t>
          </li>
        </ul>
        <t>This is a sensitive area and needs to balance user experience, privacy
expectations, and implementation feasibility.</t>
      </section>
      <section anchor="recordkeeping-and-local-management">
        <name>Recordkeeping and Local Management</name>
        <t>This architecture does not define how devices act on privacy policies or how
departures from policy are detected or remediated.
Future work may include:</t>
        <ul spacing="normal">
          <li>
            <t>Optional participant status reporting models and device-side implementation expectations.</t>
          </li>
          <li>
            <t>Recordkeeping mechanisms for correlating policy delivery and acknowledgment records.</t>
          </li>
          <li>
            <t>State models that distinguish current, stale, and needs-reassociation participant status.</t>
          </li>
          <li>
            <t>Deconfliction strategies for devices unable to meet all user-defined constraints.</t>
          </li>
          <li>
            <t>Deployment-local management options, such as notifications or inventory display.</t>
          </li>
        </ul>
      </section>
      <section anchor="user-interface-design">
        <name>User Interface Design</name>
        <t>The user-facing interface used to author, modify, and review privacy preferences is out of scope.
Future design guidance may address:</t>
        <ul spacing="normal">
          <li>
            <t>User experience design principles for presenting privacy concepts clearly and accessibly.</t>
          </li>
          <li>
            <t>Models for progressive disclosure of policy impact.</t>
          </li>
          <li>
            <t>Multi-user and household-role-specific control models (e.g., parental vs. administrative roles).</t>
          </li>
        </ul>
      </section>
      <section anchor="interoperability-testing-and-reference-implementations">
        <name>Interoperability Testing and Reference Implementations</name>
        <t>Future work may also include:</t>
        <ul spacing="normal">
          <li>
            <t>Development of reference implementations of the PPD protocol, PPD service endpoint, and policy-authority components.</t>
          </li>
          <li>
            <t>Interoperability testing across devices and vendors.</t>
          </li>
          <li>
            <t>Conformance guidelines and self-certification procedures.</t>
          </li>
        </ul>
      </section>
    </section>
    <section anchor="iana-considerations">
      <name>IANA Considerations</name>
      <t>This document has no IANA actions.</t>
    </section>
    <section anchor="internationalization-considerations">
      <name>Internationalization Considerations</name>
      <t>In contexts where privacy preferences or taxonomy elements involve user-facing or vendor-defined string identifiers, additional work may be required to:</t>
      <ul spacing="normal">
        <li>
          <t>Define string normalization and comparison rules, particularly for internationalized text.</t>
        </li>
        <li>
          <t>Support identifier consistency across diverse vendors and locales.</t>
        </li>
        <li>
          <t>Consider alignment with <xref target="RFC7564"/> for handling Unicode-aware identifiers in a secure and interoperable way.</t>
        </li>
      </ul>
    </section>
    <section anchor="security-considerations">
      <name>Security Considerations</name>
      <t>For a privacy framework to be effective, it needs to support the expression of user preferences and protect those preferences during transmission, retrieval, and acknowledgment.
This section outlines safeguards for confidentiality, authenticity, integrity, and metadata minimization during PPD operations.</t>
      <section anchor="secure-policy-dissemination">
        <name>Secure Policy Dissemination</name>
        <t>Communication between PPD participants and the PPD service endpoint needs protection against unauthorized access and tampering.
When the PPD service endpoint and policy authority are distinct, deployments also need to preserve policy authenticity and integrity across that boundary.
Discovery mechanisms can identify candidate PPD service endpoints, but discovery alone is not sufficient to establish that an endpoint is authorized to present household policy.
The companion protocol specification document, <xref target="I-D.draft-dsmullen-ppd-protocol"/>, defines explicit participant-facing security profiles and the accountability properties they need to provide. Future deployment profiles still need to identify concrete cryptographic mechanisms, such as encryption and mutual authentication, so that legitimate participants can retrieve privacy policies and detect modification.
Those deployment profiles also need to protect the binding between the authenticated participant-facing service endpoint and the policy state it presents.</t>
      </section>
      <section anchor="anonymity-and-metadata-protection">
        <name>Anonymity and Metadata Protection</name>
        <t>Even when privacy policies themselves do not contain sensitive personal information, the act of retrieving or acknowledging a policy can reveal characteristics about the household, such as the types of devices in use, specific user preferences, or behavioral patterns over time.
<xref target="RFC7258"/> cautions against protocol designs that expose unnecessary metadata, treating the accumulation of such information as a legitimate technical threat.
This framework takes that warning seriously: metadata exposure during policy retrieval and device onboarding needs to be minimized to avoid turning privacy infrastructure into a new source of privacy leakage.
Concepts from <xref target="RFC9577"/> may help inform this effort. <xref target="RFC9577"/> introduces techniques for authorization without identification, enabling a client to prove it is authorized without revealing who it is.
While <xref target="RFC9577"/> is optimized for pseudonymous web authentication over the public internet and assumes a centralized token issuer model, its core ideas, particularly around unlinkable token presentation, could be adapted to the PPD protocol to reduce metadata correlation and minimize household identifiability during policy exchanges.
However, this needs careful analysis, as the assumptions of <xref target="RFC9577"/> do not fully align with the goals or context of a local, user-governed home network.</t>
      </section>
      <section anchor="policy-integrity">
        <name>Policy Integrity</name>
        <t>Devices need assurance that the policy retrieved is authentic and unaltered.
Integrity protections, such as digital signatures, are necessary to ensure that users' preferences cannot be tampered with in transit or at rest by other devices, malicious actors, or misconfigurations.
If policy is obtained through a participant-facing service from a distinct policy authority, integrity protections also need to cover the policy-instance identifier and any freshness metadata presented through that service.</t>
      </section>
      <section anchor="device-authentication">
        <name>Device Authentication</name>
        <t>Devices participating in the privacy framework need an authentication model before accessing the PPD service endpoint.
This limits policy dissemination to known, authorized participants and helps users maintain trust in the integrity of their home network's privacy relationships.
If the PPD service endpoint and policy authority are distinct, the deployment also needs a way to preserve the authenticity of policy state presented through the participant-facing service.
By aligning with the concerns raised in <xref target="RFC7258"/> and incorporating ideas from <xref target="RFC9577"/> where appropriate, this framework seeks to protect users not only from overt data collection, but also from silent inference and passive metadata surveillance.
At the same time, it avoids treating anonymity as an end in itself.
The goal is to support privacy with recordkeeping, where user-defined preferences are signaled consistently, devices are identifiable only as much as necessary for the exchange, and the user retains visibility into what occurs within their domain.</t>
      </section>
      <section anchor="policy-acknowledgment-and-recordkeeping">
        <name>Policy Acknowledgment and Recordkeeping</name>
        <t>PPD participants need a way to acknowledge receipt of the applicable privacy policy instance.
This acknowledgment should be recorded and verifiable so that the household can determine which participants have seen the current policy.
The record needs to bind the participant identity, the acknowledged policy instance, and the time or sequence context in which the acknowledgment was made.
For devices that rely on a backend service, the record also needs to distinguish between acknowledgment by the local device and acknowledgment by the backend service acting on behalf of that device.
This record is important because it creates a reciprocal acknowledgment path.
In many current deployments, the household user is asked to acknowledge device or vendor policy terms, but there is no comparably strong household-controlled record that the participant was presented with the household's own privacy policy.
An authenticated and integrity-protected acknowledgment record allows the household to show that presentation and acknowledgment occurred, which can support later accountability or review even when the architecture does not define automated enforcement.
The companion protocol specification document, <xref target="I-D.draft-dsmullen-ppd-protocol"/>, defines baseline acknowledgment semantics and the protection properties acknowledgment mechanisms need to provide. Future deployment profiles still need concrete mechanisms that remain practical for constrained home-network devices.
At minimum, the selected mechanism needs to provide:</t>
        <ul spacing="normal">
          <li>
            <t>participant authentication sufficient to bind the acknowledgment to the device or backend service that made it;</t>
          </li>
          <li>
            <t>policy-instance integrity so that the acknowledged policy can be identified unambiguously;</t>
          </li>
          <li>
            <t>freshness or sequencing so that an old acknowledgment cannot be replayed as evidence of current association;</t>
          </li>
          <li>
            <t>verifiability sufficient for the acknowledgment record to function as a protected receipt of policy presentation and acknowledgment; and</t>
          </li>
          <li>
            <t>a way to retain or export the acknowledgment record without exposing more household metadata than necessary.</t>
          </li>
        </ul>
        <t>A policy acknowledgment is not, by itself, an assertion about subsequent device behavior.
Any local response to non-participation or other local observations is outside the baseline signaling mechanism defined by this architecture.</t>
      </section>
    </section>
  </middle>
  <back>
    <references anchor="sec-combined-references">
      <name>References</name>
      <references anchor="sec-normative-references">
        <name>Normative References</name>
        <?line 608?>

<reference anchor="RFC2119">
          <front>
            <title>Key words for use in RFCs to Indicate Requirement Levels</title>
            <author fullname="S. Bradner" initials="S." surname="Bradner"/>
            <date month="March" year="1997"/>
            <abstract>
              <t>In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements.</t>
            </abstract>
          </front>
          <seriesInfo name="BCP" value="14"/>
          <seriesInfo name="RFC" value="2119"/>
          <seriesInfo name="DOI" value="10.17487/RFC2119"/>
        </reference>
        <reference anchor="RFC8174">
          <front>
            <title>Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words</title>
            <author fullname="B. Leiba" initials="B." surname="Leiba"/>
            <date month="May" year="2017"/>
            <abstract>
              <t>RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings.</t>
            </abstract>
          </front>
          <seriesInfo name="BCP" value="14"/>
          <seriesInfo name="RFC" value="8174"/>
          <seriesInfo name="DOI" value="10.17487/RFC8174"/>
        </reference>
      </references>
      <references anchor="sec-informative-references">
        <name>Informative References</name>
        <reference anchor="RFC7258">
          <front>
            <title>Pervasive Monitoring Is an Attack</title>
            <author fullname="S. Farrell" initials="S." surname="Farrell"/>
            <author fullname="H. Tschofenig" initials="H." surname="Tschofenig"/>
            <date month="May" year="2014"/>
            <abstract>
              <t>Pervasive monitoring is a technical attack that should be mitigated in the design of IETF protocols, where possible.</t>
            </abstract>
          </front>
          <seriesInfo name="BCP" value="188"/>
          <seriesInfo name="RFC" value="7258"/>
          <seriesInfo name="DOI" value="10.17487/RFC7258"/>
        </reference>
        <reference anchor="RFC6973">
          <front>
            <title>Privacy Considerations for Internet Protocols</title>
            <author fullname="A. Cooper" initials="A." surname="Cooper"/>
            <author fullname="H. Tschofenig" initials="H." surname="Tschofenig"/>
            <author fullname="B. Aboba" initials="B." surname="Aboba"/>
            <author fullname="J. Peterson" initials="J." surname="Peterson"/>
            <author fullname="J. Morris" initials="J." surname="Morris"/>
            <author fullname="M. Hansen" initials="M." surname="Hansen"/>
            <author fullname="R. Smith" initials="R." surname="Smith"/>
            <date month="July" year="2013"/>
            <abstract>
              <t>This document offers guidance for developing privacy considerations for inclusion in protocol specifications. It aims to make designers, implementers, and users of Internet protocols aware of privacy-related design choices. It suggests that whether any individual RFC warrants a specific privacy considerations section will depend on the document's content.</t>
            </abstract>
          </front>
          <seriesInfo name="RFC" value="6973"/>
          <seriesInfo name="DOI" value="10.17487/RFC6973"/>
        </reference>
        <reference anchor="RFC8280">
          <front>
            <title>Research into Human Rights Protocol Considerations</title>
            <author fullname="N. ten Oever" initials="N." surname="ten Oever"/>
            <author fullname="C. Cath" initials="C." surname="Cath"/>
            <date month="October" year="2017"/>
            <abstract>
              <t>This document aims to propose guidelines for human rights considerations, similar to the work done on the guidelines for privacy considerations (RFC 6973). The other parts of this document explain the background of the guidelines and how they were developed.</t>
              <t>This document is the first milestone in a longer-term research effort. It has been reviewed by the Human Rights Protocol Considerations (HRPC) Research Group and also by individuals from outside the research group.</t>
            </abstract>
          </front>
          <seriesInfo name="RFC" value="8280"/>
          <seriesInfo name="DOI" value="10.17487/RFC8280"/>
        </reference>
        <reference anchor="RFC8520">
          <front>
            <title>Manufacturer Usage Description Specification</title>
            <author fullname="E. Lear" initials="E." surname="Lear"/>
            <author fullname="R. Droms" initials="R." surname="Droms"/>
            <author fullname="D. Romascanu" initials="D." surname="Romascanu"/>
            <date month="March" year="2019"/>
            <abstract>
              <t>This memo specifies a component-based architecture for Manufacturer Usage Descriptions (MUDs). The goal of MUD is to provide a means for end devices to signal to the network what sort of access and network functionality they require to properly function. The initial focus is on access control. Later work can delve into other aspects.</t>
              <t>This memo specifies two YANG modules, IPv4 and IPv6 DHCP options, a Link Layer Discovery Protocol (LLDP) TLV, a URL, an X.509 certificate extension, and a means to sign and verify the descriptions.</t>
            </abstract>
          </front>
          <seriesInfo name="RFC" value="8520"/>
          <seriesInfo name="DOI" value="10.17487/RFC8520"/>
        </reference>
        <reference anchor="I-D.draft-dsmullen-ppd-protocol">
          <front>
            <title>Privacy Preference Declaration Protocol Specification</title>
            <author fullname="Daniel Smullen" initials="D." surname="Smullen">
              <organization>CableLabs</organization>
            </author>
            <author fullname="Brian Scriber" initials="B." surname="Scriber">
              <organization>CableLabs</organization>
            </author>
            <date day="20" month="May" year="2026"/>
            <abstract>
              <t>   This document specifies a participant-facing protocol for Privacy
   Preference Declarations (PPDs) in home networks.  The protocol is
   between a home-side PPD service endpoint and a device-side actor,
   formally the PPD participant, which is a device or a service acting
   on behalf of a device.  It defines baseline operations for endpoint
   metadata confirmation, participant registration, optional participant
   declaration, effective-policy retrieval, policy acknowledgment,
   renewal, and reassociation.  This document complements the PPD
   architecture and taxonomy documents by defining the message and
   sequencing behavior needed for interoperable policy signaling.

              </t>
            </abstract>
          </front>
          <seriesInfo name="Internet-Draft" value="draft-dsmullen-ppd-protocol-02"/>
        </reference>
        <reference anchor="I-D.draft-dsmullen-ppd-taxonomy">
          <front>
            <title>Privacy Preference Declaration Taxonomy</title>
            <author fullname="Daniel Smullen" initials="D." surname="Smullen">
              <organization>CableLabs</organization>
            </author>
            <author fullname="Brian Scriber" initials="B." surname="Scriber">
              <organization>CableLabs</organization>
            </author>
            <date day="20" month="May" year="2026"/>
            <abstract>
              <t>   This document defines the core vocabulary, comparison model, and
   extension discipline used by Privacy Preference Declarations (PPDs)
   to express atomic privacy-relevant dataflows in home networks.  It
   complements the companion PPD architecture and protocol work by
   standardizing the core fields used in participant declarations and
   household policy rules.  The core vocabulary is the mandatory shared
   semantic floor for baseline participant-facing interoperability.
   Richer ecosystem-specific vocabularies remain possible, but
   comparison-relevant non-core terms need explicit relationships to the
   shared core so they remain computable.  Baseline participant-facing
   protocol messages use compact identifiers plus taxonomy context
   rather than requiring full ontology exchange on the wire.

              </t>
            </abstract>
          </front>
          <seriesInfo name="Internet-Draft" value="draft-dsmullen-ppd-taxonomy-04"/>
        </reference>
        <reference anchor="RFC7564">
          <front>
            <title>PRECIS Framework: Preparation, Enforcement, and Comparison of Internationalized Strings in Application Protocols</title>
            <author fullname="P. Saint-Andre" initials="P." surname="Saint-Andre"/>
            <author fullname="M. Blanchet" initials="M." surname="Blanchet"/>
            <date month="May" year="2015"/>
            <abstract>
              <t>Application protocols using Unicode characters in protocol strings need to properly handle such strings in order to enforce internationalization rules for strings placed in various protocol slots (such as addresses and identifiers) and to perform valid comparison operations (e.g., for purposes of authentication or authorization). This document defines a framework enabling application protocols to perform the preparation, enforcement, and comparison of internationalized strings ("PRECIS") in a way that depends on the properties of Unicode characters and thus is agile with respect to versions of Unicode. As a result, this framework provides a more sustainable approach to the handling of internationalized strings than the previous framework, known as Stringprep (RFC 3454). This document obsoletes RFC 3454.</t>
            </abstract>
          </front>
          <seriesInfo name="RFC" value="7564"/>
          <seriesInfo name="DOI" value="10.17487/RFC7564"/>
        </reference>
        <reference anchor="RFC9577">
          <front>
            <title>The Privacy Pass HTTP Authentication Scheme</title>
            <author fullname="T. Pauly" initials="T." surname="Pauly"/>
            <author fullname="S. Valdez" initials="S." surname="Valdez"/>
            <author fullname="C. A. Wood" initials="C. A." surname="Wood"/>
            <date month="June" year="2024"/>
            <abstract>
              <t>This document defines an HTTP authentication scheme for Privacy Pass, a privacy-preserving authentication mechanism used for authorization. The authentication scheme specified in this document can be used by Clients to redeem Privacy Pass tokens with an Origin. It can also be used by Origins to challenge Clients to present Privacy Pass tokens.</t>
            </abstract>
          </front>
          <seriesInfo name="RFC" value="9577"/>
          <seriesInfo name="DOI" value="10.17487/RFC9577"/>
        </reference>
      </references>
    </references>
  </back>
  <!-- ##markdown-source:
H4sIAAAAAAAAA7V925LjxpXtO74Cp/WgS5C0JVsjuTQxdqmrZVWcvp3ulh0T
J84DSCZJuEGABsCqpif0L+db5stmr33JCwCyqiX7ReoiwUTmzp37svYl5/N5
1pd95a7yJ6/b8q5YnfLXrdu41tUrl9+4VVW0RV82db5p2vzHZu/yl66/b9r3
3ZOsWC5bd4efvr65ble7J9mq6N22aU9XeVlvmixbN6u62NPo67bY9PN1tz9W
lavnh8N6XtAvyt6t+mPr5r/9Q1Yf90vXXmVrGuMqWzV15+ru2F3lm6LqXEbv
+V32SV60rrjKr988u6Y/MI9t2xwPV/lf/5z/lf4q623+Z3ySvXcn+np9leXz
/KBLO/ildfh4h+XUuhx8UNa9a+mDvNnk/Y7G4k/X7q4kYvRtUXeHAj8/ZXeu
PtIsP8lz/3780Z8OtNh0IvTxvigrPPIn96HYHyq3WDV7fA4SXOW7vj90V7/5
TfTlb2g4Grrsd8cl0Xfd0ovrbeV+8wAdn9CvKqJf19OvbFz/64UMuCibh8Z5
6PvFrt9XT7KsOPa7pgWN6cV5vqGnZb+f3BR16ar8rQzwhL9u2i19+g/mp6v8
abGs3PNi2fF3Tmj0ZL3Qd/5phe8r+h4EoXeN3/F9WxZ1/nbVlsQ4j3/FEj9b
dPKzP9HghyNt+4J+Sm+Zz+c5/YA2e9Vn2btd2eXExMe9q3tiBPlRl9NrY3Lw
4ejKbV1U2Pddc+zcrqnWU4yX941yVEf8lvIgMR2xzHaXXz6LXf4Znbju80We
v9u5dCauxppphjk9khO79uWqpO3v+b1lt2ruXKvfdq5lznb1+tAQ789yYhz6
edntiNuPXY8J9rui90/4CfZ47+FQlbxLtL6mb1ZNRYRZ07CrY1v2WHWzKSs3
y1vXt6W7c8OfRYRq6LMTvY8mQKud8UDF6n3d3FduvXU0xMqVBz2ZNKPBD4wU
/he8YX45IDt+VeTdwa3KTbkaDpDfFx2d0zWNcUdswtMjimHCERG/y0tig4aG
qxsi15J+33eu2tB8O6ImvaA+seAgHmqO/fDnn3Z5d1x27u9HzG7pdsVd2bQL
4bp9uV5XLiO5cFv3bbM+rrDXWXarQmlOIrGmXXZrYRpjImLgfVNXJ/oHnZtV
n0Nuln1J5CZRWszyk5OJ9E1TdVmyPL8BHbPwsV67FvRYYwn0Ab2SplIxVzMB
9R1QCCQJs2bTO1INbcH0dusZfrE5dvT8DL+nk0SfL/K/7ugxOwyd63vIViJk
7j6UHRGSpnfKNq0QhpZyV7Sn/L5cuwo7RL/YE/XKVcdsQYcJo4paKlZt03We
GPR9pmzdLXB+wTndsSJWxpkIM82bg5PTVFT5vqE3LUi9eWrswf300mLrbN6Z
8b68a+45SYnUzfL7HbF7Trph3bSdngU5YnQU7koQN6c9pwPfZLJtxHYnbASz
9517QHSki10Qy9OWVMT0tBHg7R3tyToSRLwuWsZ7yJqeXt3SFHino42nt4d9
xxJWuyziWZ3x0vEOMuXpHfekSeRZP1H3gQgiu9LJ+ZX3ZXhfCc3GlG/5jd3x
cGjaXjQubwU4MjoR//Vff3zzw9Nvvvr6259/xrbtHY4nFk88nbAh9pWk364m
oVJBQLmCOOrYbkEBL5hIdhNZ9O1VuSdy0ISbDjLTNpaI3J4OMipNn9+0L2t6
WBTKIvsrb3APrUDLrolC9Cf9sWqPq5LeTrxarEnAdgXpmGrmDYp5R3NzzCnu
Q98xw+vu0AxpGkRSlrWxuul53wqS2hvmgB6is20K2AzEujvaSjqTpEJKPq48
3U3V3NPwJaSAe8+L3ULiY2y3X2SmWOjHPHESY8RVOwjnNcQZ8QqPpceJhGdN
h3pJoix8SZzTztduU9ZunbDnXdmVKlii8wjR2BB1wDbJ0aTX01CQ6MZC9Enr
Kn6yb+S1pLJIWJBU27TNXuhmT2+P5ZpFN1FO2OXf/vDN737+eaaM6faHXdGV
/3AYlcQpDv0SMyiWZUX6aUZ8UL/3f2CqdERJzmxK/RAjDzhoQROn7eJtI4Ff
8uBkWNRgceLoAmoUE4eC3dBWdvQEGCraRojvuep8DEQ021/REyAs2QJujV2H
AUxrx/MqI4hf6fx1D8mIGsTLoQdFZPc63/DIPTZ/rOv8oe9jFoAUn9rA1DSC
dK26hv6DU6byAZvFzEJStMaGNfSMt6LWYd++/erb32LfQLeSB6uLtm3uYanQ
cvYNzTeS14vs9iFjTCkzD8uOxCLGJIFL3sF75w6sq/gc5DRNekkl8oZ3h2Xl
kRTBvC23ux4vK6oTbQzoshENCWORpISnWSTGfkysO5LOkNI0D+gN2GSYSbQw
aDvjovy2eUcS6a5smxokBsXdKWj7A8l9DFc1EHvFGoIKM2HFT8ZHvYYOLVr8
i3RES4pMPKwOJ053dwYdd5LdYqGIM0I82MDimzM/sgjbFLDJ9PxDkgiDrsiw
I9bB4D1To3iA8LELQYaNCLz405m3UU/+MMLGgUkCxzC2puY0K4w6NGPBnGa9
ErldJUZTsHL5pNciS1IrkNWkajmI0S4yIcH8B1jgMEo+yg7ek3Yir6Tbd4nc
JokaFltUTe30VMlJWq9x2pW/u9N+T0a07T+mT4O34pUcquYEFiFfGfMJ8kE2
EBsOMw3mVdmybE1Ma7VnsPFiunjhTju/92ZNOjRJVrAJvKdClHewZOhgmbk9
JVOmRQob4AMDI3nlp12Qd7xrpM1IxpZqLLOqpJeXtAc4EYHr6OU7EMzMldh6
TLwMlmRsSZ11TMCJdL51laIzU8KsG2gDPq2kbv4GK+WOVS0krFiBfHh2qtID
89OklwVEC07SAwpUKC1e1SzeTeFuyDXPIEzKy0oDj5ivNxAlth44Jk+b+g5n
Ei4oXnMDPin5b7G135OEAurS5U9e/PT23ZOZ/D9/+Yr//ebZ//np9s2zG/z7
7Y/Xz5/7f2T6xNsfX/30/Cb8K/zy6asXL569vJEf06d58lH25MX1fz6RxT95
9frd7auX18+fjG0qiGAi3dKJXDtAgIHAWaKVvn/6+r///5e/J+30v0g7ffXl
l38gI1T++PbLb35Pf9yTQyNvY2Esf7IXQ9LAFS3LnqoidjmUfQHfgLi72zX3
NZteRM4v/i8o8/+u8n9frg5f/v4/9AMsOPnQaJZ8yDQbfzL6sRBx4qOJ13hq
Jp8PKJ3O9/o/k7+N7tGH//5HOoIun3/57R//IxMeYaFCO0CWai6ChtRVvwuy
k3wpZfwVFD/0jDv0bCzqeRsoDTls68CMkU05S3A72TIWi+q1MUeQgVkdcZrZ
LF66VQHDFNvpzQw+5TqVI8TLivQ23gMZFEt1UxJhPUH807Z/8kn+jghQ1k3V
bE+ZGeRX2VWe3w64Vcxds0XpXZj40GFjTbMrDpBB9+IEeOxBDOYycvfYK7fP
9i42MNiPEb8KjjyozchNAVUws/e0sgePsO8jsvPq3u281XwSScOTPQBpk7HM
KjU/gvfKrzSGZECImW4bRPy9Hy5dBEgiBiW0AtTpGuqMnR/+WFYkq7uPcKXE
oib9Tsom+yliG7+iaHfUgI+9a3GFmMfP2erwzids9AlfatoSFysn8hZEp7FB
qWp54dnsHLgPPPFzXtN1TpL/yMdqbR4HnqD1PQBrMjwk2tHMGuUg1lLrEabH
jDRAKcE5U9Ckzu289SejgzdIhzoWOsG4iizU+F0zb8aJaTDGTG2ILqyjG62j
SyzAgUGJ9chHgpUTrwTe0U9E0XbNsV25AaHlpzRWL8cBbFZj+MPRJCJzSIsR
iPiOfEyxOPSnapWHNSlSehhMivjwhK0Td4J+RF5t0wNXfR1jQTij8JViW9Kj
yniVWK2xbR1esaZfrHqB9JKdSoV59uNg/bb5Sgwz6Ol1ME1EQOlCYzhdWLJ1
Zs+fNycDG9Pbnw1JyNSViILtHMluRps2x5pP2iwnsUYz2AtT01Samt3AaiOI
9ZgaM5WBrmd14GR6o/3zhud4I8dTDXJ28GMx7MClYjraQEeSACOmjwx3/r4v
9zhe5EO04/kF4rBLQKRuqjuGqM/yMBg4PjHRmfaYqjC4CoPoCWUFEYkCMUuo
gg03kNKtSzi+3REHGj7Bko6qi6BYyEag2/EGFX5ECS54r0XkKyZB26oiwCCE
IALkkwtiIEgB+8SErS0YQoblbBs4VsjKR35M+Fjeek9Zj0Hq0YzgDprNtSqS
mKllh2KHF2bthFSUjVaxLoaRuhoRBUYTHkn/YUjnXGiGpvvUxh9MO1qGTl9G
6kvY3g39qDs0tYDcmKfERi9OcyMGUnoimIIQNuSc0ZjEWSvQk9EQGZmm5+4t
lEBqi4ZZBfdwSqWluyDjkhjozhhLjAKEx2063nNT0MfOwR3gZ1rK0m2a1iVT
XLtiDbELvxHm7OVpCkEAqvsgUDwPYgXdfFrP2x4o90dtEovej9mfECIEWnjP
hpOnHR+IuiH5XG8dFs/xqFLkspr1ngzlmlVZs6IFXKLjo8hEq3/JyC1Rarh+
XbWHIFZjbsbEMZnJGQ9AphFJ4F5sYWDFPCvvbElZ3mn095Ls9kPQZruabSIZ
AKhM1XS9YclOQnYs8aLpB5qblqE3lzRdGJ9vRiRR9oZFDqKeM78gtSAYGfzH
hFLuEvHPgPk8IXukbFo3j6PAE6SnCX7fNu9dPWbc4cZ1fdOK4UEyumlF6UDt
y1vZ9yD3oFiXK3qS7QDyURrEviBZa4UMIbgvsoA3CtYxGc2OTKS7TZTRLkPd
pgU2RxjXcQz87OFaRPbWQJtY0EhoI4EqYw3F95gOe2B+7IAAfnJsh5vqosWJ
H5r+7h4CgcFzJ15NZISIG6O5R3OJ49Bi5JeJVSDDlokLxbKFLdtgG4rlz1Ic
BoN9oSEi3pHlURcnaGefKMd463zexusbOIq1RI4H06unJ+/nixeA1HUajHFm
gTStPaXcU500RC2ucZjDgL0/yZ8D2y88NvLMjvALD00EHRuh1Rs2sekNPr5o
JnMkLkc4gvuwK5fkXlfhpQJ83AwC508h6bZHTWpBpsO6vCvXR9ZR6vMycu32
gLjzY13+/Rg8acsjYJCjdUsGx1lK9hamVPoYsGu/1Dg6SzMynFrvMigAL0f2
A3wWKBRiEAR1eyF5RepTMBgoCNYUTBpACWJcfYIUjnC2IIZ+qo8d7xKDCLcW
U6FFvyzuyq3MWCAvoYkErwcJEwz8j2ikRmDBhjqyRLrj3obbwDZuZXgE6Mlm
M2eCwSSGzjqwWJivj/gY+XnFHnvaFbS5Rw5WFdvWOYGiq+KD4krn8GVgUkje
Og3STLALpe2bMqiMtCY+6YR9eFs2pIcwiRCoazkmq2ewuS+APNv7w9b7yBWS
d8rNBiKlDzSQ2Cqyd/auQByPZunhQXZsdWH0LKeLCMgGcKoikktACCxDSqrs
dEkRIeLECOGONwhy489decDb/XlE5iCeoLPy8h2/4vXvXgO9ESiRVHH53uU3
Tf6SpMC7lkRz/hk9+bk8SoNCGfHKxoBPR/9uEI/IP6NBP7dElCj+j4SbjraN
drtYNyERgp9E7MLJWSDfH2JpE/x9zL0+QnMEcIj1svHXwa9g3fBLLRQyMLXn
QyfAh3EQrScOPa2QN0azqizHKImmx1k9B2KXI59tZ9iuBlE5wDaYiEcikzCM
+E3B8v2VAa0LkaWF7PuLn26y7EVRH+kAYvYtSQwkId0wDi178hk99LmFzr/+
6rc//8zxJfhhZKrBgPY22gFqH95hFFSa9808IZqn8SKjkXNe+oqxJXrZPpqL
DwzaLxGNPNZ6cllw1H20ZqTaJFqXrKW28ACjBPCiQGeUI+HTIjzBopSToYCJ
uWSSmUaGTuxhz8YeskLo5P8ysJ/gLtP7agbdhf3l4Md947N4OKNJlZaoKEef
5g2/ku2fENoVbwe7Y9mH6QmyqLu8M5wUz8K2MPFPQ7RLo1FgPRMZf6EtW8Lo
KhWiVsPzBX5KGst/Lz69vHMeYcVEbTKOO4/BYIY3kJmjN5xIfL3+i4ivVzdv
njO+yEzcmtnGOVM10txOFmLRb0KEl78TbvLkKXqyTg+9aCdJluD5MB/qppEj
aKpkfji2yMMilu85NMOkk5yPinyjIx1CpCyQBirITRIXUCI9ncSF2prj74PY
lML/XYhqZX53ZrpAMpVr1mDFh6Zu9ieZo/1QnkFeS+mQrxcCMZnkLrNdsaka
s6hhD+wLViwteS1EtbtoRwNQ7l+c+VDV6MVTgHsaWA/uW52FYzY+BvGZGUJB
El1+FcnutytHVnnZdJoo1mm2X8j6iVGyuyQ9iAgAaWJATrIbmnHtRV9kt4qu
yjt7swhBiEAfSJjFhvUsdfWFrvvyg1vPY2nBXojEiOIwIQ3LqVHe2ILTV5RV
Z/Zj2SOh8CYkiNDoEZKSZdcjN+9vDRAhb5PbGjlusOwZLmpqlgmcURWSaqYA
jZCMvorNdPw6lemeOpzRU1iGiuYomwATwTM7DzOVSGzrFGUp2LJjVJPeROax
Q3YJsIm/H0mEVKMsOh1ukd1MpQ+1EKw+0Y+lg8HwsukxFWO4oJAkHpZCtbIg
w+cTWUWK23SPy5iX/Hrva0XxpQfhnrNJ8t2DWfK82EnyQxPpgUrgIUDKAF+K
YWRoHN4xI5/zar3JJCmwwCaagwjWcVr9IrtWbElCXJpgy/jQbITGwhk/54iX
m0fj0RiGs5qJL0sOVk7IpbP0EhMyDd6UemgDHCn4ke3GwCv/xBTrT4e1wVUJ
VCb5EkNRaiH8QWHBcH0zxfS6C4so9xwu6SXNns4Vi4NJlGwKF584O7U6aOdB
uSMdygqeWGD4ERtziJ+Jsr6AUalmUHmdl1Ulvq4v9hi9G1ZyqQlw5ytBMB3I
CE7YKkfxlM5y9ZK9egTQmDNCwULCn44p3u7AlNNo+JTPkODrP3jWe/bhUKra
eCOIt7DTYyQLCHWsPSa8HFeysJ2VSSQYVtSvh895bjGxyGxF0r+gvu6OacKS
tG5S4FsQ/NaOazzLAEvviuk98exoMC1zX+bj2xYtYGm01mSjmO0kQX17lG0f
Y9QqPc8eiIwxftV6SvLpfY4C8flb5uen/DhMgXjReu7FheCsTLMKC97wj1Aw
IYq6jrJF5GjOkNxm6S/yThZN4r/IgdMNemQU4oKkSiTtefz8XOyHK5KCV6zh
AUu5TXmm33Gu7EXhRDxRnR5FwyG7wA/rxnZp5HAPgY8J79Q0La3+GGSJYOwD
wjMbywu7Ym8ouAEorEo4UQvFOFH5DZAINmVjplP0Pf+LN2nBeomtCW+NDU1O
6Rsm90hZjcKWkv6gIMwI7V4MdB82VyKGiIjJKiJw1xIxOk1IsSMZ4zJp6k8y
LwkjPjIuJ3bXCM3n+rkU+O8e1FF1U8+nICUtFqgVPBzUq5EYqcx2ZaARdpsm
J0tAKsl65PmOSObRVV934X3nFcK9o1Cb8lSSuRO4FP6t5kOKO/dnVHswJz2r
d6xVGPt+KsBqln2Rv9UisCLJyjGRcjl1Lk5/GJSyhFy4ykP9Z2h8LnvxC5qz
loalud1KIEZEe/Buu6fl/0OcKjijwLdxWOsVcuYZ0EntNNkHW9JdQEKaNuQk
m34Ap2w5X6rh0r2oCmiUMohZvzYwsz6bWB4QTUO3zhjMKoem62IHmUhf5E+V
ZeIU/EHkUKAnH0ztR+dbkk1ljKkaAgRxHom0jpMNZ7HJdxmFBRnZxbkN1Ykq
7ogxfIk0h0BRZrMmBlgHXy7J32f1v4Z/0bkES+2S1LlJDzwqnE5qTKIiiHIi
H3EiBzGctThRNE7jYB2T2l+iDOwE7aCBG6BlRLthyiYfcp7dD4ibBWpdV5Lm
7UtU2KpmNM7nhTUtAG7MwGvR+cg2kHNTnYIDWgWPkyWTxZK8bJrAriQ8v2KO
fe5wiNumsQ2zOUYxCMPFtJK909xhX/TFAEBAkDC5AGGPKow708eamx3ylzWF
bZ6ksEUyS/2vRHWqZuro624jljUXqx737JMD/FjxKZG0ywSUWhaq5H8gulQl
mwxxwnYWy5JIeDxe74YQ9VkNLLYy72fEeHMN16TPik92RldiN2/3HKCK4TYi
ZmNFP4+qZ74sCINVbGyHqOWJDUtJ97Iw86gMEOrwLQoGxh0lBGolDVhud3Oy
kl01kNxpLeJj8rgfADlv+7gulWHpmcJRWjbIQiXEEDQqOhtWK4r5eZDtl1SB
9w+3v0jUyCPLhAO9TAHTaZJIdROEuq8xLxQqZjw1ODHkTXIoTRJ26AsJ21s0
Ck9axV5RbYFp7VD9pmHiPEZqpRL4h2MLBQowdZZWbsQTZRMmSqrgw0vyoNyc
hpxCVHfbolJpvCW7gNOIWGavle4yI7/ouNmGRFhRpe5aN5wQaZojR9kGla+a
JB27F2Xni0FB3wlHJARE64v9PM4YDxN1AKOkJRi3o5nBSjhMyaUCyclIyiR/
PlR0k3DtZyKUOzR58dW7hmlc2rL4UHtlyKXFkDNRHw/JOjFnr+hD9W3ZaUyU
z8vV0O0TWAFhTh8EXVtlfpZ2RnhcHfO0roYsyCIVbAWd8QLpa4c8WT43Y0Vo
2RBql2SKXrNSE7mQABCGC5QDXbV0IjPgkdclziH0Q5+N6aIRveytzWB0eIPb
6g+6FN1KqxCJlsVFYRnHFou2LQMUFWJvSeWuD8QZAeTkXaErxe38ZjHRHcl+
//PPHOM7/6CN/fPPoheuYxjg1R0kn7s3XM/E61AOSksPrGDTwMLSdEx7/Ar6
+43TipKnZq2gQPj10B1PNAXo0dRsykg9SJYnxo70Q1hpxscBFfoz2hqSgAJL
L2nE+3Ld7xajzkQ0EqdpcTryprhDh5QK8dV7h/9OWWt8BKSUiVvIcPm8iEzM
69fZPyKxs3zQPond7Y7TrPCrgRFmwZzpogMaLA7x3TvapdrMUtWibB9MDM02
eokuXeImS2jgKn828CC504KeISlklf4eWnhSgzIXwrDsUCFPbJG/QPaS1Pib
w6qmv996j9uhB08az+vibg1kDHWWinCmqAcetWJ5aUyTS/Wv0lxAH2T0jBnp
T9p3mppG904PRTF9F5G4zt7LehAr+HM+JpgPgPKhH+CbdPhiTyl+BzdFrCeo
2zBSGNL8ANP4IIAamaH9ABbZDWrEJoIwn8YNQbTBzDgu8NYnRPsIgcqUhP05
z787H6TVZI5QqkbHgAxnmlbcMmgceVkMIOokxho9zAUnl2NDRea10uPw16ni
EYGCu1C8LXillmpFMpVpgk5/xMB90mBsjNJ/RNlJckiTwpMikYTfsTLRV09U
f/ySYhR695kChtvNo0MDSYTIB4h84DPzwR+HMBQ6CYng15mNg7oaORzFTxaP
mZVGPGYI/16sgcgeroEQ+5oGSsogymEVBEcGs4vhB1URZ9daZGeDQpxK4UrW
JchmGY/xYGQLBQORSpXYNMuhDxYgijPIjGUUIZR0/QJAmm7uKbPazzMSIGTL
xfH4qDzoMcGd6Yy7MIqiBtgj4jGVwwkBNdU80TGiTOfXPlXhe7XQp2RgSOsK
WoNLBB/ReiZjuE+sYy/jpHtPKgChYitXtPUvycjJHsjICcrysvZjMzJDOWZ4
caJvFtn3Wj5m9V/FOFo+0FETrMqbBMlzSaXCL/wo3ZlJekkfacl/iFE0yGMx
pTmlJgfBvyFOaHm1090y2VkUWw4xbhJ7ltoySI3B7sepUIW1b1Lwc+YFWbNf
lnWUTzaUV1kkr8gpLWtFyX3+E3daPK+/6/VocqzkrM2a7F6Ya6abp0RtAUX3
4Fyy2aAT/ZZpD4T+Yg1zkhk8YJPvaDRojW2rQ2leadA4ls/VpngRMMOQcbMn
ZxC4DoYrEN4viKLbI/zgZSkFB0s6JpZi4TnsLLlEY46IxsYKIjBaRBwZbr49
kea+cyQv8W6Ra4MmtJZzOOGGJp5n1BHkMR7o7KEUTu87zzy97FRFvGfQ0tyO
aNLhMRMQIWJrD5KrHbRrOKhqbcJCfCTJdPzf7kTGv7md2mXDF8hc5deSJ7lB
5cqa85QMhvvMLbYLWkGzxERItszI5Vrm3D+t+ly6s3m5xZ02tJJpuu4fHPBW
OcB8lauLPSfytH3EKNptcjfEccaxmoutIx6VT8nZsa0V3v0zcijPqap/dsKo
dyHV7ZhIqxypFzHZpRnB5ZRK4L4rqUGx0klvAFz90v4bw+YbYLIHM1oe22tD
U3pHYgiWpSTwepUmSUUPp2D67EsLs3HLQs4a12TWkJSZxvQ+qu1HqANxIZTy
i/t/hLYWum83oQPHv7b9xvnTONl/I06OSfpzg1Wecgf7Pn8DmLgD9vbqYJVJ
sW0r3OTzGUBWFEBwdGLuE3XW8fDMMjxqvtK3aDWb5m9I8x3QIkRDVSm903o/
fmus+oWpPIzG0xSfWEmhQkzjE5PVwRZ7iEtXhBmd9DQt2k7VXVt27Mh2VouX
ZBcmxDSPS1Wjwes+dsFDoKGY1kiKCROiYryOGJlT/hsdXMxisuhZGvlZx+Vo
/t1RdBHA2bLY1iR9yxXLqO+skykXS/ttqN226cvgcNYQ4mrqcBWlQnAhMVxL
3Jr7mtQgagwPPa+IC+sT8J925UUIzYba3itizjGmP6jzVetlWNjPom/cWeBI
rsWIfECuuzPZUT503aS2+CJ7i/BcNBmTNVqs5kMvKnsk0D6UIFxIlkoReIEw
bH5AS+RBGQnJyMof+ij0GiHOnbcTuUiUS36SvtjnaxlCic3YERCFYK0P3JZ7
eEiblSjPSM23KUMl5/7DWRxShTqxtCDNrTkdnPR55R7Wom7i5m6owOWGZ/Al
19o1L37Y90vjaGPSJQ0vgEVXy+GK+7dlkusLAbpgHHs+jJqKHBV6MDI+bHwm
cDLNPeMWe9w4sdBMLDEc3TDKI/L278eiYvcg3xT7kntQHH1fbSKYCbG49O87
nUDhC7Xw/GOjN684544TnwzliQPd3M7O96dii2yshYScRPRsbBCo1Xqy/oTB
XbyQKMPQCNBM7bovVW4Plwlx0/CJMqHs11l9Zl1mURfE/FKRThXBIIZfGF7h
71vIzqL2CQwd4QmT1mU2Qk4vIBQXSnTUN/+IwpzQaON2KrTCDDsI+owbWYWE
CW9zaUJBrOwwljCWtedgEj4C/OCIsGrfqb3/js3hjsjQZ8GpDmjEWTFogXGt
CxpgDiGBoVFjQUOSl4AKj0/o5kS4RCb9niNo4SJKMbuAPMwyaaowRBtiZCOq
9xph7En4NcPRbtQm5LRRDWcPbZ+ScfGtEw6fLEXKBoZsmKJH4lmspB2gPDQm
NhRtfjYBRQOBT/FYf7eAmt25qUzO5cY8+t5p785ET2dWm1T2KtgEldYOlq7u
vWELoYHbDGTRqjxA/ZCLsREI0J5k/hDdoe1Zubg5/+EouG4Sy0cz3mMFUaB5
vqIKeQqJFSAVyQ4BY76qoeX96p1VjSObUksbSCeVByTtEFdl6A9P/9oe9RoK
hOa5O8hUOErtxLLWfAXnwx7Dmps81NxErb0mtlsDLxcqLqNg36CjV1RzU/7C
mNCwPKrsgvPI9s7HVXBYYxZ4wtyv69f0xsp+YW+sIaUwh73LfETioc5ZWk12
YcGZxowW+S8pLkv7E2Zpf8LJOOQlKYh0jlOmZVE2QvTTUWWU7GsczA6MkUWM
8ZGQ56OSbh6EPDV/KPP5Q9512KP/Se0QhpMew2S+NWuf8IjxAJylNV0+zsnF
/3fuhD0bx5IsThul07FeiFO4skGDhpAnIgowlkZkLowT1bjvs2TFeNmIeVlw
1vpuIWDR6tMhH1tj1eaExMdkkGFT5NsSIsHD2+RhvWzqOfteNII153plPp/0
e+qywTUY/goLM2unKoJ8Midk1yxKILO6UCLl4BmcrTZ0j+ZXmDZOEciox5cW
Hk4nQXRhbqg5nirFkV4Dsb3lPsAFkm835K4gtRS9Hs8nT7MYIKXfdUho4XTV
6LWXC43AxmtpDBUa0YVsy2GfrqjPFYAMeacdOS4f0jzqqIV1XElEzIQu9lFD
nNDzzcJuca1m+ZG9015OdGG7UJY0C19yMnaQtTqtQee7cS01V6SlY0zc4Saj
fdrFOd3XZPvKlsoR6px0BRuUgIX8Y3qDpzWS/3GVAtJ7kZxaFaeoeNIHcgOv
FAp0ll0KrlkV13QOFuv9M9evIPlQ8dXn2uBERLNHiDRO4yHU5PYN3xVlbHaJ
VUlC32T4Im5KfJe2j/G5llF/Nt/yn9VYT1qN3EB1qYObfiEWFtz0pMqCPCHW
Owjseps+7kyhDe6jvioBaUCDFfK8P/S4Xq+pM1xCSediFbdU2ZN+tFZrFimL
7i2YUntZovZEqtp+MHZsmBfCrT/KVUSqqa5k/+A8qzW7RMlZVyY91uRmtZP1
e/sifzFQeJPDmFZUHEh6zAh00+wLn/kYitU4YhgaSX2RPxNK8QsYtLdFhZds
uFioCq0xG85lIvXAeIO7a6q7OGCnaXcgNi58Q+Si1PIP8SZF7Op1bfPQhDJa
YhFd84VQMu9OEJN79kvwoviauj28IywMffeSih8rTLl++/T2FjoXqpoJnb1r
zLuIiq4mqBbByThCPYCxuch1PTJO1FwkS1S8MQ/OiLRLTsRGJEZv4I0ZXIYM
Rb8BstOt4Fu7AoiAi9Wc4KPpacONfF//2+8BfrGtU1ohGI5zcnmVJPJFZNYq
CjPHZMbiXBmAJt1VZ9LXvV+xnpB8drtREVDkiehoVRSiiLlkOJRdsUyW6xL8
dnibjoPO5hFKt73pipx+OiVdinBIU+XTRmx2IX3c7LUzFmp6d5Q0ePLSR2+o
fLczmxK/2MgypLMaT1nTq/kawr+RaKr5xptCyJWNm1NF1QnS2RL4SSzLRR4N
QllMuI3B6nFk3zeDMvvFvPliLSBzUfnbV3zYigEUxK6yQewqUWWhj1Qyv2nv
PhPG5hSVG6E6+/Ahmb23pJY7ycgJnnueeO5+oliupRysC1zchw5jbd19vuCX
6DUxkxiCmEptfId0P+NmeFJCMgkt8HGBaeFWvReppPdbN6xHsjJU5vnr17fo
80AeRWHXDfvUnbi/5a7hWL8/yEV0byx4RsM4mv2zLCr25PjNoZHozERzNr79
c1COtYFiksoca4I5DF49Z2sqhLCmsvPOOz+4HrXnBgqDAlmJoN2jBIZOHcxx
CXUaKtlKw2W7UwLqlpvirFHhFU5YBMQzX01GcyU2plCe77cXNz6Ys5U2IE7a
IvSLAW0GkTpOMK4kr8WjFdzCaNhkfy+AGsaSImAGJnRK4m2NSzZ9iqVngEGG
6ni9cgIA1tBs4tL10qW3mR1rqwHbO9fzDVnJbaLR8ZQxfdhy5DwJcHrGvO6k
mn5gYQvfpQlGOLmk/ESQxRoqHC+7ZUXg65nW7hn+h6DnZKHjUFIZMymU6W8v
BV9p+Rfz1U/pEbPn/X2zo3DtyGpmUVcZM3C1DncQ/0JbOOoAzZYbNt45yznA
7EK4v2RDmX8VKkWkYsg6cqJ4dFwkpvylslKkHe3dXbcY3i3HxaefW8e7tNY+
f+c6n7r1xle/3SYHhxTQ8IhyY674nN4gstscrDg3FNKlZ9C7OZzGpbp4diG+
q1GCENQIkQpvkcbr6W09oxurzb7hPgqN+PGYYGSHST5etZmvcOms99g99taJ
P3d7/fKatXUoFx0aOHpxIz9pqFKmt46PTOjRWLe1v8RYa+GnWB/+n9lAZrvi
OOJqmOSQNaMrJ9VWjUzHWWw8+G1eujxKRooVvY5grTkm7ez2yFXPifG6sQYc
gQYYm/OEQ/+EyKz2Vn3oKG6NHuIrwNUe1c1laorBzdvBRndsWsuVs2aq/1SX
ZLi6eXHP99RFBrVebsrdCS3AFepFpV0+ys6tPni4kz800XWtUZW5FJV5XJhD
Nd4MMH+IA2jJ5WAsHYZ3m+EYScOpYZR/LR3O2R7alzyKj6+iPGSsxxZnMji6
YuPILWvXphzrjSU9ahFwiATOQhhQ23Va/C6+5tsmxz33DerTbNW3QnDLS6OZ
u70mbGfZ06QbskVEJ5sgnQ3KCKmVcMy5WyDIPVRnSHIXqa74yR6qAgbir879
jnFhlqPmB3sLMhrAp32n4VXfWz8qhI7bcqaVfL+yjk/jZt0RTd1Lp7Ecnxtg
Jepxk9HpUoEhMPgvzsumswPbdLLqNa7ol3xqY5i0vt3uB5Iia3eKNovLerxr
dD5F234RNsGStVft6dCTgVAcdkmqWDC36BzjGROu+yNfixmlXPCR7hrZhIps
QemVn54FbfsjTW5GhrtYzSxC2OjyIJncwT61sAHbmvxxk9n/aZ3wI+p7hhkN
WhwWcv+18rJu6tPezsYLEzKv/anOsmcAyRklGa0aKFSH/ha+Wb52VYs8NH8X
QQT7a4BypWaOjwUmHSyTIh6h/p1DHMOgEgiDVRdh39HdoXFD75Bi5vw9lEdU
qnl7cKgS9D4qAdAlvZQdaL1jAUXJmWrCr77+ljThivZHYvAqA/0pFJvYWu1y
fikJyNoxVMlCRkg+i2+u4gN03B8rzVDayHqSwAn84IhXGWbkVON+h4HsIoqg
LbUQH2kIRVsrz5REsup0FdQLT5HPot7sMZVRa5cJ1MumkKSk4H4701Dqitw1
Jf3jKG80Dkq760t3FEn/Crnt9iy5CO+5pflT8xrYJxb6/+Hrb74h+sPO2rnq
4HvKYe3S3H2RPhldwi0kA+oTp/qpZvWdc9SQMTHh4BcKa64qk+MSiBoVeNkY
wrjSyKyRx6ACUQuSzq1jX1Fox55P545rHFAUBaFiJBVa/sqP/HCkSa3UKHS9
VR0d95LM7JAmpVYi3x9F1sAxxBNKiLdGrLZiaGwWLVQj8Swt4L36xO9DRyIl
y8qAcuKiQxQ/jl0UaXAP2se1PIoQmHBW3onUnO2AKZOUMX2N6CL7sbl3XMPC
uy8cuSKzASBSaKdS+DwS3xGImC3ZB5VlcgXMEHFGr0ApJGDnQgou2XKeicPg
uxykYcyoi/KtmSBZpoFn0QOYUxtuwYykd7jDUDmMmUDAN1pZ77inrB83sspi
JL7c4oZviX1oHyFOYfbCqG/iLCOJw3yaWMMhui3WnGX0ISQBC7mUxnec685Z
kBIm9BFGODorCB1I/6YVSbuHrRRfsMQ5RRGMutTM1NDU4oIG1IoTXywwzmQt
p+iU6uNVOFpnU+x8mc0js+zi4sdwyRSX+oQzHVgibSCmofGxFyScUw9FgxTU
a+KVQiuqW851NbZmSp3H6mKnAVThO79msYgbuQuQwp3eVmQ5xVrvpEsY5jfa
Bd96UqLbbtvo8qHO55n9c8pFw3Z31ns4ch6GiZkR2CSG1HQW5XmuXGTfqyix
rrgaZJXrOPK2KLskeCV2hTgsJCTJl1U+gIie0IACcfCFLUQ+hub71AAIqZBq
a8om8cWSSEDhMcH2ej+5JuGzeIdHwwST6w3LSgIHhk9JWyVB6PwRIDly58h2
l94W2jafO/vKNb0kKtg46KKa7WCNduoNgSSWBvxO5S+HkIKLH9/0NuwZJmRJ
kNth5ruEghXStW6ss6S7bVBBlq2DbmGG5nrxaTmqppNCtF1710jK+rjxX6F3
e8aXpOJWsQYnKFEd1+MLcxMEfnQNsakWY/G4ojNJ9T7fuXPUInoA24couRBf
rycgTjKSmW+VGOlSQ2bJwGfbH/oe85ZKGLu92n41WJ+leT1xSjrvnhTSXe6L
EjYMLMpBXA57rZxX90lT2AEhrIfcgiGrJFesRRNQ2DjD3PiZZjXzOiKZNGhS
ab7g4I0aekyS5icCKz5C+fD90lprKDKLd1vnhtibD5pbTmtpqU7dxX66UgR2
y1d0nfxGDusIItaQsCGSyt6rGxGxbbjKTMDL+LJO3y9PetMBPxYwFTeTAnBt
4raPcw0EVM7a00WW1+V7tJL5ks7CZZjpqSGZVw/89gR/mqscdmfiYIg5IZKb
UgZyD5FE6b8T3749se923YAV73BrcpWaU40IJajIcSLnvf5BfsE4sBkSbaI+
ff9iWCoU5A5kkaYhBCgqAigjGGrY8DlAfb8Qmxo3D/Anf8/l0P7KPsV+fae6
pLLd94klfak94jTfyyqCQgpD1PCCp8qhhTNVRRzpTLBHLyoHpPBFLnbIhlJD
uoKiVWbJrTBGBrK372KxPyV2x3fCR00wqhMGD6Z1EMZsVjUeMsWpGKwheCl8
odqJfStiabxJwIWJvHS8znSWnIaIYP6KrsmTSjTzGbIMzISjHenYiQv9Jg6t
NdLyGlt7jTccarWQxvQ8RkX13KsnCA9vm3H6rzdcOOd3upxYYOtZqGeb5XK9
u2tl/gy+RS1GRx2Fp5NPBx2SpQxPfEUtLQxptRagHieNhoTRuGXHxqoYR1lE
tND5fC4BNy594T89rEYf/A8MDu0SZ54AAA==

-->

</rfc>
