<?xml version="1.0" encoding="UTF-8"?>
<rfc
  xmlns:xi="http://www.w3.org/2001/XInclude"
  version="3"
  category="exp"
  ipr="trust200902"
  submissionType="IETF"
  docName="draft-godoy-fischer-mafp-00"
  xml:lang="en">

  <front>
    <title>Microphone Access Fairness Protocol (MAFP)</title>

    <author fullname="Henri Alves de Godoy" initials="H." surname="Godoy">
      <organization>State University of Campinas (UNICAMP)</organization>
      <address>
        <postal>
          <country>Brazil</country>
        </postal>
        <email>henri@unicamp.br</email>
      </address>
    </author>

    <author fullname="Douglas Fernando Fischer" initials="D." surname="Fischer">
      <address>
        <email>fischerdouglas@gmail.com@</email>
      </address>
    </author>

    <date year="2026" month="January" day="23"/>

    <abstract>
      <t>
        This document specifies the Microphone Access Fairness Protocol
        (MAFP), an Experimental protocol intended to improve fairness in
        access to microphones during technical events, forums, panels,
        and other interactive sessions. The protocol documents commonly
        observed behaviors, informal control mechanisms, and failure
        modes associated with microphone access in both in-person and
        remote participation environments.
      </t>
    </abstract>
  </front>

  <middle>

    <section anchor="introduction">
      <name>Introduction</name>
      <t>
        In technical meetings, the microphone represents a shared and
        finite resource. Access to this resource is typically governed by
        informal human-driven procedures that lack formal specification,
        predictability, and reproducibility.
      </t>
      <t>
        Despite advances in networking, distributed systems, and
        resource allocation, microphone access control continues to rely
        on ad-hoc moderation techniques. These techniques frequently
        result in unfair allocation, participant frustration, and
        inefficient use of limited session time.
      </t>
      <t>
        This document does not attempt to fully resolve these challenges.
        Instead, it provides a structured description of commonly
        observed behaviors and proposes a lightweight protocol framework
        to improve perceived fairness.
      </t>
    </section>

    <section anchor="terminology">
      <name>Terminology</name>
      <t>
        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 RFC 2119,
        except when such interpretation is overridden by real-time
        moderation decisions.
      </t>
      <t>
        A Microphone Holder (MH) is the participant currently in
        possession of the microphone. A Waiting Speaker (WS) is a
        participant who has indicated a desire to speak but has not yet
        been granted access. The Moderator (MOD) is the entity
        responsible for microphone arbitration.
      </t>
      <t>
        A Question That Is Actually a Comment (QTAC) refers to an
        utterance that presents itself as a question while containing no
        interrogative intent. A Last Question Promise (LQP) is a
        statement indicating that no further microphone access will be
        granted, without a guarantee of enforcement.
      </t>
    </section>

    <section anchor="problem-statement">
      <name>Problem Statement</name>
      <t>
        Operational experience shows that microphone access frequently
        exhibits disproportionate usage by a small subset of
        participants, starvation of participants with concise or
        well-formed questions, non-deterministic moderation decisions,
        and inconsistent enforcement of time constraints.
      </t>
      <t>
        In the absence of explicit policy, microphone access effectively
        operates under a best-effort emotional fairness model, which does
        not scale well with audience size or remote participation.
      </t>
    </section>

    <section anchor="design-goals">
      <name>Design Goals</name>
      <t>
        The MAFP is designed to improve perceived fairness in microphone
        access, limit prolonged monopolization of shared audio
        resources, provide minimal access guarantees, and allow graceful
        degradation under time pressure.
      </t>
      <t>
        The protocol assumes deployment in environments where consensus
        is imperfect, time is finite, and human behavior remains
        unpredictable.
      </t>
    </section>

    <section anchor="architecture-overview">
      <name>Architecture Overview</name>
      <t>
        The protocol assumes the existence of a Microphone Resource, a
        Human Queue that may or may not be ordered, a Moderator Control
        Plane, and an Audience Data Plane. The system is inherently
        stateful, although state transitions are rarely documented and
        often only observable after microphone allocation has occurred.
      </t>
    </section>

    <section anchor="microphone-access-policies">
      <name>Microphone Access Policies</name>
      <t>
        Speakers should be allocated time proportional to estimated
        question length, prior speaking history, and observable audience
        reaction. Implementations may also consider queue position,
        remote participation latency, and the likelihood that a request
        is a QTAC.
      </t>
      <t>
        Speech that exceeds a reasonable duration without a clear
        interrogative structure may be reclassified as a QTAC and subject
        to early termination. Detection mechanisms should be conservative
        to avoid misclassifying novel interrogative forms.
      </t>
    </section>

    <section anchor="moderator-behavior">
      <name>Moderator Behavior Considerations</name>
      <t>
        Moderators must appear neutral and should make a reasonable
        attempt to honor request order. Moderators may forget all
        previous state at any time.
      </t>
      <t>
        Decisions made by the moderator are considered non-reproducible
        events and should not be appealed during the same session. A
        moderator declaring a Last Question Promise should either
        enforce it or clearly label it as aspirational.
      </t>
    </section>

    <section anchor="failure-scenarios">
      <name>Failure Scenarios</name>
      <t>
        Common failure scenarios include deferred handling via offline
        discussion, declaration of a final question followed by
        additional allocations, microphone failure during transitions,
        queue reordering based on physical proximity, and starvation of
        remote participants due to unmodeled latency.
      </t>
      <t>
        In such cases, fallback mechanisms such as chat submission or
        hallway discussion may be invoked.
      </t>
    </section>

    <section anchor="operational-considerations">
      <name>Operational Considerations</name>
      <t>
        A speaker <bcp14>MUST NOT</bcp14> monopolize the microphone under the
        assumption that silence implies consent.
      </t>
    </section>

    <section anchor="security-considerations">
      <name>Security Considerations</name>
      <t>
        The MAFP does not protect against intentional or accidental
        misuse of the microphone. Attack vectors include excessive
        verbosity, conversational hijacking, comment injection, and
        social engineering techniques such as appeals to urgency or
        authority.
      </t>
    </section>

    <section anchor="iana-considerations">
      <name>IANA Considerations</name>
      <t>
        This document makes no request of IANA.
      </t>
    </section>

    <section anchor="acknowledgements">
      <name>Acknowledgements</name>
      <t>
        The authors acknowledge moderators operating under severe time
        constraints, participants who consistently fail to obtain
        microphone access, and audio hardware that fails at critical
        moments.
      </t>
    </section>

    <section anchor="call-for-contributions">
      <name>Call for Contributions</name>
      <t>
        This document is open for community contribution. Contributors
        are encouraged to submit text that is both technically plausible
        and operationally recognizable.
      </t>
    </section>

  </middle>


</rfc>
