Skip to main content

IPv7: Identity-Centric Network Protocol for Security, Proxy Mitigation, and Operability
draft-subbiah-ipv7-00

Document Type Active Internet-Draft (individual)
Author Arunkumar Subbiah
Last updated 2026-04-25
RFC stream (None)
Intended RFC status (None)
Formats
Stream Stream state (No stream defined)
Consensus boilerplate Unknown
RFC Editor Note (None)
IESG IESG state I-D Exists
Telechat date (None)
Responsible AD (None)
Send notices to (None)
draft-subbiah-ipv7-00
IETF                                                          A. Subbiah
Internet-Draft                                               Independent
Intended status: Standards Track                              April 2026
Expires: 27 October 2026

IPv7: Identity-Centric Network Protocol for Security, Proxy Mitigation,
                            and Operability
                         draft-subbiah-ipv7-00

Abstract

   This document specifies a network-layer protocol, IPv7, that extends
   the Internet Protocol model with an identity-carrying address form
   and an origin-validation mechanism intended to mitigate abuse of
   residential proxy infrastructure.  IPv7 replaces purely numerical
   source addressing with a hierarchical identity string and a Variable-
   Length Identity Block (VLIB) that carries an Ephemeral Identity Token
   (EIT), provider and tenant identifiers, role/policy signalling, and
   an Origin Signature verifiable by the originating provider.  The
   protocol enables routers to apply policy and reputation signals at
   the network layer while limiting disclosure of a subscriber's long-
   term identity to intermediate systems.  This document addresses
   growing security challenges in Internet-connected devices (IoT),
   including smart TVs, appliances, and other residential endpoints that
   are vulnerable to residential proxy exploitation and botnet
   infection.

Status of This Memo

   This Internet-Draft is submitted in full conformance with the
   provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF).  Note that other groups may also distribute
   working documents as Internet-Drafts.  The list of current Internet-
   Drafts is at https://datatracker.ietf.org/drafts/current/.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   This Internet-Draft will expire on 3 October 2026.

Subbiah                  Expires 27 October 2026                [Page 1]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

Copyright Notice

   Copyright (c) 2026 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents (https://trustee.ietf.org/
   license-info) in effect on the date of publication of this document.
   Please review these documents carefully, as they describe your rights
   and restrictions with respect to this document.  Code Components
   extracted from this document must include Revised BSD License text as
   described in Section 4.e of the Trust Legal Provisions and are
   provided without warranty as described in the Revised BSD License.

Table of Contents

   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   4
     1.1.  Conventions and Terminology . . . . . . . . . . . . . . .   5
     1.2.  Key Design Pillars  . . . . . . . . . . . . . . . . . . .   5
     1.3.  Goals and Capabilities  . . . . . . . . . . . . . . . . .   6
   2.  Non-Goals . . . . . . . . . . . . . . . . . . . . . . . . . .   6
   3.  Problems with IPv4 and IPv6 . . . . . . . . . . . . . . . . .   7
     3.1.  Identity Masking via Residential Proxies  . . . . . . . .   7
     3.2.  Stateless Security (Dumb Headers) . . . . . . . . . . . .   7
     3.3.  Complexity and Human Error  . . . . . . . . . . . . . . .   7
     3.4.  Lack of Native Trust Tiers  . . . . . . . . . . . . . . .   8
   4.  IPv7 Solutions  . . . . . . . . . . . . . . . . . . . . . . .   8
     4.1.  Identity-Centric Addressing . . . . . . . . . . . . . . .   8
     4.2.  Eliminating Residential Proxies via Source-Provider
           Validation  . . . . . . . . . . . . . . . . . . . . . . .   8
     4.3.  Built-In Reputation and Hardware-Level Filtering  . . . .   8
     4.4.  Granular Policy Enforcement . . . . . . . . . . . . . . .   9
     4.5.  Human-Readable Auditing . . . . . . . . . . . . . . . . .   9
   5.  Darknet Diaries Case Studies: IPv7 as the Fix . . . . . . . .   9
     5.1.  Episode 172: SuperBox . . . . . . . . . . . . . . . . . .   9
     5.2.  Episode 128: Gollumfun (Part 1) (Fraud and Identity
           Abuse)  . . . . . . . . . . . . . . . . . . . . . . . . .   9
     5.3.  Episode 110: Spam Botnets . . . . . . . . . . . . . . . .  10
     5.4.  Episodes 45-46: Xbox Underground (Credential Reuse and
           Lateral Movement) . . . . . . . . . . . . . . . . . . . .  10
   6.  Additional Motivation from Public Incident Reporting  . . . .  10
     6.1.  Botnets Monetised as Residential Proxy Infrastructure . .  10
     6.2.  Credential Abuse and the Limits of Source Addressing  . .  11
   7.  Technical Specification . . . . . . . . . . . . . . . . . . .  11
     7.1.  IPv7 Packet Header Format . . . . . . . . . . . . . . . .  11
       7.1.1.  Fixed Header Section (40 bytes) . . . . . . . . . . .  11
       7.1.2.  Variable-Length Identity Block (VLIB) . . . . . . . .  11
     7.2.  How Routers Process IPv7 Headers  . . . . . . . . . . . .  11

Subbiah                  Expires 27 October 2026                [Page 2]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

       7.2.1.  Three-Stage Processing  . . . . . . . . . . . . . . .  11
   8.  Operational Considerations  . . . . . . . . . . . . . . . . .  12
     8.1.  Key Management and Rollover . . . . . . . . . . . . . . .  12
     8.2.  First-Hop Deployment Model  . . . . . . . . . . . . . . .  12
     8.3.  Telemetry and Troubleshooting . . . . . . . . . . . . . .  12
     8.4.  Policy and Misconfiguration Risk  . . . . . . . . . . . .  12
     8.5.  Interconnection Considerations  . . . . . . . . . . . . .  13
     8.6.  Manageability Considerations  . . . . . . . . . . . . . .  13
   9.  Deployment and Transition Considerations  . . . . . . . . . .  13
     9.1.  Incremental Deployment Models . . . . . . . . . . . . . .  13
     9.2.  Coexistence and Negotiation . . . . . . . . . . . . . . .  13
     9.3.  Naming and Discovery Considerations . . . . . . . . . . .  13
     9.4.  Middleboxes, Firewalls, and Translation . . . . . . . . .  13
   10. Privacy Model: Hybrid Anonymity . . . . . . . . . . . . . . .  14
     10.1.  Ephemeral Identity Tokens (EIT)  . . . . . . . . . . . .  14
     10.2.  ISP-Level Verification . . . . . . . . . . . . . . . . .  14
     10.3.  Reputation without Identification  . . . . . . . . . . .  14
     10.4.  Optional Selective Disclosure  . . . . . . . . . . . . .  14
   11. Advanced Security Features  . . . . . . . . . . . . . . . . .  14
     11.1.  Quantum-Resistant Cryptography . . . . . . . . . . . . .  15
     11.2.  Multi-Layer Authentication . . . . . . . . . . . . . . .  15
     11.3.  Built-In DDoS Mitigation . . . . . . . . . . . . . . . .  15
     11.4.  Reputation-Based Filtering . . . . . . . . . . . . . . .  15
   12. Advanced Routing Features . . . . . . . . . . . . . . . . . .  15
     12.1.  Trust-Aware Path Selection . . . . . . . . . . . . . . .  15
     12.2.  Policy-Aware Forwarding  . . . . . . . . . . . . . . . .  16
     12.3.  SLA Enforcement  . . . . . . . . . . . . . . . . . . . .  16
     12.4.  Real-Time Media (Voice and Video) Quality Signalling . .  16
     12.5.  On-Demand Streaming Video Delivery Optimisation  . . . .  16
   13. IPv4 vs IPv6 vs IPv7 Comparison . . . . . . . . . . . . . . .  16
   14. Use Case: Botnet Attack Mitigation  . . . . . . . . . . . . .  16
     14.1.  Scenario . . . . . . . . . . . . . . . . . . . . . . . .  17
     14.2.  IPv4/IPv6 Failure  . . . . . . . . . . . . . . . . . . .  17
     14.3.  IPv7 Solution  . . . . . . . . . . . . . . . . . . . . .  17
     14.4.  Cost Shift . . . . . . . . . . . . . . . . . . . . . . .  17
   15. Use Case: Interactive Conferencing (Voice and Video)  . . . .  17
     15.1.  Scenario . . . . . . . . . . . . . . . . . . . . . . . .  18
     15.2.  IPv4/IPv6 Limitations  . . . . . . . . . . . . . . . . .  18
     15.3.  IPv7 Approach  . . . . . . . . . . . . . . . . . . . . .  18
     15.4.  Operational Notes  . . . . . . . . . . . . . . . . . . .  18
   16. Use Case: On-Demand Streaming Video . . . . . . . . . . . . .  18
     16.1.  Scenario . . . . . . . . . . . . . . . . . . . . . . . .  18
     16.2.  IPv4/IPv6 Limitations  . . . . . . . . . . . . . . . . .  19
     16.3.  IPv7 Approach  . . . . . . . . . . . . . . . . . . . . .  19
     16.4.  Operational Notes  . . . . . . . . . . . . . . . . . . .  19
   17. Implementation Considerations . . . . . . . . . . . . . . . .  19
     17.1.  Signature Verification Performance . . . . . . . . . . .  19
     17.2.  Trust Level Updates via Gossip Protocol  . . . . . . . .  19

Subbiah                  Expires 27 October 2026                [Page 3]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

     17.3.  EIT Generation and Rotation  . . . . . . . . . . . . . .  20
     17.4.  Backward Compatibility and Incremental Deployment  . . .  20
     17.5.  Scalability Considerations . . . . . . . . . . . . . . .  20
     17.6.  Resiliency and High Availability . . . . . . . . . . . .  20
     17.7.  AI/ML-Assisted Policy and Telemetry (Non-Normative)  . .  20
     17.8.  Futuristic Extensions (Non-Normative)  . . . . . . . . .  20
   18. Security Considerations . . . . . . . . . . . . . . . . . . .  21
     18.1.  Signature Key Compromise . . . . . . . . . . . . . . . .  21
     18.2.  Replay Attacks . . . . . . . . . . . . . . . . . . . . .  21
     18.3.  Role Escalation  . . . . . . . . . . . . . . . . . . . .  22
     18.4.  Privacy Leakage via Traffic Analysis . . . . . . . . . .  22
     18.5.  Trust Level Depletion Attack . . . . . . . . . . . . . .  22
   19. IANA Considerations . . . . . . . . . . . . . . . . . . . . .  22
   20. References  . . . . . . . . . . . . . . . . . . . . . . . . .  22
     20.1.  Normative References . . . . . . . . . . . . . . . . . .  22
     20.2.  Informative References . . . . . . . . . . . . . . . . .  23
   21. Conclusion  . . . . . . . . . . . . . . . . . . . . . . . . .  23
   Author's Address  . . . . . . . . . . . . . . . . . . . . . . . .  23

1.  Introduction

   The current Internet architecture (IPv4 and IPv6) is built upon the
   principle of "reachability first, security second."  IP addresses
   identify connection points or topological locations, not the identity
   or intent of the sender.  This fundamental architectural gap has
   enabled the proliferation of residential proxy networks - a multi-
   billion-dollar market where malicious actors mask their identity
   behind legitimate consumer IP addresses to conduct fraud, credential
   stuffing, and distributed denial of service (DDoS) attacks.

   With the explosive growth of Internet-connected devices (IoT) -
   including smart TVs, robotic vacuum cleaners, refrigerators, washing
   machines, and other consumer appliances - the security challenges
   have intensified.  Most of these devices run Linux or Android-based
   systems with customizable network stacks that can be modified at the
   kernel level.  However, current IPv4 and IPv6 lack the native
   mechanisms to authenticate the true origin of traffic or enforce
   network-layer policy based on device identity and trust.

   IPv7 defines an address and header structure that allows an IPv7 node
   to (1) validate the binding between an asserted provider and the
   packet origin, and (2) convey coarse-grained policy and reputation
   signals to forwarding devices.  In IPv7, the source address is
   treated as an authenticated assertion rather than a purely
   topological locator.  This document describes the on-wire format,
   processing model, and security considerations for these mechanisms.

Subbiah                  Expires 27 October 2026                [Page 4]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

1.1.  Conventions and Terminology

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
   "OPTIONAL" in this document are to be interpreted as described in RFC
   2119.

   IPv7 Node: A host or router that originates, forwards, or terminates
   IPv7 packets.

   Ephemeral Identity Token (EIT): A time-bound token carried in the
   User Identity component of the Variable-Length Identity Block (VLIB).

   Provider ID: An identifier for the originating ISP or provider domain
   used for signature verification and policy decisions.

   Origin Signature: A cryptographic signature that enables a verifier
   to validate the source/provider binding and detect unauthorised
   proxying.

   Variable-Length Identity Block (VLIB): A structured identity block
   that carries EIT, service, location, provider, tenant, role/policy,
   and signature components.

   Source-Provider Validation (SPV): A validation procedure that checks
   whether an IPv7 packet's claimed Provider ID matches the provider
   that can validate the Origin Signature.

1.2.  Key Design Pillars

   Identity-Carrying Addressing: The address syntax includes explicit
   service, location, provider, tenant, and role/policy components to
   support operational attribution.

   Origin Validation: Packets include an Origin Signature that enables
   Source-Provider Validation (SPV) at, or near, the first hop.

   Trust and Reputation Signalling: Packets carry compact trust/
   reputation indicators to support fast-path filtering and traffic
   management.

   Privacy Preservation: The EIT mechanism is intended to limit
   disclosure of a subscriber's long-term identity to intermediate
   routers while enabling accountability by the originating provider.

Subbiah                  Expires 27 October 2026                [Page 5]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

1.3.  Goals and Capabilities

   IPv7 is designed to achieve the following objectives:

   Signal-Based Endpoint Integrity Monitoring: While IPv7 cannot perform
   malware analysis on endpoints, it SHOULD enable provider-edge routers
   to receive and act upon signals from machine-learning (ML) analysers.
   These signals allow routers to flag packets from potentially
   compromised devices based on behavioral analysis, traffic pattern
   anomalies, and reputation scoring.  This approach shifts
   responsibility for anomaly detection to the provider's infrastructure
   rather than requiring it at the endpoint.

   Fraud Mitigation through Network-Layer Attribution: IPv7 is intended
   to reduce the value and effectiveness of unauthorized residential
   proxying by providing cryptographic binding between a packet's
   asserted origin and the originating provider.  This enables targets
   to distinguish legitimate residential traffic from proxy-relayed
   abuse, supporting fraud prevention at the network layer.

   Provider-Operated Identity Domains: IPv7 assumes that providers
   operate local validation domains with independent policy.  This
   architecture supports jurisdictional compliance and operational
   flexibility while maintaining cryptographic integrity across domain
   boundaries.

   Legal Process Support: IPv7 provides a framework for identity
   disclosure under law enforcement requests.  The exact legal and
   policy mechanisms for such disclosures remain jurisdictional and are
   outside the scope of this protocol specification, but the protocol is
   designed to support such processes at the provider level.

   Network-Layer Policy Enforcement: By including role, policy, and
   trust-level information in the address itself, IPv7 enables routers
   to enforce intent-based policies without relying solely on
   application-layer controls or external reputation databases.

2.  Non-Goals

   IPv7 does not perform endpoint-side malware analysis or endpoint
   integrity verification.

   IPv7 does not replace application-layer authentication or
   authorization mechanisms (e.g., MFA, session management, and access
   control).

   IPv7 does not prevent social engineering or application-layer
   credential theft attacks.

Subbiah                  Expires 27 October 2026                [Page 6]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

   IPv7 does not mandate a single global identity provider or global
   identity registry; the protocol assumes provider-operated validation
   domains.

   IPv7 does not specify legal processes for identity disclosure; such
   processes remain jurisdiction-dependent and outside the scope of this
   protocol.

3.  Problems with IPv4 and IPv6

3.1.  Identity Masking via Residential Proxies

   IPv4 and IPv6 addresses identify connection points, not people.
   Residential proxies exploit this by routing malicious traffic through
   legitimate home IPs, making it difficult for servers to distinguish
   legitimate residential activity from automated abuse without
   additional signals or application-layer analysis.

   The market for residential proxies generates billions in annual
   revenue, while victims lack consistent protocol-level mechanisms to
   authenticate source/provider binding and often rely on application-
   layer detection and third-party mitigation.

3.2.  Stateless Security (Dumb Headers)

   IPv4/v6 headers contain routing information but no metadata about
   sender intent or reputation.  Security is always an add-on
   (firewalls, WAFs) rather than built in.

   This design creates high overhead for deep packet inspection and
   forces applications to implement their own reputation, trust, and
   policy enforcement, duplicating work across millions of servers.

3.3.  Complexity and Human Error

   IPv6 addresses (e.g., 2001:0db8:85a3::8a2e:0370:7334) can be
   difficult for operators to read and reason about directly, increasing
   the risk of configuration errors and reliance on tooling for
   interpretation.

   This difficulty impairs auditing, troubleshooting, and incident
   response, making security logs cryptic and preventing operators from
   quickly understanding "who is attacking us."

Subbiah                  Expires 27 October 2026                [Page 7]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

3.4.  Lack of Native Trust Tiers

   In IPv4 and IPv6, packets are typically forwarded without an
   explicit, standardised trust tier at the network layer.  Trust and
   prioritisation are commonly implemented using local policy and
   external systems after traffic is received.

   DDoS attacks receive the same network resources as legitimate
   traffic.  Routers cannot prioritise based on sender reputation,
   forcing targets to over-provision bandwidth or rely on external
   mitigation services.

4.  IPv7 Solutions

4.1.  Identity-Centric Addressing

   IPv7 represents addresses as hierarchical identity strings.  An IPv7
   source address MUST include sufficient components to support provider
   validation and policy enforcement.

   Format:
   [EIT]/service.location.provider.tenant.role.trustlevel.reputationscope.[Origin_Signature]

   Example: eit_7f3a9c2b/
   web.nyc.exampleisp.home.guest.medium.local.ed25519sig_...

4.2.  Eliminating Residential Proxies via Source-Provider Validation

   IPv7 defines Source-Provider Validation (SPV) using the Provider ID
   and Origin Signature components.  An IPv7 validating node (typically
   the first-hop router) MUST verify that the Origin Signature is valid
   under a public key associated with the asserted Provider ID.  If
   validation fails, the packet MUST be dropped.  Providers SHOULD
   rotate signing keys and deploy key protection mechanisms (e.g., HSMs)
   to reduce the impact of key compromise.

4.3.  Built-In Reputation and Hardware-Level Filtering

   By including trust level and reputation scope in the address itself,
   routers can perform hardware-level filtering.  Instead of a software
   firewall consulting a database, a router can detect an indicator such
   as "trustlevel.low" and apply local policy (e.g., throttling or
   quarantine).  This can reduce per-packet CPU cost and avoid
   duplicating reputation logic across endpoints.

Subbiah                  Expires 27 October 2026                [Page 8]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

4.4.  Granular Policy Enforcement

   The role and policy fields enable intent-based networking.  An IPv7
   node MAY indicate that a packet is associated with a specific access
   intent (e.g., "read-only" or "admin").  If a packet asserts a role
   that is not authorised for a given path, a policy-enforcing device
   SHOULD reject the traffic before it reaches the destination service.

4.5.  Human-Readable Auditing

   IPv4 Example: 192.0.2.1

   IPv7 Example: malicious_user/
   bot.london.exampleisp.home.guest.0.global

   This representation is intended to improve operator attribution
   during troubleshooting and incident response.

5.  Darknet Diaries Case Studies: IPv7 as the Fix

5.1.  Episode 172: SuperBox

   Consumer network devices (e.g., illicit streaming boxes) may exhibit
   malicious behaviour, including unsolicited outbound connections,
   local network scanning, and traffic proxying.  Such devices can be
   used to conceal an attacker's origin behind a residential connection
   and to generate abuse traffic from "clean" residential IP space.

   In IPv7, the first-hop validating node performs Source-Provider
   Validation (SPV) over the Provider ID and Origin Signature.  Packets
   that traverse an unauthorised proxy path will fail validation and
   MUST be dropped.  This allows abuse originating from compromised
   residential environments to be filtered near the source, reducing
   reliance on application-layer bot detection.

5.2.  Episode 128: Gollumfun (Part 1) (Fraud and Identity Abuse)

   Underground cybercrime ecosystems enable fraud at scale, including
   credential theft, account takeover, and carding.  These activities
   are operationally supported by infrastructure that reduces
   attribution (e.g., use of compromised devices, proxies, and
   anonymisation techniques).

Subbiah                  Expires 27 October 2026                [Page 9]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

   IPv7's Source-Provider Validation (SPV) and trust/reputation
   signalling are intended to improve attribution and early filtering of
   abuse traffic.  Where fraud operations rely on unauthorised proxying
   through residential networks, first-hop validation can reduce the
   effectiveness of such infrastructure by requiring origin/provider
   binding for forwarded traffic.

5.3.  Episode 110: Spam Botnets

   Large-scale botnets can be formed from compromised endpoints and used
   to send spam or generate abuse traffic at high volume.  These botnets
   commonly rely on weak authentication, unpatched software, and an
   inability for the network to attribute or rate-limit traffic close to
   its true source.

   IPv7 supports first-hop origin validation (SPV) and explicit rate
   enforcement.  A validating node MAY apply local policy to rate-limit
   or discard packets based on the Trust/Reputation field and the Rate-
   Limit Token.  This enables abuse traffic to be constrained near its
   origin and reduces reliance on downstream filtering alone.

5.4.  Episodes 45-46: Xbox Underground (Credential Reuse and Lateral
      Movement)

   Attackers frequently use stolen credentials and credential reuse to
   gain initial access, then pivot laterally inside trusted networks to
   reach higher-value systems.  Network-layer forwarding does not
   typically account for authenticated role assertions.

   IPv7 allows forwarding devices to apply policy based on role/policy
   signalling carried in the VLIB.  Role and policy assertions that
   affect forwarding decisions SHOULD be integrity-protected so that on-
   path modification causes validation failure.  This supports network-
   layer containment by preventing packets asserting a low-privilege
   role (e.g., "guest") from being forwarded to high-privilege
   destinations (e.g., "admin-only").

6.  Additional Motivation from Public Incident Reporting

6.1.  Botnets Monetised as Residential Proxy Infrastructure

   Recent public reporting describes large botnets composed of
   compromised consumer devices (e.g., routers, cameras, and low-cost
   streaming boxes) that are not only used for volumetric DDoS but are
   also rented out as residential proxy endpoints.  Under IPv4 and IPv6,
   the network layer does not provide an origin-authentication
   primitive; downstream destinations observe traffic as originating
   from the compromised subscriber's IP address and cannot distinguish

Subbiah                  Expires 27 October 2026               [Page 10]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

   legitimate use from unauthorised proxying without extensive
   application-layer detection.

   IPv7 is intended to enable earlier filtering by requiring Source-
   Provider Validation (SPV) and by supporting explicit rate
   enforcement.  A first-hop validating node can drop packets that fail
   provider binding and can apply local rate policy using the Trust/
   Reputation field and Rate-Limit Tokens.

6.2.  Credential Abuse and the Limits of Source Addressing

   Annual incident and breach analyses consistently identify stolen
   credentials and exploitation of exposed services as dominant initial
   access vectors.  Once valid credentials are obtained, an attacker's
   traffic is often indistinguishable at the IP layer from legitimate
   user traffic.

   IPv7's role/policy signalling and trust-aware forwarding are intended
   to provide additional containment and prioritisation options at the
   network layer; however, they are not a replacement for strong
   authentication and application-layer authorisation.

7.  Technical Specification

7.1.  IPv7 Packet Header Format

   IPv7 packets begin with a fixed 40-byte header for rapid processing
   in routers, followed by a variable-length identity block (VLIB)
   containing hierarchical identity information.

7.1.1.  Fixed Header Section (40 bytes)

   The fixed header contains routing, flow identification, and trust
   signalling fields for fast-path processing.

7.1.2.  Variable-Length Identity Block (VLIB)

   The VLIB contains hierarchical identity information including EIT,
   provider, tenant, service, location, role, and Origin Signature
   components.

7.2.  How Routers Process IPv7 Headers

7.2.1.  Three-Stage Processing

   Fast Path (Fixed Header): The router examines the Trust/Reputation
   octet.  Packets below a locally configured threshold MAY be rate-
   limited or discarded, particularly under congestion.

Subbiah                  Expires 27 October 2026               [Page 11]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

   Validation Path: The router performs SPV by verifying the Origin
   Signature against the asserted Provider ID.  Packets that fail
   validation MUST be dropped.

   Routing Path: The router uses provider, location, and role/policy
   information to select the next hop and apply local forwarding policy.

8.  Operational Considerations

8.1.  Key Management and Rollover

   Operational deployments MUST define key lifecycle procedures for
   Origin Signature verification, including issuance, rotation, and
   revocation.  Routers SHOULD support caching of provider public keys
   with explicit freshness bounds.  Providers SHOULD support seamless
   rollover (e.g., overlapping validity windows) to avoid traffic loss
   during rotation events.

8.2.  First-Hop Deployment Model

   SPV is expected to be enforced primarily at, or near, the source
   attachment network (e.g., access edge, CPE gateway, or provider
   first-hop router).  Core and transit networks SHOULD NOT be required
   to maintain subscriber-specific state to forward IPv7 traffic.
   Operators MAY deploy SPV selectively (e.g., only for certain roles,
   tenants, or destination prefixes) during incremental rollout.

8.3.  Telemetry and Troubleshooting

   To support incident response, implementations SHOULD provide counters
   and logs for SPV outcomes (e.g., signature failures, key-fetch
   failures, replay rejections) and SHOULD allow operators to correlate
   drops with Provider ID, tenant, and role/policy fields.

8.4.  Policy and Misconfiguration Risk

   Because role/policy signals may influence forwarding, operators MUST
   consider the risk of misconfiguration causing unintended denial of
   service.  Implementations SHOULD support safe defaults (e.g., treat
   unknown roles as least privilege) and SHOULD support staged policy
   rollout with monitoring before enforcement.

Subbiah                  Expires 27 October 2026               [Page 12]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

8.5.  Interconnection Considerations

   Inter-domain deployment requires agreement on how Provider IDs are
   formed and how validation keys are distributed or discovered.
   Operators SHOULD support multiple authorised Provider IDs for multi-
   homed environments and SHOULD define local policy for traffic
   arriving from domains that do not support SPV.

8.6.  Manageability Considerations

   Implementations SHOULD expose operational configuration and telemetry
   for SPV and policy decisions via existing management frameworks
   (e.g., vendor CLIs, NETCONF/RESTCONF data models where available).
   At a minimum, operators SHOULD be able to configure trust thresholds,
   rate-limit behaviour, and fallback policy for key unavailability.

9.  Deployment and Transition Considerations

9.1.  Incremental Deployment Models

   IPv7 is expected to be deployed incrementally.  Early deployment
   models include: (1) dual-stack hosts and routers that support IPv7
   alongside IPv4/IPv6; (2) IPv7-in-IPv6 (or IPv7-in-IPv4) encapsulation
   between IPv7-capable edges; and (3) translation gateways that
   terminate IPv7 and originate IPv4/IPv6 flows toward legacy
   destinations.

9.2.  Coexistence and Negotiation

   Where both peers support IPv7, an endpoint SHOULD prefer IPv7 for
   flows that benefit from origin validation and policy signalling.
   Name resolution mechanisms are expected to evolve to return both IPv7
   and IPv6/IPv4 locators where applicable.

9.3.  Naming and Discovery Considerations

   This document does not specify new DNS resource record types.  A
   standards-track evolution of IPv7 would be expected to define how
   IPv7 locators and/or identity strings are represented for name
   resolution and service discovery.

9.4.  Middleboxes, Firewalls, and Translation

   Existing middleboxes that rewrite packet headers may interfere with
   IPv7 if they modify fields covered by the Origin Signature.
   Deployments SHOULD treat the VLIB as integrity-protected metadata and
   SHOULD avoid in-path modification.

Subbiah                  Expires 27 October 2026               [Page 13]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

10.  Privacy Model: Hybrid Anonymity

   IPv7 is designed to support accountability while limiting identity
   disclosure to on-path devices.  The protocol uses Ephemeral Identity
   Tokens (EITs) such that intermediate routers can forward traffic and
   apply coarse-grained policy without requiring access to a
   subscriber's long-term identifier.

10.1.  Ephemeral Identity Tokens (EIT)

   The User Identity field does not contain the user's legal name or
   permanent identifier.  Instead, it contains an Ephemeral Identity
   Token (EIT) - a time-bound opaque value renewed per session
   (typically every 24 hours).  Intermediate routers see a "Verified"
   identity with a "High Trust" score but cannot determine the user's
   identity across sessions.

10.2.  ISP-Level Verification

   The originating provider is assumed to maintain the ability to map an
   EIT to subscriber records using local operational data (e.g.,
   provisioning, authentication, and accounting logs).  The mechanisms
   and conditions under which such mappings are disclosed to third
   parties are out of scope for this document and depend on applicable
   policy and law.

10.3.  Reputation without Identification

   IPv7 separates conduct from identity.  If an EIT is flagged for
   malicious activity, the reputation of the associated Tenant ID or
   Provider ID is affected - not the individual user.  This creates a
   system of collective responsibility, incentivising ISPs and network
   operators to monitor tenant health without inspecting private user
   data.

10.4.  Optional Selective Disclosure

   Only the destination server (after mutual authentication) can request
   the true identity behind an EIT.  Users can choose to publish their
   EIT to public key mapping via distributed PKI (DNS CERT records,
   blockchain), allowing applications to verify identity without
   revealing their legal name to network operators.

11.  Advanced Security Features

Subbiah                  Expires 27 October 2026               [Page 14]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

11.1.  Quantum-Resistant Cryptography

   IPv7 incorporates a quantum-resistant signature option using
   CRYSTALS-Dilithium (NIST PQC standard).  Devices negotiate the
   strongest mutually supported algorithm, enabling smooth migration
   from classical to post-quantum cryptography without protocol
   redesign.

11.2.  Multi-Layer Authentication

   Layer 1 (Origin Verification): Origin Signature validated at first-
   hop router.

   Layer 2 (Identity Verification): EIT token validated against user's
   public key certificate.

   Layer 3 (Policy Enforcement): Role/Policy validated against
   destination ACL.

11.3.  Built-In DDoS Mitigation

   IPv7 packets include a Rate-Limit Token field.  ISPs issue time-
   limited tokens allowing N packets per second.  Routers enforce tokens
   at the hardware level, preventing botnet traffic from overwhelming
   targets.  Malformed or forged tokens are silently dropped.

11.4.  Reputation-Based Filtering

   The Trust Level field is dynamically updated.  Nodes observing
   malicious behaviour (failed authentication, port scans, botnet
   signatures) decrement the Trust Level.  Once below threshold (e.g.,
   50/255), packets are rate-limited or quarantined.  Trust can be
   restored through ISP intervention or automatic recovery.

12.  Advanced Routing Features

12.1.  Trust-Aware Path Selection

   IPv7-BGP routing protocols compute paths based on topological
   distance and trust scores of intermediate ASes.  A path with lower
   cost may be rejected if intermediate routers have low reputation,
   preventing traffic hijacking through compromised ASes.

Subbiah                  Expires 27 October 2026               [Page 15]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

12.2.  Policy-Aware Forwarding

   Routers make decisions based on Role/Policy, not just destination.  A
   packet with a .guest role cannot reach .admin-only resources.  Access
   control shifts from the application layer to the network layer,
   reducing server-side processing.

12.3.  SLA Enforcement

   Service Level Agreements can be enforced at the network layer using
   IPv7's explicit service identifiers and trust signals.

12.4.  Real-Time Media (Voice and Video) Quality Signalling

   Interactive voice and video conferencing is sensitive to latency,
   jitter, and packet loss.  IPv7's explicit QoS Class and role/policy
   fields enable routers to apply local policy (e.g., low-latency
   forwarding, congestion handling, or admission control) based on
   authenticated context.  Where deployed, trust/reputation and provider
   binding can also assist operators in prioritising interactive media
   flows from validated sources during congestion.

12.5.  On-Demand Streaming Video Delivery Optimisation

   On-demand streaming video remains throughput-driven but is sensitive
   to rebuffering and sudden congestion.  IPv7's explicit service
   identifier, location, provider/tenant context, and QoS class can
   enable policy-aware routing decisions such as preferring local
   caches, applying per-tenant traffic engineering, and providing
   differentiated treatment for premium or latency-sensitive segments.
   SPV and trust signalling can help reduce abuse that leverages
   residential proxy infrastructure to scrape or attack streaming
   services.

13.  IPv4 vs IPv6 vs IPv7 Comparison

   A comprehensive comparison table comparing IPv4, IPv6, and IPv7
   across key attributes (addressing, security, policy, scalability, and
   operability) would be included in a full implementation.

14.  Use Case: Botnet Attack Mitigation

Subbiah                  Expires 27 October 2026               [Page 16]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

14.1.  Scenario

   A malicious actor in Region A wants to perform credential stuffing on
   a bank in Region B.  To bypass the bank's firewall, the attacker
   rents access to a residential proxy service routing traffic through a
   legitimate home router in Region C (compromised via weak
   credentials).

14.2.  IPv4/IPv6 Failure

   The Packet Path: Attacker (192.0.2.5) - Proxy (198.51.100.14) - Bank

   The Mechanism: The proxy server strips the attacker IP and replaces
   it with a residential IP.

   Bank Result: The bank sees a legitimate residential customer.

   Detection: Not generally feasible at the network layer using IP
   addressing alone.  The bank typically relies on bot-detection and
   fraud controls at higher layers (with false positives and false
   negatives).

14.3.  IPv7 Solution

   The Packet Header:
   [eit_7f3a9c]/web.london.exampleisp.home.guest.low.[sig_ed25519]

   Step 1 (Identity Bind): EIT is cryptographically tied to the home
   router's hardware.

   Step 2 (Signature Verification): The attacker cannot generate a valid
   Origin Signature without the provider's signing key.

   Step 3 (The Drop): The first ISP router notices a Provider ID/Origin
   Signature mismatch.  The packet is dropped at the edge.  The attacker
   never reaches international cables or the bank.

14.4.  Cost Shift

   In IPv4/v6, many defensive controls against proxy abuse are applied
   at the destination (or by downstream mitigation providers).  IPv7
   shifts some enforcement capability toward the source attachment
   network by enabling first-hop validation and policy-based dropping.
   This can increase attacker cost in scenarios that depend on
   unauthorised proxying through residential networks.

15.  Use Case: Interactive Conferencing (Voice and Video)

Subbiah                  Expires 27 October 2026               [Page 17]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

15.1.  Scenario

   An enterprise user participates in an interactive voice/video
   conference (e.g., Microsoft Teams or Zoom) from a residential access
   network while other household devices concurrently generate bulk
   traffic (e.g., software updates or large downloads).  The
   conferencing application requires low latency and low jitter to
   maintain conversational quality, but the access uplink and upstream
   aggregation links are intermittently congested.

15.2.  IPv4/IPv6 Limitations

   Under IPv4 and IPv6, routers generally forward packets without an
   authenticated statement of application intent.  QoS treatment is
   often inferred from DSCP markings or local policy, but DSCP is not
   consistently preserved or honoured across networks.  As a result,
   conferencing traffic competes with bulk flows during congestion.

15.3.  IPv7 Approach

   With IPv7, an endpoint MAY indicate interactive media intent using
   the QoS Class field and MAY include a role/policy appropriate for
   real-time collaboration.  Where local policy permits, the first-hop
   validating node can apply Source-Provider Validation (SPV) and then
   provide preferential queueing or congestion handling for validated
   real-time flows (e.g., lower latency treatment) while rate-limiting
   or deprioritising bulk traffic.

15.4.  Operational Notes

   Operationally, real-time flows benefit from stable queueing and
   bounded latency.  Implementations SHOULD expose counters for drops
   and queueing outcomes by QoS Class and role/policy to support
   troubleshooting.  Deployments MUST consider privacy implications of
   any per-application classification.

16.  Use Case: On-Demand Streaming Video

16.1.  Scenario

   A user streams on-demand video (e.g., Netflix or Amazon Prime Video)
   from a nearby cache while other traffic on the access network is
   bursty.  The streaming service adapts bitrate based on measured
   throughput and rebuffering events.  During peak periods, some paths
   to caches exhibit congestion or intermittent loss, reducing quality
   of experience (QoE).

Subbiah                  Expires 27 October 2026               [Page 18]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

16.2.  IPv4/IPv6 Limitations

   In IPv4 and IPv6 deployments, cache selection and traffic engineering
   commonly rely on DNS-based steering, anycast, and proprietary
   telemetry.  The IP layer itself provides limited authenticated
   context about service intent, tenant boundaries, or delivery class.

16.3.  IPv7 Approach

   IPv7 can provide explicit, coarse-grained delivery context via the
   Service ID, location, provider/tenant components, and QoS Class.
   Operators MAY use this context to implement policy-aware routing such
   as preferring local caches, applying per-tenant traffic engineering,
   or selecting congestion-avoidant paths during peak demand.  Where
   streaming abuse relies on unauthorised proxying through residential
   networks, SPV and trust/reputation signalling can provide additional
   inputs for early filtering and rate enforcement near the source
   attachment network.

16.4.  Operational Notes

   QoE (quality of experience) is impacted by sustained throughput,
   start-up delay, and rebuffering frequency.  Operators SHOULD consider
   how QoS Class and service identifiers interact with congestion
   management to avoid starving other traffic.

17.  Implementation Considerations

17.1.  Signature Verification Performance

   Ed25519 signature verification at line rate is feasible on modern
   hardware.  CPUs with SHA-3 and Ed25519 instructions (AVX-512, ARM
   SVE) verify approximately 100,000 signatures per second per core.
   This is sufficient for backbone routers processing millions of
   packets per second using dedicated crypto accelerators (e.g., Intel
   QuickAssist, ARM TrustZone).

17.2.  Trust Level Updates via Gossip Protocol

   Rather than updating Trust Levels on every malicious packet (massive
   state overhead), IPv7 uses a gossip protocol where nodes periodically
   exchange reputation data.  Low-trust addresses are marked in Bloom
   filters for fast negative lookups, reducing memory footprint.

Subbiah                  Expires 27 October 2026               [Page 19]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

17.3.  EIT Generation and Rotation

   EITs are generated using HASH(user's_long_term_public_key +
   session_nonce + ISP_signing_key).  ISPs can verify EIT validity
   without learning which specific user generated it.  EITs are valid
   for 24 hours; users request new tokens via an authenticated channel.

17.4.  Backward Compatibility and Incremental Deployment

   Detailed deployment and transition considerations are discussed in
   Section 10.  Implementers SHOULD ensure that signed fields are
   treated as integrity-protected metadata and that any translation or
   tunnelling preserves the deployment model's security assumptions.

17.5.  Scalability Considerations

   IPv7 is intended to support Internet-scale routing and forwarding by
   minimising per-flow state in the network core.  SPV is expected to
   occur primarily at, or near, the first hop; intermediate domains are
   not required to perform subscriber-specific validation.  Trust/
   reputation processing is designed for fast-path evaluation using
   compact fields and locally configured policy thresholds.

17.6.  Resiliency and High Availability

   Deployments MUST consider failure modes for validation and key
   management.  If SPV cannot be performed (e.g., key retrieval
   failure), an implementation SHOULD support a locally configurable
   policy that defines whether traffic is dropped, rate-limited, or
   forwarded with reduced trust.

17.7.  AI/ML-Assisted Policy and Telemetry (Non-Normative)

   This document does not require artificial intelligence (AI) or
   machine learning (ML).  However, operators MAY use ML models to
   improve operational outcomes, for example: predicting QoE (quality of
   experience) degradation for interactive media, detecting anomalous
   traffic patterns that correlate with proxy abuse, and tuning local
   policy thresholds for trust/reputation and rate enforcement.

17.8.  Futuristic Extensions (Non-Normative)

   Intent negotiation: A future extension could allow endpoints to
   negotiate acceptable QoS classes and policy constraints without
   relying on out-of-band configuration.

Subbiah                  Expires 27 October 2026               [Page 20]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

   In-network telemetry signals: Standardised metadata for congestion
   exposure or path health could improve troubleshooting and closed-loop
   traffic engineering without payload inspection.

   Privacy-preserving feature export for ML-based operations: A future
   extension could define standardised, coarse-grained telemetry
   features for ML-assisted anomaly detection and QoE prediction without
   requiring payload inspection or stable user identifiers.

   Federated learning for cross-domain models: A future extension could
   enable federated learning approaches in which multiple operators
   train shared models for anomaly detection or QoE prediction while
   keeping raw telemetry local.

   Multipath-aware identity binding: A future extension could support
   binding a single identity to multiple simultaneous paths (e.g.,
   multi-access) while preserving provider validation properties.

   Programmable policy objects: A constrained policy object format could
   enable richer, interoperable policy enforcement while remaining
   bounded for router fast paths.

   Cryptographic agility: Explicit algorithm agility beyond the initial
   signature set to support post-quantum transitions and future
   primitives.

   Encrypted metadata: Selective encryption of portions of the VLIB
   could reduce metadata leakage while retaining verifiability where
   required.

18.  Security Considerations

18.1.  Signature Key Compromise

   If a provider signing key is compromised, an attacker could forge
   Origin Signatures for that provider.  Providers SHOULD protect
   signing keys using appropriate controls (e.g., hardware security
   modules) and SHOULD define rapid revocation and rollover procedures.
   Validating nodes SHOULD support key freshness bounds and SHOULD be
   able to reject signatures from revoked keys based on locally
   configured policy.

18.2.  Replay Attacks

   Attackers could capture and replay valid packets.  Mitigation: IPv7
   packets include a timestamp and a nonce.  Routers maintain a short-
   lived cache of (packet_hash, nonce) pairs and reject duplicates.

Subbiah                  Expires 27 October 2026               [Page 21]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

18.3.  Role Escalation

   An attacker could attempt to assert a higher-privilege role (e.g.,
   "admin").  Role/policy fields that affect forwarding or access
   decisions SHOULD be integrity-protected (e.g., covered by the Origin
   Signature) so that unauthorised modification causes validation
   failure.  Destination systems SHOULD continue to perform application-
   layer authorisation independent of network-layer signalling.

18.4.  Privacy Leakage via Traffic Analysis

   An adversary could correlate behavioural patterns (timing, volume) to
   de-anonymise users.  Mitigation: Users employ packet padding and
   random inter-packet delays.  ISPs provide "privacy mode" where
   multiple users share a single EIT during a session.

18.5.  Trust Level Depletion Attack

   An attacker might attempt to influence trust signals to cause
   legitimate traffic to be deprioritised.  Implementations that accept
   trust updates from the network MUST authenticate and authorise such
   updates.  Nodes SHOULD apply rate-limits and sanity checks to
   reputation inputs and SHOULD support administrative override
   procedures for recovery from erroneous or malicious downgrades.

19.  IANA Considerations

   This document makes no request of IANA at this stage.  If the
   mechanisms described here are progressed on the standards track, IANA
   actions would be required to allocate protocol code points and
   establish registries to support interoperability, including
   allocation of an IP version number for IPv7, creation of registries
   for Next Header types, Role/Policy encodings, and Signature
   Algorithms.

20.  References

20.1.  Normative References

   RFC 2119 - Key words for use in RFCs to Indicate Requirement Levels

   RFC 8200 - Internet Protocol, Version 6 (IPv6) Specification

   RFC 8032 - Edwards-Curve Digital Signature Algorithm (EdDSA)

   FIPS 204 - Module-Lattice-Based Digital Signature Standard (CRYSTALS-
   Dilithium)

Subbiah                  Expires 27 October 2026               [Page 22]
Internet-Draft   IPv7: Identity-Centric Network Protocol      April 2026

20.2.  Informative References

   Darknet Diaries Podcast - Episodes 172, 128, 110, 45, 46, 13

   NIST SP 800-207 - Zero Trust Architecture

   NIST Post-Quantum Cryptography Standardization Project

   RFC 9330 - Low Latency, Low Loss, and Scalable Throughput (L4S)
   Internet Service: Architecture

   EU Regulation (EU) 2016/679 - General Data Protection Regulation
   (GDPR)

21.  Conclusion

   This document describes IPv7, an identity-carrying, origin-validating
   network-layer protocol intended to reduce abuse enabled by
   unauthenticated source addressing and residential proxy
   infrastructure.  IPv7 introduces an explicit identity structure
   (VLIB), an Origin Signature for Source-Provider Validation, and
   fields that support policy and reputation signalling.

   With the rapid proliferation of IoT devices running Linux and
   Android-based systems, the ability to modify network stacks at the
   kernel level creates both opportunities and challenges.  IPv7
   provides a framework for these devices to authenticate their origin,
   signal their intent and policy requirements, and participate in a
   trustworthy network ecosystem.

   Further work is required to evaluate deployability, incremental
   transition mechanisms, and interoperability with existing IP
   networks, and to refine the protocol into a form suitable for
   standards-track consideration and eventual RFC publication.

Author's Address

   Arunkumar Subbiah
   Independent
   Email: arunkumar.subbiah@apexadversary.com

Subbiah                  Expires 27 October 2026               [Page 23]