Quantum-Native Architectural Tenets and Philosophy for the Quantum Internet
draft-cacciapuoti-qirg-quantum-native-architecture-01
This document is an Internet-Draft (I-D).
Anyone may submit an I-D to the IETF.
This I-D is not endorsed by the IETF and has no formal standing in the
IETF standards process.
| Document | Type | Active Internet-Draft (individual) | |
|---|---|---|---|
| Authors | Angela Sara Cacciapuoti , Marcello Caleffi , Jessica Illiano , C. De Risi , A. Abane , Joaquin Chung | ||
| Last updated | 2026-04-20 | ||
| 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-cacciapuoti-qirg-quantum-native-architecture-01
Quantum Internet Research Group A. S. Cacciapuoti
Internet-Draft M. Caleffi
Intended status: Informational J. Illiano
Expires: 22 October 2026 C. De Risi
University of Naples Federico II
A. Abane
National Institute of Standards and Technology
J. Chung
Argonne National Laboratory -- Data Science and Learning Division
20 April 2026
Quantum-Native Architectural Tenets and Philosophy for the Quantum
Internet
draft-cacciapuoti-qirg-quantum-native-architecture-01
Abstract
This document extends RFC 9340 by outlining a set of quantum-native
architectural tenets for the design and evolution of the Quantum
Internet. These principles should not be interpreted as dogmas, but
as pragmatic guidelines and criteria for harnessing the unique
properties of quantum entanglement within networked systems. Such
design perspectives, while departing from the classical Internet,
remain aligned with a foundational insight: the principle of constant
change, articulated in RFC 1958.
The document specifies quantum-native extensions to the Quantum
Internet framework, defining an entanglement packet switching
paradigm and an explicit separation between the Quantum Data Plane
and Quantum Control Plane. It introduces Quantum Internet Addressing
to extend quantum semantics into control and coordination, and
generalizes the classical forwarding concept to quantum packets.
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."
Cacciapuoti, et al. Expires 22 October 2026 [Page 1]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
This Internet-Draft will expire on 22 October 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.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 4
4. Architectural Overview . . . . . . . . . . . . . . . . . . . 5
4.1. Quantum Data Plane (QDP) . . . . . . . . . . . . . . . . 6
4.2. Quantum Control Plane (QCP) . . . . . . . . . . . . . . . 8
4.3. Hierarchy & EDC . . . . . . . . . . . . . . . . . . . . . 9
5. Quantum Internet Addressing (QA) . . . . . . . . . . . . . . 11
5.1. Quantum Packet . . . . . . . . . . . . . . . . . . . . . 11
5.1.1. Quantum Packet Structure . . . . . . . . . . . . . . 12
6. Generalized Quantum Forwarding (GQF) . . . . . . . . . . . . 12
6.1. Role within the Architecture . . . . . . . . . . . . . . 13
7. Quantum-Native Principles . . . . . . . . . . . . . . . . . . 13
8. Routing Models and Repeater Realizations Support . . . . . . 14
8.1. Routing-Model Agnosticism . . . . . . . . . . . . . . . . 14
8.2. Repeater-Generation Agnosticism . . . . . . . . . . . . . 15
9. Multi-Domain Routing and Forwarding . . . . . . . . . . . . . 16
9.1. Intra- and Inter-Domain Routing . . . . . . . . . . . . . 16
9.2. Quantum Addressing . . . . . . . . . . . . . . . . . . . 17
9.3. Quantum Forwarding . . . . . . . . . . . . . . . . . . . 17
10. Security Considerations . . . . . . . . . . . . . . . . . . . 18
11. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 18
12. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 18
13. Informative References . . . . . . . . . . . . . . . . . . . 18
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 19
Cacciapuoti, et al. Expires 22 October 2026 [Page 2]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
1. Introduction
The Quantum Internet will interconnect quantum devices to enable
distributed quantum functionalities built upon shared entanglement.
RFC 9340 [RFC9340] laid the initial foundation for such an Internet,
defining its motivation and goals. At that time, the role of a
distinct Quantum Control Plane (QCP) was explicitly declared out of
scope. Subsequent work on multiplane quantum network architectures
has further highlighted the need for control functions
[QI-MULTIPLANE]. This document revisits that open question left by
RFC 9340, by arguing that explicit control is necessary. However,
this alone is not sufficient: control must also evolve beyond
classical coordination toward quantum-native orchestration. To this
end, the document presents an architectural framework that makes this
distinction explicit.
Recent theoretical progress [CalCac25], [CacCal26] has underscored
the critical importance of revisiting the open question of how should
we design the control plane of the Quantum Internet? Quantum
entanglement is a non-local and volatile resource, and it is
therefore inherently stateful [IllCal22] from a networking
perspective. Indeed, in a quantum network, local operations at one
node can instantaneously affect correlated states at remote nodes, by
dynamically reconfiguring entanglement relations among nodes and
thereby modifying the connectivity graph induced by entanglement. As
entanglement relations evolve across nodes and over time, continuous
monitoring, timely state dissemination, and latency-aware control
become necessary for the effective exploitation of entanglement.
Accordingly, the network has to track and expose explicitly resource
descriptors, including, at a minimum, fidelity, residual coherence
time, and ownership (i.e., the nodes participating in a given
entanglement relation). If left uncoordinated, these entanglement
features can trigger the amplification principle [CalCac25], whereby
uncontrolled entanglement resources cause routing ambiguities,
resource inefficiencies and, ultimately, network instability. It is
evident that under these entanglement features an explicit control is
necessary. However, a purely classical control cannot maintain a
consistent global view as quantum networks scale, becoming the
limiting factor for performance and scalability. In summary,
scalable quantum network architectures require not only explicit
tracking and management of entanglement resources, but also control
mechanisms that evolve beyond classical coordination.
Building on the above considerations, this document updates and
extends the architectural principles defined in RFC 9340 [RFC9340],
by introducing a set of quantum-native architectural tenets for the
design of the Quantum Internet. These tenets should not be intended
as dogmas, but as pragmatic guidelines to harness the unique physical
Cacciapuoti, et al. Expires 22 October 2026 [Page 3]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
properties of quantum entanglement within networked systems. In this
sense, the proposed approach echoes the enduring ``principle of
constant change'' articulated in RFC 1958 [RFC1958], reaffirming that
adaptability remains the cornerstone of Internet evolution.
Specifically, this document introduces a quantum-native control and
forwarding architecture composed of the following interlocking
components:
* Quantum Data Plane (QDP): the operational plane that carries and
manipulates entangled qubits (ebits) for applications such as
teleportation. It generalizes the classical notion of forwarding
to the quantum domain through Generalized Quantum Forwarding
(GQF).
* Quantum Control Plane (QCP): the entanglement-orchestrator plane
that manages entanglement resources throughout their entire life-
cycle by relying on the Entanglement-Defined Controller (EDC), a
distributed control entity analogous to a Software-Defined
Networking (SDN) controller, but operating on entanglement
resources to maintain global coherence.
Accordingly, this document defines three core mechanisms -- Quantum
Internet Addressing (QA), Generalized Quantum Forwarding (GQF), and
Entanglement-Defined Controller (EDC) -- that together form the basis
for scalable, entanglement-driven coordination across heterogeneous
quantum domains.
2. Scope
This document is Informational. It proposes architectural tenets and
guidance for researchers and implementers. It is not a protocol
specification. Terminology and notation adhere to monospaced ASCII
presentation for clarity in IRTF review contexts.
3. Terminology
This section defines key terms used throughout this document. Some
definitions extend those introduced in RFC 9340 [RFC9340], reflecting
the evolution from classical coordination to quantum-native
orchestration.
* Quantum Data Plane (QDP): The QDP is the operational plane of the
Quantum Internet that generalizes entanglement forwarding to the
quantum domain through GQF. It is responsible for the generation
of elementary link-level entanglement and for the execution of
quantum operations, including entanglement swapping and
purification.
Cacciapuoti, et al. Expires 22 October 2026 [Page 4]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
* Quantum Control Plane (QCP): The QCP is the logical plane of the
Quantum Internet responsible for orchestrating entanglement
resources by exploiting quantum-addressing abstractions and
relying on the Entanglement-Defined Controllers (EDCs)
* Entanglement Service Provider (ESP): The ESP is a network entity
of the QDP that provides entanglement connectivity to both end-
nodes and peer ESPs. ESPs collectively form the ``entangled-
backbone''.
* Entanglement-Defined Controller (EDC): The Entanglement-Defined
Controller is the logical entity of the QCP responsible for
orchestrating entanglement resources across ESPs through
reconfiguration, monitoring and policy enforcement. It maintains
a view of the network’s entanglement state, to enable scalable and
adaptive control.
* Quantum Internet Addressing (QA): An addressing scheme in which
node identifiers are represented as quantum states. It serves as
addressing abstraction of the QCP.
* Generalized Quantum Forwarding (GQF): A forwarding abstraction
that generalizes the classical prefix-matching forwarding to the
quantum domain.
The above terminology forms the conceptual foundation of this
document. The QDP and QCP represent the two key planes of the
architecture. Within these planes, ESPs (network entities) implement
entanglement forwarding and maintenance, while EDCs (logical
entities) orchestrate the entanglement resources. QA and GQF provide
quantum-native abstractions for addressing and stateful forwarding.
4. Architectural Overview
This section extends Section 5 of [RFC9340].
The Quantum Internet is an entanglement-packet switching network,
where entangled qubits (ebits) replace classical packets as the basic
network units, carrying quantum correlations across network nodes.
This paradigm is not merely an optimization or a refinement of the
classical packet-switching paradigm, but a fundamental departure
imposed by the unique constraints of quantum mechanics. Indeed,
while the goal of the classical packet-switching paradigm is to
determine the best next-hops toward a set of nodes (routing) and to
forward packets along those hops from source to destination,
entanglement-packet switching aims to distribute and manipulate
entanglement among quantum nodes, ultimately entangling the source
and destination(s) regardless of their physical location.
Cacciapuoti, et al. Expires 22 October 2026 [Page 5]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
By inheriting statefulness and non-locality, this switching model
departs from the end-to-end principle [RFC1958], [RFC3439], a key
tenet of the classical Internet design. Indeed, it requires in-
network operations and persistent state awareness across all phases
of the entanglement life-cycle -- from generation and distribution to
storage and final utilization. A broader discussion of the notion of
statefulness adopted in this document is provided in [IllCal22].
Combined with the sophisticated and resource-intensive nature of
state-of-the-art quantum hardware, this paradigm advocates
concentrating the complexity inside the network, while keeping the
edges simple. As a consequence, it mandates a clear decoupling
between the QDP, which handles qubit operations, and the QCP, which
orchestrates entanglement.
+--------------+-----------------------------------+-----------------------+
| Network |Classical Internet |Quantum Internet |
| Feature | | |
+--------------+-----------------------------------+-----------------------+
|Resource |Communication links change slowly |Entanglement is |
|Persistence |relative to packet forward |ephemeral and depleted |
| |dynamics |upon use |
+--------------+-----------------------------------+-----------------------+
|Control plane |It is grounded in classical |It is grounded in |
| |topological-driven abstractions |quantum-native |
| |such as IP-style addressing |abstractions, such as |
| | |quantum addressing |
+--------------+-----------------------------------+-----------------------+
|Data plane |Packet forwarding |Generalized quantum |
| | |forwarding |
+--------------+-----------------------------------+-----------------------+
4.1. Quantum Data Plane (QDP)
The QDP constitutes the operational plane of the Quantum Internet.
It provides the substrate on which entanglement-based connectivity --
also referred to as quantum connectivity -- is established and
maintained among remote nodes. Within the QDP, network entities
exchange and manipulate ebits to establish, extend, and refresh
quantum correlations across the network.
The QDP supports a set of primitives, such as the generation of
elementary link-level entanglement and the execution of quantum
operations including entanglement swapping and purification
[AbaCub25].
Unlike its classical counterpart, the QDP does not act on user
information directly (data in the traditional sense), but it operates
on entanglement that applications later exploit (e.g., quantum
Cacciapuoti, et al. Expires 22 October 2026 [Page 6]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
teleportation and distributed quantum processing). Each
entanglement-link represents a consumable network resource that must
be created, maintained, and periodically refreshed as coherence
decays.
The QDP interfaces closely with the QCP to expose real-time KPIs and
metrics such as fidelity, coherence time, and link availability.
These metrics support adaptive entanglement management and allow the
QCP to optimize resource allocation, path selection, and recovery
procedures. The logical interface between the QCP and QDP may be
realized through classical or quantum signaling channels, functioning
analogously to the control-to-data interface in software-defined
networks [KreRam14]. Detailed protocol specifications are out of
scope for this document.
┌─────────────────────────────┐ ┌────────────────────────────┐
│ │ │ │
│ Plane Abstraction │ │ Network and │
│ │ │ Logical Entities │
│ │ │ │
│┌───────────────────────────┐│ │┌──────────────────────────┐│
││ (1) (QCP) ││ ││ (3) (EDC) ││
││ Quantum Control Plane: ││ ││ Entanglement-Defined ││
││ Entanglement-Orchestrator ││ ││ Controller: ││
││ ││ ││ QCP's Logical Entity ││
││ ││ ││ ││
│└─────────────^─────────────┘│ │└─────────────^────────────┘│
│ │ │ │ │ │
│ Controller Functions: │ │ Controller Functions: │
│ - Reconfiguration │ │ - Reconfiguration │
│ - Monitoring │ │ - Monitoring │
│ - Policy Enforcement │ │ - Policy Enforcement │
│ │ │ │ │ │
│┌─────────────v─────────────┐│ │┌─────────────v────────────┐│
││ (2) (QDP) ││ ││ (4) (ESP) ││
││ Quantum Data Plane : ││ ││ Entanglement Service ││
││ ││ ││ Provider: ││
││ Operational Plane ││ ││ QDP's Network Entity ││
││ ││ ││ ││
│└───────────────────────────┘│ │└──────────────────────────┘│
└─────────────────────────────┘ └────────────────────────────┘
Figure 1: Mapping between plane abstractions and architectural
entities, showing: (1) the QCP as the entanglement-orchestration
plane, (2) the QDP as the operational plane, (3) the EDC as the
QCP logical entity, and (4) the ESPs as QDP network entities.
Cacciapuoti, et al. Expires 22 October 2026 [Page 7]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
4.2. Quantum Control Plane (QCP)
The QCP is the entanglement-orchestration plane of the Quantum
Internet, maintaining a consistent view of entanglement resources
across the network.
In this document, the term control plane is adopted from classical
networking terminology, where it denotes the network-wide logic that
controls packet forwarding among a network, as well as the
configuration and management of these devices and their services
[yang2004rfc3746], [KurRos10]. By analogy, the QCP orchestrates the
entanglement life-cycle -- from generation to distribution and
exploitation. Unlike its classical counterpart, the QCP must account
for the stateful and non-local nature of entanglement, while
coherence-time constraints demand time-aware coordination across the
network.
As indicated in Section 1, effective tracking and management of
entanglement resources are essential for scalable quantum network
architectures. However, if such tracking relies solely on classical
control and signaling, the resulting coordination overhead and
latency prevent the system from maintaining a consistent global view,
ultimately hindering scalability. Even in classical networks, where
entanglement is absent, it has been shown that the number of control
messages required per topology change (namely, the updating
communication overhead) cannot scale better than linearly on
Internet-like topologies [KriCla07]. In the quantum setting, this
challenge becomes even more pronounced due to the intrinsic
statefulness and fragility of entanglement, and it is further
exacerbated when multipartite entanglement is considered [IllCal22].
The QCP coexists with the classical control plane, complementing
rather than replacing it. While the QCP introduces quantum-native
mechanisms for entanglement orchestration, both the QCP and the QDP
rely on an underlying classical communication substrate. Indeed,
classical signaling remains necessary for the functioning of a
quantum network, including, for example, the transmission of
measurement outcomes required to trigger conditional operations at
remote nodes, as also discussed in [RFC9340]. Accordingly, the
architecture described in this document can be interpreted as
augmenting classical control mechanisms rather than replacing them.
A detailed characterization of the interaction between the classical
control plane and the QCP/QDP architecture is outside the scope of
this document and remains an open research question.
Cacciapuoti, et al. Expires 22 October 2026 [Page 8]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
Architecturally, the QCP defines a distinct yet tightly coupled
control logic above the QDP. The QCP interfaces directly with ESPs,
which expose local entanglement capabilities, while ensuring
consistent entanglement resource policies through EDCs.
4.3. Hierarchy & EDC
+--------------------------------+-----------------------------------+
| Classical Tenets | Quantum-Native Tenets |
+--------------------------------+-----------------------------------+
| Complexity located at the | Complexity concentrated in the |
| network edges | core network |
+--------------------------------+-----------------------------------+
| Stateless core network | Stateful core network |
+--------------------------------+-----------------------------------+
| End-to-end protocol design | Network-mediated protocol design |
+--------------------------------+-----------------------------------+
Building on the above considerations, the network architecture is
organized into a two-tier structure, that distinguishes between ESPs
and quantum-edge nodes:
* Bottom Tier (tier-1): Edge quantum nodes (e.g., quantum
processors, sensors, and cryptographic devices) consume
entanglement resources to support quantum applications. These
edge nodes primarily connect to nearby ESPs via short-range
quantum links.
* Top Tier (tier-2): ESPs form the entanglement-core network. They
provide end-to-end entanglement-based connectivity to the lowest
tier by proactively maintaining entangled resources among each
other. The ESPs can be interconnected via long-range quantum
links, such as optical fibers, and they are equipped with the
sophisticated and resource-intensive infrastructure required for
entanglement generation and distribution.
Overall orchestration is achieved through EDCs: distributed logical
entities that maintain coherent global or partial topological views.
EDCs act as the quantum-native counterpart of SDN controllers,
linking control logic directly to quantum states and enabling state-
aware, entanglement-driven routing.
EDCs perform primarily three control-plane functions:
* Reconfiguration: Dynamic management and reallocation of
entanglement resources among ESPs.
Cacciapuoti, et al. Expires 22 October 2026 [Page 9]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
* Monitoring: Assessment of fidelity, coherence time, and
availability of entanglement resources across ESPs.
* Policy enforcement: Application of global policies for routing,
resource allocation, and entanglement-loss recovery.
Although an EDC reflects a centralized control logic, the
architecture supports multiple, potentially federated controllers.
These EDCs coordinate to share partial topological knowledge and
enforce consistent entanglement resource policies, while preserving
local autonomy and scalability.
+---------------------------------------+
| (2) |
| Entanglement Defined Controller (EDC) |
+---------^-----^-----^-----------------+
| | |
|Reconfiguration
| | | Quantum Control Plane (QCP)
| |Monitoring
| | |
| | |Policy Enforcement
----------------------------+-----+-----+-----------------------------
| | |
Entanglement-defined | | | Quantum Data Plane (QDP)
Networking | | |
| | |
+---------------------v-----v-----v--------------------------+
| Top Tier (1) |
| Tier 2 |
| +-------+ +-------+ +--------+ |
| | | | | | | |
| | ESP_1 +----------+ ESP_2 +-----...-------+ ESP_n | |
| | | | | | | |
| +-------+ +-------+ +--------+ |
+------------------------------------------------------------+
Figure 2: Entanglement-Defined Network Architecture showing: (1)
ESPs forming a virtual mesh via proactive entanglement sharing
(dashed lines), (2) the EDC responsible of the QCP
functionalities.
Cacciapuoti, et al. Expires 22 October 2026 [Page 10]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
5. Quantum Internet Addressing (QA)
The architectural decoupling of the QCP and QDP is a necessary
condition for scalability, but it is not sufficient. To manage in-
network operations and maintain persistent state awareness required
by entanglement, the control plane itself must be designed to embrace
quantum principles and phenomena for effectively controlling
entanglement dynamics. This requirement follows once again from the
non-local nature of quantum entanglement: entanglement proximity
cannot be confined to physical distance or restricted to fixed
topological neighborhoods. As a result, control mechanisms based on
locality and topological-driven addressing, such as IP, are
inherently unable to efficiently track, respond to, or propagate
entanglement state evolution across the network. Although
alternative approaches have been proposed in classical networking,
they have not seen widespread adoption in current networking
architectures. Regardless, such approaches remain fundamentally
grounded in classical abstractions and are not designed to natively
capture or manipulate the quantum-state semantics of entanglement
resources. A fundamental rethinking of network addressing and
control mechanisms is therefore needed to embed quantum behavior
directly into the node identifiers, thereby elevating the control
plane to a quantum-native level.
Quantum addressing (QA) provides the logical foundation for this
quantum-native control model. QA does not replace classical
addressing; rather, QA complements it by enabling control and
forwarding functions to be expressed directly on quantum states.
Accordingly, each network node is associated with two types of
identifiers: i) a classical address, such as an IP address, required
for classical communications and signaling; and ii) a quantum
address, represented by a quantum state |A> of an N-qubit system.
Since qubit states can exist in superposition, a sequence of N-qubits
can encode a single node identity, i.e., a single quantum network
address, or a superposition of node identities, each corresponding to
a distinct network address. In this way, a single quantum address
can represent a set of quantum nodes, independently of their physical
or topological location, inherently supporting compactness of routing
tables.
5.1. Quantum Packet
In the entanglement-packet switching paradigm, packet forwarding
relies on the manipulation of shared entanglement resources across
network nodes. Therefore the QA model requires a corresponding
quantum packet structure that supports quantum-native forwarding and
routing operations.
Cacciapuoti, et al. Expires 22 October 2026 [Page 11]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
5.1.1. Quantum Packet Structure
A quantum packet consists of a quantum header and a quantum payload.
* Quantum Header: carries quantum addresses that enable network
nodes to interpret and forward (in the generalized sense) quantum
packets according to the quantum-routing logic. The header may
also carry additional control information [CacCal26].
* Quantum Payload: carries the entanglement qubits |e_i>
constituting the network resource manipulated by packet-processing
operations. Depending on the underlying logic, such resources may
be consumed, transformed, stored, or left unchanged. The payload
is not required to be strictly tied to the immediate communication
objective identified by the header, and may also include
entanglement resources maintained to support subsequent operations
or network-level optimization [CacCal26].
The following ASCII diagram illustrates the conceptual structure of a
quantum packet for documentation only. The model is not limited to
bipartite entanglement.
+---------------------------------+----------------------------------+
| Quantum Header | Quantum Payload |
+---------------------------------+----------------------------------+
| - Quantum Address |A> | - Entangled qubits |e_i> |
| - Optional metadata | (bipartite or multipartite) |
+---------------------------------+----------------------------------+
6. Generalized Quantum Forwarding (GQF)
End-to-end entanglement distribution can be logically divided into
two distinct phases: routing and forwarding [AbaCub25]. Routing
determines the entanglement path (or, more generally, an entanglement
graph), according to the selected routing metric, while forwarding
performs the quantum operations on the entangled resources required
to sustain quantum connectivity.
In classical networks, the forwarding logic follows a match-and-
forward paradigm, where the destination address is extracted from the
packet header and matched against the routing table. In the Quantum
Internet, this logic is generalized toward entanglement manipulation,
enabling forwarding decisions that act directly on quantum states in
accordance with quantum-native principles.
Cacciapuoti, et al. Expires 22 October 2026 [Page 12]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
6.1. Role within the Architecture
Within the architecture, forwarding operations arise from the
interaction between the QDP and the QCP, through the EDCs and ESPs:
* QCP provides control support for forwarding, by maintaining
awareness of network connectivity, tracking entanglement
resources, and by enforcing policies across ESPs.
* ESP performs the quantum operations required for forwarding based
on locally available entanglement resources, packet-carried
control information, and, when available, policies provided by the
QCP.
Forwarding decisions require the capability to operate directly on
quantum identifiers. This is enabled by the quantum header in the
packet, which carries the quantum equivalent of the source address
and destination address.
7. Quantum-Native Principles
This section extends the architectural principles provided in
[RFC9340] by introducing a set of quantum-native principles that
guide the design and operation of a scalable Quantum Internet. These
principles reflect the physical properties of entanglement and the
architectural requirements arising from entanglement-driven
networking.
* Entanglement Packet Switching.
The architecture adopts an entanglement-packet switching paradigm,
in which entangled bits (ebits) serve as the fundamental network
units. These ``quantum packets'' carry quantum correlations
across network nodes.
* Explicit Plane Decoupling.
The architecture explicitly separates the QDP from the QCP. This
decoupling is necessary for scalability, since the stateful and
non-local nature of entanglement resources requires in-network
operations and persistent state awareness.
* Quantum Addressing.
The network logic adheres to a quantum-native control model
grounded in the quantum addressing (QA), which serves as its
logical abstraction.
* Stateful Core Network, Lightweight Edges.
The network core -- formed by ESPs -- is inherently stateful.
Conversely, edge nodes remain lightweight.
Cacciapuoti, et al. Expires 22 October 2026 [Page 13]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
* Entanglement-Aware Metrics.
Routing and orchestration decisions rely on entanglement-aware
metrics, including fidelity, residual coherence time in quantum
memories, entanglement purification overhead, and entanglement
availability.
* Hybrid Control Coexistence.
The architecture must support the coexistence of classical and
quantum control planes.
8. Routing Models and Repeater Realizations Support
The architectural framework defined in this document is agnostic to
both the specific routing model and quantum repeater generation
adopted in an implementation. The abstractions introduced by the
QDP, the QCP, and QA are designed to remain valid across different
operational regimes of entanglement generation and across different
physical realizations of quantum repeaters.
8.1. Routing-Model Agnosticism
The QCP orchestrates entanglement resources while the QDP executes
the underlying quantum operations. This separation naturally
accommodates different routing models, including regimes in which
entanglement generation occurs proactively (i.e., independently of a
specific request) as well as regimes in which it is triggered
reactively in response to explicit service requests.
In quantum routing, control decisions may precede the actual
availability of entanglement. This stems from the ephemeral nature
of entanglement resources, which may be consumed by operations such
as entanglement swapping or may decohere over time. Accordingly, the
architecture supports two complementary routing views:
* a physical-proximity view, driving hop-by-hop entanglement
generation and distribution; and
* an entanglement-proximity view, reflecting the currently available
entanglement connectivity and enabling routing and forwarding
decisions.
When a service requires entanglement connectivity, the QCP -- through
the EDCs -- may either exploit already-established entanglement links
or trigger additional link-level generation based on the physical-
proximity view. The resulting entanglement resources are then
reflected in the entanglement-proximity view maintained by the
control plane. This mechanism applies independently of whether
entanglement generation is proactive or reactive.
Cacciapuoti, et al. Expires 22 October 2026 [Page 14]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
QA plays a key role in enabling this flexibility. QA is not
envisioned as a routing-only mechanism and does not require
preexisting entanglement. Rather, QA provides a quantum-native
control abstraction that enables the QCP to represent and operate on
sets of candidate nodes, paths, or domains in a compact manner. In
particular, QA enables reasoning over potential connectivity before
entanglement resources are instantiated, supporting scalable
orchestration decisions such as which links to activate, which paths
to provision, or which domains to involve in distributed
coordination.
GQF operates on the basis of the quantum identifiers encoded in the
quantum packet header, allowing forwarding decisions to adapt
naturally to different routing regimes. For example, forwarding may
exploit already available entanglement resources as indicated by the
entanglement-proximity view, or trigger additional entanglement-
generation procedures guided by the physical-proximity view when such
resources are unavailable. In both cases, the quantum header and QA
abstractions provide the necessary information for GQF to apply the
appropriate sequence of operations on the available entanglement
resources.
8.2. Repeater-Generation Agnosticism
Within this architecture, a repeater is treated primarily as a
network function rather than as a fixed network entity. Accordingly,
the architectural tenets remain invariant across repeater
generations, while the physical mechanisms implemented in the QDP may
evolve with technological progress.
In first- and second-generation repeater networks, where bipartite
entanglement generation, purification, and swapping are the dominant
primitives, QA naturally indexes and operates on entanglement
resources and entanglement-based connectivity. The entanglement-
proximity view maintained by the QCP reflects the dynamic overlay
formed by these resources, while GQF implements the corresponding
sequence of swapping and forwarding operations.
In third-generation repeater networks, where quantum error correction
and logical qubit transmission may become prevalent, the operational
primitives evolve, and shared elementary link-level entanglement may
no longer be the dominant abstraction. In this regime, QA continues
to provide a scalable control abstraction, but the entities it
represents evolve from individual entanglement resources toward
logical entanglement resources. Nevertheless, the architectural
separation between QDP and QCP, the role of EDCs in orchestration,
and the forwarding logic expressed through GQF remain unchanged.
Cacciapuoti, et al. Expires 22 October 2026 [Page 15]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
9. Multi-Domain Routing and Forwarding
While the future deployment structure of the Quantum Internet remains
uncertain, it is reasonable to consider the possibility that, as the
network scales, it may evolve toward a multi-domain environment
composed of independently operated infrastructures, similarly to the
classical Internet. If such an evolution occurs, inter-domain
routing and coordination would naturally arise. The architectural
tenets described in this document are sufficiently flexible to
accommodate such a scenario.
The following section therefore illustrates, as an example of the
flexibility of the architectural abstractions, how the separation
between QDP, QCP, and QA can support intra-domain and inter-domain
routing and forwarding in a multi-domain deployment. This discussion
is intended to be illustrative rather than prescriptive.
9.1. Intra- and Inter-Domain Routing
In a large-scale deployment composed of multiple administrative
domains, routing functionality naturally separates into intra-domain
and inter-domain components, similarly to the distinction between
Interior Gateway Protocols (IGPs) and Exterior Gateway Protocols
(EGPs) in the classical Internet.
Within each administrative domain, the QCP maintains the information
required to populate ESP routing tables. This intra-domain routing
function maintains a consistent view of local ESP connectivity and
distributes the corresponding forwarding policies across the domain.
Across domains, routing information is exchanged between boundary
ESPs belonging to different administrative domains. Because inter-
domain routing spans independently operated networks, it cannot rely
on a single centralized controller [CalCac25]. Instead, routing
information propagates through distributed coordination between
domain-level controllers and gateway ESPs.
Mechanisms inspired by classical inter-domain routing, such as BGP,
may be envisioned to exchange reachability information and candidate
paths across domains while preserving routing autonomy [Liu2024QBGP].
However, in the quantum networking context, routing decisions must
also account for the probabilistic nature of entanglement generation
and the temporal fluctuations in the availability of entanglement
resources. This aligns naturally with the abstractions introduced in
this document, where routing information may reference sets of
candidate forwarding targets rather than a single deterministic next
hop, allowing the control plane to reason about multiple potential
nodes, paths and domains simultaneously.
Cacciapuoti, et al. Expires 22 October 2026 [Page 16]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
The QCP therefore combines two complementary roles:
* maintaining intra-domain and inter-domain entanglement-
connectivity graph and populating ESP routing tables, and
* coordinating with peer controllers and/or gateway ESPs to exchange
physical and entanglement reachability information across domains.
9.2. Quantum Addressing
As discussed in the previous section, rather than associating
forwarding entries strictly with individual nodes, QA enables routing
information to reference sets of nodes or candidate forwarding
targets through quantum-state representations.
This capability is particularly useful when the control plane must
decide which resources should be maintained or generated, and how
forwarding logic should be configured, even if end-to-end
entanglement has not yet been established. In this sense, QA
provides a quantum-native abstraction to reason about entanglement
connectivity and control actions independently of when entanglement
generation occurs.
In the context of inter-domain routing, QA can represent reachable
nodes, domains, or established entanglement in a compact and
expressive form and support coordination among federated or
distributed EDC instances when disseminating reachability information
or provisioning candidate paths.
9.3. Quantum Forwarding
Once routing information has been installed at ESPs, forwarding
decisions are executed locally through GQF in the QDP. Instead of
matching classical address prefixes, ESPs evaluate forwarding rules
defined over the quantum identifiers contained in the quantum packet
header. These rules determine how the ESP should manipulate locally
available entanglement resources to progress the forwarding
operation. Upon receiving a forwarding request, an ESP consults its
routing table and selects the appropriate next hop according to the
installed policies. The ESP then performs the quantum operations
required to extend entanglement toward that next hop using locally
available resources. These operations may include generating
elementary link-level entanglement with the next ESP or performing
entanglement swapping using stored entangled qubits.
Cacciapuoti, et al. Expires 22 October 2026 [Page 17]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
10. Security Considerations
As an Informational document, this draft does not propose any
specific mechanisms to ensure security. The security considerations
provided in RFC9340 apply for this document as well.
11. IANA Considerations
This memo includes no requests to IANA.
12. Acknowledgments
This document is based on work funded by the European Union under
Horizon Europe ERC-CoG grant QNattyNet, n.101169850. Views and
opinions expressed are however those of the author(s) only and do not
necessarily reflect those of the European Union or the European
Research Council Executive Agency. Neither the European Union nor
the granting authority can be held responsible for them.
13. Informative References
[RFC9340] Kozlowski, W., Wehner, S., Van Meter, R., Rijsman, B.,
Cacciapuoti, A. S., Caleffi, M., and S. Nagayama,
"Architectural Principles for a Quantum Internet",
RFC 9340, DOI 10.17487/RFC9340, March 2023,
<https://www.rfc-editor.org/rfc/rfc9340>.
[QI-MULTIPLANE]
Lopez, D., Martin, V., Lopez, B., and L. M. Contreras, "A
Multiplane Architecture Proposal for the Quantum
Internet", Work in Progress, Internet-Draft, draft-irtf-
qirg-qi-multiplane-arch-01, February 2026,
<https://datatracker.ietf.org/doc/draft-irtf-qirg-qi-
multiplane-arch/01/>.
[CalCac25] Caleffi, M. and A. S. Cacciapuoti, "Quantum Internet
Architecture: unlocking Quantum-Native Routing via Quantum
Addressing", IEEE Transactions on Communications 74,
DOI 10.1109/tcomm.2025.3650397, 2026,
<http://dx.doi.org/10.1109/TCOMM.2025.3650397>. Invited
Paper.
[CacCal26] Cacciapuoti, A. S. and M. Caleffi, "A Quantum Internet
Protocol Suite: Beyond Layering", IEEE Transactions on
Network Science and Engineering 2026,
DOI 10.1109/TNSE.2026.3679795, 2026,
<https://doi.org/10.1109/TNSE.2026.3679795>. Invited
Paper.
Cacciapuoti, et al. Expires 22 October 2026 [Page 18]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
[IllCal22] Illiano, J., Caleffi, M., Manzalini, A., and A. S.
Cacciapuoti, "Quantum Internet Protocol Stack: a
Comprehensive Survey", August 2022,
<https://www.sciencedirect.com/science/article/abs/pii/
S1389128622002250>.
[RFC1958] Carpenter, B., "Architectural Principles of the Internet",
RFC 1958, 1996, <https://www.rfc-editor.org/rfc/rfc1958>.
[RFC3439] Bush, R. and D. Meyer, "Some Internet Architectural
Guidelines and Philosophy", RFC 3439, 2002,
<https://www.rfc-editor.org/rfc/rfc3439>.
[AbaCub25] Abane, A., Cubeddu, M., Mai, V. S., and A. Battou,
"Entanglement routing in quantum networks: A comprehensive
survey", IEEE Transactions on Quantum Engineering 2025,
2025, <10.1109/TQE.2025.3541123>.
[KreRam14] Kreutz, D., Ramos, F. M. V., Verissimo, P. E., Rothenberg,
C. E., Azodolmolky, S., and S. Uhlig, "Software-defined
networking: A comprehensive survey", Proceedings of the
IEEE 2014, 2014, <10.1109/JPROC.2014.2371999>.
[yang2004rfc3746]
Yang, L., Dantu, R., Anderson, T., and R. Gopal,
"Forwarding and Control Element Separation (ForCES)
Framework", 2004.
[KurRos10] Kurose, J. and K. Ross, "Computer Networks: A Top-Down
Approach", 2010.
[KriCla07] Krioukov, D., Claffy, K. C., Fall, K., and A. Brady, "On
compact routing for the Internet", SIGCOMM Computer
Communication Review 37, 2007,
<https://doi.org/10.1145/1273445.1273450>.
[Liu2024QBGP]
Liu, M., Li, Z., Cai, K., Allcock, J., Zhang, S., and J.
C. S. Lui, "Quantum BGP with Online Path Selection via
Network Benchmarking", IEEE INFOCOM 2024,
DOI 10.1109/INFOCOM52122.2024.10621359, 2024,
<https://doi.org/10.1109/INFOCOM52122.2024.10621359>.
Authors' Addresses
A. S. Cacciapuoti
University of Naples Federico II
Email: angelasara.cacciapuoti@unina.it
Cacciapuoti, et al. Expires 22 October 2026 [Page 19]
Internet-Draft Quantum-Native Architectural Tenets and April 2026
M. Caleffi
University of Naples Federico II
Email: marcello.caleffi@unina.it
J. Illiano
University of Naples Federico II
Email: jessica.illiano@unina.it
C. De Risi
University of Naples Federico II
Email: caterina.derisi@unina.it
A. Abane
National Institute of Standards and Technology
Email: amar.abane@nist.gov
J. Chung
Argonne National Laboratory -- Data Science and Learning Division
Email: chungmiranda@anl.gov
Cacciapuoti, et al. Expires 22 October 2026 [Page 20]