Entanglement-Enabled Communication

Updated 1 February 2026

Entanglement-enabled communication is a paradigm where shared quantum entanglement is exploited to achieve higher success rates and novel functionalities beyond classical limits.
Adaptive protocols and structured resource hierarchies enable simulation of broader quantum correlations, validating performance improvements in both finite and asymptotic regimes.
Experimental demonstrations and scalable finite-blocklength codes confirm that entanglement can reduce error rates and enhance capacities in realistic noisy and lossy channels.

Entanglement-enabled communication refers to communication protocols wherein entanglement acts as a resource to enhance classical or quantum information transmission capabilities beyond what is possible with classical resources or unentangled quantum protocols. The operational paradigm typically assumes that the communicating parties share entangled quantum states prior to the communication act; this resource is then leveraged—often in conjunction with adaptive local measurements and classical/quantum signaling—to achieve elevated success probabilities, transmission rates, or even new functionalities that are fundamentally unreachable classically. The landscape of entanglement-enabled communication encompasses assistive roles in zero-error communication, resource inequality hierarchies, experimental realizations over bosonic and multi-mode channels, theoretical extensions to non-causal processes, and protocols requiring only local measurements. The resulting advantages are rigorously characterized for both finite and asymptotic blocklengths, as well as for point-to-point, multiparty, and networked communication settings.

1. Foundational Protocol Structures and Adaptive Measurement

The canonical architecture for entanglement-enabled communication involves two parties (often designated Alice and Bob) sharing an entangled state ρ{AB}, usually a pure bipartite state |ψ⟩{AB}. For a typical classical communication task, Alice receives input x and Bob input y from finite sets, and Bob must output b. In the standard, non-adaptive protocol, the communication proceeds via: (i) Independent local measurements on their respective subsystems, producing measurement results m (Alice) and b' (Bob); (ii) a classical communication round where Alice sends m to Bob; (iii) Bob produces the output b according to a deterministic function of m and b'. The mathematical structure is characterized by local Positive Operator-Valued Measures (POVMs):

$p(b|x,y) = \sum_{m=1}^D \sum_{\vec{b}: b_m = b} \langle\psi|A_{m|x}\otimes B_{\vec{b}|y}|\psi\rangle,$

with commutation constraints on the measurement operators that define the non-adaptive regime.

The adaptive protocol generalizes this by allowing Bob’s measurement to depend explicitly on Alice’s communicated value m. Upon receipt of m, Bob locally selects a measurement from a family {B_{b|y,m}}, producing outcome b:

$p(b|x,y) = \sum_{m=1}^D \langle\psi|A_{m|x}\otimes B_{b|y,m}|\psi\rangle.$

This adaptive mechanism strictly expands the class of attainable input-output correlations. In Random Access Codes (RACs), for example, the best non-adaptive entanglement-assisted trit yields a maximal success probability $p_{\text{win}}^{\text{NA EA trit}} \leq 0.9082$ , while the adaptive protocol achieves $p_{\text{win}}^{\text{A EA trit}} = (1/4)(3+1/\sqrt{2}) \approx 0.9268$ , a strict improvement (Pauwels et al., 2022).

2. Resource Hierarchies and Comparative Strength

Entanglement-enabled protocols yield strict resource hierarchies. In the prepare-and-measure scenario, it is established that:

With a single entanglement-assisted bit (EA bit) and adaptive measurements, all qubit correlations (including those involving general POVMs) can be simulated.
With a non-adaptive EA bit, only projective qubit correlations are attainable, and not the full set of qubit output distributions.

This establishes the hierarchy (using the bracket notation “[]” and an arrow for resource superiority):

$[\text{EA qubit}] \Rightarrow [\text{EA bit}] \Rightarrow [\text{qubit}] \Rightarrow [\text{bit}]$

where strict superiority holds at each level except between EA qubit and EA bit, where adaptive simulation equalizes the power for certain tasks. In dense coding, for instance,

$[q \rightarrow q] + [qq] \succeq 2[c \rightarrow c]_{\text{EA}}$

and, critically, an adaptive EA bit can simulate a quantum channel, but not in the non-adaptive setting (Pauwels et al., 2022).

3. Asymptotic Capacity, Zero-Error Communication, and Channel Separation

Entanglement can boost asymptotic communication rates, especially in the zero-error context. For classical channels described by a confusability graph $G$ , the independence number $\alpha(G)$ defines the unassisted one-shot zero-error capacity. In the entanglement-assisted setting, the capacity $C_0^E(G)$ can strictly exceed $C_0(G)$ . Explicitly, for the graph $sp(6,F_2)$ (the orthogonality graph of the $E_7$ root system), $C_0(G) = \log 7$ while $C_0^E(G) = \log 9$ , a strict separation. Further, the entanglement-assisted zero-error capacity in the limit of many channel uses can reach the ordinary Shannon capacity, closing the gap between zero-error and arbitrarily small error communication rates (Leung et al., 2010).

A notable case is the multiple-access channel defined by shared environmental variables and adversarial jamming. There exist instances where, without entanglement, the capacity region collapses to zero irrespective of shared randomness, while with entanglement, strictly positive rates are possible—thus, entanglement enables communication where none is feasible classically (Nötzel, 2019).

4. Experimental Demonstrations and Capacity Surpassing

Controlled experiments have verified entanglement-enabled communication advantages in both discrete and continuous-variable regimes:

For lossy and noisy bosonic channels, implementation of entanglement-assisted communication (EACOMM) using a two-mode squeezed vacuum (TMSV) source and phase-conjugate receiver (PCR) has yielded mutual information rates up to 14.6% above the Holevo-Schumacher-Westmoreland (HSW) capacity, with bit error rates reduced by up to 69% under the same power constraints (Hao et al., 2021). The experimental protocol leverages the phase-sensitive cross-correlation, which survives even when entanglement is broken by high channel loss or noise.
In multi-mode fibers, the entanglement-assisted capacity, under a fixed power budget and increasing the number of spatial modes (by tuning refractive index contrast), grows logarithmically with the mode count $M$ while the classical capacity saturates. Asymptotically, the capacity ratio $C_E/C_S$ diverges as $\log M$ (Sekavčnik et al., 2022).

5. Protocol Scalability and Practical Implementations

Scaling entanglement-enabled protocols to high dimensions and finite blocklengths has been explored:

Protocols based on high-dimensional entanglement accomplish perfect stochastic message delivery (e.g., in a two-input, one-output selection task) for $n$ -dimensional entangled states, with experimentally demonstrated fidelity exceeding the highest possible for non-maximal entanglement, and with only separable measurements at the receiver (Zhang et al., 7 Feb 2025).
Explicit constructions of entanglement-assisted Reed–Solomon codes for finite blocklength are attainable by using mixed alphabets (field elements and their extensions), leveraging superdense coding segments. The tradeoff relation $R \leq (1-\delta)(1+e)$ (with $R$ the code rate, $\delta$ the erasure fraction, $e$ the per-use entanglement rate) interpolates smoothly between unassisted and fully entangled regimes (Prasad et al., 2023).

Product measurement protocols, as opposed to entangled joint measurements, suffice for achieving strict entanglement advantages in semi-device-independent scenarios, with every entangled two-qubit Werner state certified operationally as advantageous under symmetric task extensions (Bakhshinezhad et al., 2024).

6. Extensions to Quantum Causal Structure and Non-Distillable States

Entanglement resources extend beyond capacity enhancement to simulating classical communication with no predefined causal order. By leveraging LOCC protocols aided by sufficient shared entanglement, any acausal classical correlation (arbitrary classical communication structure) can be faithfully simulated, with entanglement cost depending on the structure's complexity. This demonstrates a duality between "entanglement + causally ordered communication" and "classical communication without causal order" (Akibue et al., 2016).

Remarkably, undistillable bound entangled states, which cannot be converted to pure maximally entangled pairs, can still provide a communication complexity advantage in multi-party tasks. Even with a fully bound tripartite state, the quantum protocol succeeds with a provably higher probability than any classical protocol for a given bit budget, as witnessed by a Bell inequality violation (Epping et al., 2012).

7. Implementation in Correlated Quantum Systems and Reference-Frame Independence

The short-range ground-state entanglement in strongly correlated quantum systems can be harnessed for both classical and quantum communication, obviating the need for long-range entanglement engineering. For instance, dimerized antiferromagnetic spin chains enable dense-coding-style classical communication and enhanced remote state preparation fidelities purely by local unitaries and natural spin dynamics (Yang et al., 2011).

In the absence of a shared spatial or polarization reference frame, the three SU(2)-invariant relational parameters of a two-qubit pure state (local Bloch vector lengths, relative angle, and phase) offer three distinct channels for frame-independent message encoding. The average information gain per use in such invariant-parameter communication is significantly enhanced by maximal entanglement relative to any separable protocol (Beheshti et al., 2019).

References

Adaptive protocols and resource hierarchy: (Pauwels et al., 2022)
Channel separation and asymptotic zero-error: (Leung et al., 2010, Nötzel, 2019)
Experimental capacity surpassing: (Hao et al., 2021, Sekavčnik et al., 2022)
High-dimensional and finite-blocklength codes: (Zhang et al., 7 Feb 2025, Prasad et al., 2023)
Simulating non-causal communication, non-distillable states: (Akibue et al., 2016, Epping et al., 2012)
Short-range entanglement and reference-frame independence: (Yang et al., 2011, Beheshti et al., 2019)
Product measurement protocols and semi-device-independent certification: (Bakhshinezhad et al., 2024)

This synthesis establishes that entanglement-enabled communication yields fundamental, certified, and often unbounded advantages across an array of physically and operationally motivated scenarios, with rigorous theoretical bounds and scalable architectures now converging with experimental realizations.