Teleportation-Based Scheme

• Teleportation-based schemes are protocols that use pre-shared entanglement, joint measurements, and classical communication to remotely implement quantum operations like state transfer, gate application, filtering, and amplification.
• They encompass diverse variants—including correction-based, port-based, gate, and continuous-variable protocols—that strategically relocate operational complexity from direct interactions to resource preparation and conditional decoding.
• Practical implementations span optical systems, cavity-QED setups, and fault-tolerant quantum computing, emphasizing trade-offs between entanglement cost, fidelity, and error sensitivity.

A teleportation-based scheme is a protocol in which the operational task—state transfer, gate application, filtering, amplification, transduction, or encrypted evaluation—is realized through shared entanglement, a joint measurement, and classical side information, rather than by direct physical transmission or direct interaction alone. In the literature, this designation covers correction-based Bennett-style teleportation, correction-free port-based teleportation, gate teleportation, continuous-variable feed-forward protocols, and hybrid constructions that deliberately redistribute where complexity, channel knowledge, or correction effort resides (Mozrzymas et al., 2017, Wu et al., 2021, Brun et al., 2015).

1. Operational structure and defining components

At the most abstract level, teleportation-based schemes separate a target transformation into three layers: resource preparation, measurement, and conditional decoding. In standard discrete-variable teleportation this takes the familiar form of a shared maximally entangled pair, a Bell-state measurement, and a Pauli correction on the receiver. In continuous-variable teleportation, the same logic appears as a shared entangled Gaussian resource, a Bell-type homodyne measurement, and a feed-forward displacement. In gate-teleportation settings, the gate is precompiled into the resource or measurement basis, so the output appears as the desired transformed state up to known byproduct operators (Gao et al., 2010, Fulek et al., 2018, Wu et al., 2021).

The cryptographic formulation of this structure is explicit in the encrypted-gate primitive

$$EG[U]:|\alpha\rangle \rightarrow \big((a,b),X^a Z^b U|\alpha\rangle\big),$$

where the byproduct Pauli is not an inconvenience but the encryption itself. In the optical gate-teleportation formulation of Gottesman and Chuang, the same principle appears as a $U$-rotated Bell basis: a Bell measurement on the input and part of a pre-entangled ancilla teleports the input through a gate already embedded in the resource, leaving only known single-qubit Pauli corrections on the output (Fulek et al., 2018, Gao et al., 2010).

Continuous-variable schemes instantiate the same architecture with quadratures. In microwave–optical transduction, the input mode and one half of a microwave–optical entangled state interfere on a $50/50$ beam splitter, the quadratures $q_-$ and $p_+$ are measured, and the optical mode is displaced with gain $\kappa$. Depending on $\kappa$, the effective channel is a thermal attenuator, a thermal amplifier, or, at $\kappa=1$, an additive white Gaussian noise channel (Wu et al., 2021). In the asymptotic port-based quantum correction teleportation family, Bob performs a classical selection task followed by a quantum correction, making explicit that “which port?” and “which unitary?” can be treated as separate decoding layers rather than mutually exclusive designs (Kim et al., 2024).

2. Probabilistic transfer, asymmetric information, and compressed state classes

A prominent discrete-variable theme is the deliberate redistribution of channel knowledge. In probabilistic teleportation through a non-maximally entangled channel

$|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$

earlier schemes required Bob to know $(a,b)$ in order to choose a parameter-dependent recovery unitary. The sender-only scheme instead moves the “entanglement matching” to Alice’s side by a generalized measurement $U$0, leaving Bob with only Pauli-frame corrections $U$1. The total success probability remains

$U$2

so the novelty is not improved efficiency but elimination of Bob’s need to know the channel parameters (Wei et al., 2012).

The same paper uses an assistant, Charlie, to exhibit two distinct communication patterns. When only Charlie knows both partially entangled channels, the total process decomposes into a conventional hop followed by the new sender-knows-only hop, with total success probability

$U$3

When Charlie knows neither channel, Alice uses the sender-knows-only protocol on the first hop and Bob uses the conventional protocol on the second, again with

$U$4

The first scenario makes Charlie an active controller who alone knows whether the full transmission succeeded; the second turns him into a passive relay (Wei et al., 2012).

A different form of structural simplification appears in bidirectional teleportation of GHZ-like states. For input families

$U$5

local CNOT compression reduces each $U$6-qubit state to one unknown qubit tensored with $U$7 registered qubits $U$8. The actual bidirectional task then becomes two ordinary one-qubit teleportations, and two Bell states suffice for the whole exchange. This replaces the six-qubit cluster-state resource used by Zhou et al. for a representative asymmetric task with Bell-pair channels, Bell-state measurements, and Pauli corrections only (Sisodia, 2021).

3. Port-based, recoverable, and asymptotic interpolation schemes

Port-based teleportation (PBT) is the canonical teleportation-based scheme in which Bob does not apply a correction unitary to the selected output system. In deterministic PBT, Alice performs a POVM on the input and her halves of $U$9 shared entangled pairs, sends an outcome $50/50$0, and Bob simply keeps port $50/50$1. The full jointly optimized deterministic problem is solved by the teleportation matrix $50/50$2: the optimal fidelity is

$50/50$3

and for $50/50$4,

$50/50$5

The same Perron eigenvector determines both the optimal resource state and the optimal POVM (Mozrzymas et al., 2017).

The price of this correction-free receiver is large entanglement and imperfect finite-$50/50$6 performance. That limitation is sharpened by two later perspectives. First, recoverable PBT shows that PBT and standard teleportation are not disjoint extremes: for qubits with odd $50/50$7, the failure branch of probabilistic PBT can be measured in a basis that makes Bob’s total $50/50$8-port system recoverable by a global unitary, while the success branches retain the usual single-port interpretation. In parallel, port-based superdense coding yields the bound

$50/50$9

for the success probability of probabilistic PBT, and monogamy/asymmetric-cloning arguments show that deterministic PBT cannot converge to perfect fidelity faster than order $q_-$0 (Ishizaka, 2015).

Second, recycling analysis studies what happens if the unused ports are retained for later rounds. For deterministic PBT, the one-round recycled-resource fidelity can be written explicitly in group-theoretic data for arbitrary $q_-$1, and multi-round degradation satisfies

$q_-$2

A striking outcome is that optimizing single-shot teleportation fidelity does not necessarily improve recyclability: in the qubit case the optimal resource can even be slightly worse for recycling than the non-optimal one (Studziński et al., 2021).

The asymptotic PBQCT family makes the interpolation between standard teleportation and PBT explicit at the level of joint measurements. For qudits, signal states of the form

$q_-$3

encode both a port label $q_-$4 and a generalized Bell label $q_-$5. When there is only one signal state per port, the decoding reduces to PBT-style selection. When all $q_-$6 generalized Bell states per port are used, the resulting protocol is a parallelized form of standard teleportation. Intermediate signal sets define genuinely mixed decoding rules in which Bob first selects a port and then applies a correction (Kim et al., 2024).

4. Optical, cavity-QED, and photonic gate-teleportation realizations

Several teleportation-based schemes are built around cavity reflection and photonic interference rather than abstract Bell measurements. In low-$q_-$7 cavity QED, photonic Faraday rotation can replace a direct Bell analyzer. An arbitrary multipartite atomic state is teleported by entangling Bob’s atoms with linearly polarized photons, sending the photons to Alice, reflecting them from cavities containing Alice’s atoms, and then applying local Hadamards and product-basis measurements. The Bell-state information is converted into photon–atom product outcomes, and Bob’s recovery is always a tensor product of single-qubit Pauli operators. The protocol is deterministic in logical structure, works in the virtual-excitation regime, and the paper estimates an overall success rate scaling as

$q_-$8

with about three hours required for a successful two-qubit teleportation at $q_-$9 photons per second under the stated detector assumptions (Chen et al., 2010).

A related long-distance atomic protocol uses entangled coherent states (ECS) and one-sided atom–cavity systems as state-dependent mirrors. Reflection implements

$p_+$0

and the effective Bell discrimination is reduced to ON/OFF detection after beam-splitter mixing with ancillary coherent states. The single-round success probability is

$p_+$1

while failure is nondestructive: $p_+$2 Because Alice can refrain from measuring her atom in the failure branch, the repeated-attempt success probability becomes

$p_+$3

For $p_+$4, one-shot performance is already near unit success with unit fidelity (Mishra et al., 2012).

Gate teleportation in linear optics was demonstrated directly in the Gottesman–Chuang sense. One experiment used a six-photon interferometer to realize a teleportation-based CNOT gate: two input polarization qubits were Bell-measured against a four-photon cluster-like resource $p_+$5, and the remaining two photons carried the CNOT output up to Pauli byproducts. The measured truth-table fidelity was

$p_+$6

and entangling capability was verified by producing an output Bell state with fidelity

$p_+$7

A second experiment used four-photon hyper-entanglement to realize a teleportation-based controlled-phase gate with complete single-photon Bell-state analysis, obtaining $p_+$8, $p_+$9, a process-fidelity bound $\kappa$0, and a concurrence lower bound $\kappa$1 (Gao et al., 2010).

5. Continuous-variable, multimode, and transduction-based variants

Continuous-variable teleportation has been used not only for state transfer but also for resource engineering. One modification inserts a cubic phase gate

$\kappa$2

after displacing one squeezed resource mode by $\kappa$3. In the resulting CZ-based teleportation circuit, the $\kappa$4-quadrature error remains

$\kappa$5

but the $\kappa$6-quadrature error becomes approximately

$\kappa$7

For $\kappa$8 dB squeezing and $\kappa$9, the modified scheme beats the original one for $\kappa$0; in the vacuum-state example, $\kappa$1 yields $\kappa$2 with postselection, and $\kappa$3 yields $\kappa$4 in the most probable region without postselection (Zinatullin et al., 2021).

A second CV example uses teleportation to implement noiseless amplification. The desired probabilistic filter $\kappa$5 is encoded into a non-Gaussian two-mode entangled resource, then applied by CV teleportation with postselection on Bell outcomes near $\kappa$6. For the weak-state amplifier

$\kappa$7

the pure-state teleportation action is

$\kappa$8

showing that finite squeezing attenuates the input by $\kappa$9 before amplification. The paper’s main technical point is that, because of the photon-number correlations of the two-mode squeezed vacuum, the required non-Gaussian resource can be prepared using only photon subtraction and auxiliary Gaussian squeezed states, rather than direct photon addition on the signal mode (Fiurášek, 2021).

Teleportation can also be reinterpreted as a transduction architecture. In microwave–optical conversion, the electro-optic device is used as an entanglement generator rather than a direct beam-splitter converter. The teleportation channel then has added thermal noise

$\kappa=1$0

while the direct-conversion benchmark has a nonzero quantum-capacity threshold

$\kappa=1$1

in the ideal extraction-efficiency limit. The teleportation-based scheme has a nonzero quantum-capacity lower bound for arbitrarily small cooperativity, and the paper reports higher fidelity or success probability for coherent, cat, and GKP-state transduction in the low-$\kappa=1$2 regime (Wu et al., 2021).

Finally, multimode CV teleportation has been extended to the “complete state of light” by decomposing an unknown coherent field into pixel modes,

$\kappa=1$3

and teleporting each pixel mode independently with multimode two-mode squeezing. For one pixel the average fidelity is

$\kappa=1$4

and for an $\kappa=1$5-pixel image

$\kappa=1$6

In the realistic SPDC model, the effective squeezing $\kappa=1$7 is ring-shaped across the image plane, so teleportation quality depends on spatial position as well as squeezing strength (Permaul et al., 25 Feb 2025).

6. Computation, cryptography, control, and sensing

Teleportation-based schemes are central in fault-tolerant and delegated quantum information processing because they convert hard inline operations into measurement and ancilla-preparation tasks. In multi-qubit CSS fault tolerance, logical Clifford operations are implemented by logical-operator measurement inside large memory blocks, while logical qubits requiring a non-Clifford gate are teleported into a processor code with a transversal $\kappa=1$8 gate and then teleported back. The architecture uses memory blocks $\kappa=1$9, processor blocks $|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$0, and Steane-style syndrome extraction throughout; the authors explicitly present the scheme as universal FTQC without magic-state distillation, with all circuits for computation and error correction transversal (Brun et al., 2015).

In quantum homomorphic encryption, gate teleportation is used to hide non-Clifford correction structure. The primitive

$|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$1

lets the client retain the byproduct key while the server continues on Pauli-encrypted data. This is crucial for $|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$2 because

$|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$3

introduces a $|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$4-error depending on a hidden $|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$5-key bit. Using $|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$6, the paper constructs two non-interactive, perfectly secure QHE schemes, GT and VGT. Their decryption complexity depends on the total number $|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$7 of $|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$8 gates; VGT is proved optimal with $|\psi_{23}\rangle=a|0_20_3\rangle+b|1_21_3\rangle,$9-quasi-compactness (Fulek et al., 2018).

Teleportation-based CV cryptography uses the protocol itself as the key-generation mechanism. Alice chooses from the coherent-state alphabet $(a,b)$0, groups it into real and imaginary bases, performs a teleportation Bell measurement, and Bob chooses between two gain settings corresponding to his basis guess. After a final displacement, Bob assigns bit $(a,b)$1 to vacuum and bit $(a,b)$2 to any light. Against the incoherent beam-splitting attack studied in the paper, the protocol is deterministic, uses direct reconciliation, requires no postselection, and can generate a secure key beyond the usual $(a,b)$3 loss point, although the key rate vanishes exactly at $(a,b)$4 where Bob and Eve are symmetric (Luiz et al., 2014).

Control and metrology applications further broaden the term. In QCECT for satellite control and communication, a qubit

$(a,b)$5

encoding satellite coordinates is first protected by a 3-qubit repetition code, teleported through a 5-qubit controlled scheme, rotated on the ground by $(a,b)$6 and $(a,b)$7, and then teleported back; the protection is limited to single-qubit bit-flip errors (Thacker et al., 2015). In gravitational-wave detectors, continuous-variable teleportation is used as a substitute for long low-loss filter cavities: in the ideal limit of perfect EPR entanglement, teleportation-based post-filtering reproduces conventional frequency-dependent post-filtering, but for realistic Einstein Telescope parameters finite EPR noise, readout loss, and phase error make post-filtering nearly equivalent in sensitivity to teleportation-based pre-filtering (Nishino, 25 Sep 2025).

Taken together, these works indicate a recurring tradeoff. Teleportation-based schemes often remove a direct interaction, a trusted receiver-side correction, or a bulky filtering element, but the cost is shifted into more demanding ancilla states, enlarged measurement spaces, deferred corrections, sender-side generalized measurements, or stronger sensitivity to finite squeezing and loss. This suggests that “teleportation-based” should be understood not as a single protocol family with fixed operational meaning, but as a design principle for relocating where quantum complexity is paid.