betaFeedback

WorksPhysics and Astronomy

Secret-shared teleportation of a two-qudit gate with a deferred recipient

Abstract

An unknown quantum operation, like an unknown quantum state, cannot be perfectly copied or stored for later use. A specific gate can be embedded in an entangled resource and teleported, but its destination is fixed when the resource is created. Here we show that the capability to apply a two-qudit entangling gate can instead be secret-shared among many parties: the shares are distributed in advance, and a later key release lets exactly one of them, chosen after distribution, exercise it. A provider encrypts half of a Bell pair held with a user into nn shares, one per candidate node. The selected node, given the key qudits, completes an exact non-local CSUMk\mathtt{CSUM}^{k} gate with the user's data, in every measurement branch and for any qudit dimension, with the power kk chosen by the user alone. Any coalition of fewer than nn candidate nodes, or the key register alone, holds no information, unselected nodes cannot tell whether a selection occurred, and the key cannot decrypt twice. For two candidates the registers form a ((3,5))((3,5)) threshold scheme. Encoder and decoder are explicit O(n)O(n) circuits. The construction supports jointly authorized and anonymous gate application, advancing distributed quantum computing and quantum network capabilities.

Quantum information cannot be copied [1], and neither can an unknown quantum operation: no finite quantum program can store an arbitrary unitary for later execution [2], and copies of a gate are necessarily imperfect [3]. What can be stored is a specific gate, frozen into an entangled resource and released later by local measurement and classical feedforward [4, 5, 6]. Gate teleportation of this kind is the elementary operation of distributed quantum computing and of a quantum internet [7, 8]: it applies a non-local entangling gate between data held at different nodes without moving the data, and it has been demonstrated between logical qubits, between separated trapped ions and across optical network links [9, 10, 11]. In every such protocol, however, the resource is bound to its destination when it is created. The half of the entangled pair that completes the gate is sent to one node, and whoever holds it holds the gate.

Secret sharing offers a different way to hold a resource. A threshold scheme splits a secret among nn parties so that only authorized subsets recover it and all other sets learn nothing [12]. Its quantum analog shares an unknown quantum state in the same way [13, 14, 15], with constructions based on graph states and on absolutely maximally entangled (AME) states [16, 17], states that are maximally entangled across every bipartition, and demonstrations with photons and optical modes, recently in eleven dimensions [18, 19]. In all of these, the shared object is a state. Authority over an operation has also been shared, in quantum operation sharing [20], controlled remote implementation of operations [21] and threshold-controlled teleportation [22]. There, however, the recipient is fixed in advance, and every controller must measure and announce an outcome, so each can veto the operation and all of them observe it. Here, we ask: can the capability to act, namely, applying an entangling gate to data that never enter the scheme, be secret-shared so that the node exercising it is chosen after distribution and the unselected sharers take no part?

A gate shared in this way would meet three requirements of an entanglement-based network: joint authorization, anonymity and early distribution. Authority to use a gate can be split among independent custodians, none of whom individually holds anything of value, as threshold cryptography does for classical keys. Moreover, nodes that are not chosen should learn neither the data nor whether anyone was chosen, a requirement shared with anonymous quantum communication [23] and with secure multiparty and delegated quantum computation [24, 25]. Because remote entanglement is slow to establish [11], the user’s entanglement should be in place before the destination is known. However, the obvious ways of deferring the destination fail at least one of these requirements. For example, a provider that shares a Bell pair with the user and later swaps its half onto the chosen node defers the choice, but the chooser already holds the gate. A Greenberger–Horne–Zeilinger (GHZ) state spread over the user and all candidates, as in quantum information secret sharing [13, 15], requires every unselected node to measure and announce its outcome, so any one of them can veto the gate, and it leaves each share classically correlated with the user (Supplementary Section S13).

Encrypted cloning [26] supplies the missing ingredient. It splits a qubit into nn signal copies, each indistinguishable from noise, any one of which can be decrypted once by whoever holds an entangled key. Qudit versions of encrypted cloning have recently been developed [27, 28], with Ref. [28] relating the construction to AME states and secret sharing. Here we apply encrypted cloning not to data but to one half of a gate-teleportation resource, and show that the result is a threshold quantum secret-sharing scheme for a non-local two-qudit gate. Specifically, a provider distributes encrypted shares to nn cancandidate nodes. Releasing nn key qudits later lets exactly one node complete an exact CSUM⁡\operatorname{CSUM} with the user, in every measurement branch, for any dimension dd and for data entangled with arbitrary external systems, while every coalition of fewer than nn candidates, or the key register alone, holds nothing. We derive the access structure of this encoder in closed form; for two candidates the five distributed registers form a ((3,5))((3,5)) threshold scheme, as anticipated in Ref. [28]. The user alone decides which power CSUM⁡k\operatorname{CSUM}^{k} is applied, and kk is not revealed in any transmitted register or public message. Encoder and decoder are explicit circuits of O(n)O(n) gates, independent of dd, and every claim is proved in closed form. We work with qudits because high-dimensional carriers are increasingly available [29, 30, 31] and bring higher capacity and noise tolerance to cryptographic tasks [32]. The construction is novel already for d=2d=2; for d>2d>2 it relies on two features of the arithmetic of the qudit Clifford group [33, 34] that are not present for qubits: a flat phase angle that depends on the parity of dd, which we show is necessary for the quadratic-phase encoder, and sign bookkeeping in the teleportation stage that makes every power of CSUM⁡\operatorname{CSUM} available.

Protocol

The protocol runs in three stages as depicted in Figure 1.

Deferred-recipient teleportation of a two-qudit gate

Figure 1. Deferred-recipient teleportation of a two-qudit gate. In stage 1, a maximally entangled pair ∣Φ⟩BABC|\Phi\rangle_{B_A B_C} is shared between Alice, who holds BAB_A, and the provider, who holds BCB_C; it may be prepared by either party. The provider also prepares nn Bell pairs ∣Φ⟩SjKj|\Phi\rangle_{S_j K_j} and applies the encrypted-cloning encoder UencU_{\mathrm{enc}} to BCB_C and the signal qudits S1,…,SnS_1,\ldots,S_n, then sends one signal to each of nn candidate sites and withholds the key qudits K1,…,KnK_1,\ldots,K_n. Each share alone carries no information, and the provider keeps nothing that can complete the gate. Alice’s data qudit DAD_A is untouched at this stage. The recipient (Bob) is chosen only after distribution in the next stage: releasing the key register to one site lets it apply UdecU_{\mathrm{dec}} and recover BCB_C, with its entanglement with BAB_A, at unit fidelity. All other shares (dashed, crossed) stay maximally mixed, their holders cannot tell that a selection was made, and the spent key cannot decrypt them. In the final stage, Alice applies CSUM⁡(DA→BA)\operatorname{CSUM}(D_A \to B_A) and Bob applies CSUB⁡(SB→DB)\operatorname{CSUB}(S_B \to D_B). One computational-basis and one Fourier-basis measurement, two publicly exchanged dits and the Weyl corrections XdmBX_d^{m_B} on DBD_B and ZdmAZ_d^{m_A} on DAD_A complete an exact CSUM⁡(DA→DB)\operatorname{CSUM}(D_A \to D_B) in every measurement branch, although DAD_A and DBD_B never interact directly. Replacing the two gates by CSUM⁡k1\operatorname{CSUM}^{k_1} and CSUB⁡k2\operatorname{CSUB}^{k_2} yields CSUM⁡k1k2\operatorname{CSUM}^{k_1k_2}.

Stage 1 — distributing the resource. A provider, who may be Alice herself, prepares a maximally entangled pair ∣Φ⟩BABC|\Phi\rangle_{B_A B_C} and gives BAB_A to Alice. By applying UencU_{\mathrm{enc}} it encodes BCB_C into nn signal qudits S1,…,SnS_1,\ldots,S_n, each initially Bell-paired with a key qudit KjK_j, sends SjS_j to candidate site jj and retains the key register (K1,…,Kn)(K_1,\ldots,K_n). Alice’s data qudit DAD_A plays no part and need not yet exist.

Stage 2 — one site redeems the key. At any later time and by any criterion external to the protocol, the provider selects a site, Bob, and releases the key register to that site alone. Bob applies a decoder UdecU_{\mathrm{dec}} to (SB,K1,…,Kn)(S_B,K_1,\ldots,K_n) and recovers BCB_C with its full entanglement with BAB_A, at fidelity exactly one: (BA,SB)(B_A,S_B) is then a perfect Bell pair between Alice and Bob. Every other share is left untouched and maximally mixed. Operational requirement: the key register must reach Bob intact over an authenticated quantum channel [35]; the keys alone carry no information, but an unselected site that obtained them could redeem the gate in Bob’s place.

Stage 3 — local operations complete the gate. Alice and Bob complete a non-local gate by local operations and one round of classical communication. Alice applies CSUM⁡(DA→BA)\operatorname{CSUM}(D_A \to B_A) and measures BAB_A in the computational basis (outcome mAm_A); Bob applies CSUB⁡(SB→DB)\operatorname{CSUB}(S_B \to D_B) and measures SBS_B in the Fourier basis {F†∣m⟩}\{F^\dagger|m\rangle\}, i.e. applies FF (defined below) and reads out mBm_B in the computational basis; after exchanging the two dits, Bob applies XdmAX_d^{m_A} to DBD_B and Alice applies ZdmBZ_d^{m_B} to DAD_A. Only Stage 1 requires quantum communication with all candidates, and the two dits of Stage 3 may be announced publicly (Methods). Operational requirement: the dits need integrity but no secrecy [36]; altering them by (δA,δB)(\delta_A,\delta_B) replaces CSUM by (ZdδB⊗XdδA)(Z_d^{\delta_B}\otimes X_d^{\delta_A})CSUM, a definite Weyl error rather than a randomization (Methods).

Trust assumptions and anonymity. The provider is trusted to follow the protocol and knows the recipient, as does Alice. Candidate sites are colluding semi-honest: they may store, measure and pool shares but do not deviate. Anonymity refers to these unselected sites: their registers and the public transcript (mA,mB)(m_A,m_B) are independent of whether, and which, site was selected (Supplementary Section S13).

Encoding and decoding circuits

The construction uses the shift and clock operators Xd∣j⟩=∣j+1⟩X_d\lvert j\rangle=\lvert j+1\rangle and Zd∣j⟩=ωj∣j⟩Z_d\lvert j\rangle=\omega^j\lvert j\rangle with ω=e2πi/d\omega=e^{2\pi i/d} and all labels taken modulo dd; the discrete Fourier transform Fkj=ωjk/dF_{kj}=\omega^{jk}/\sqrt{d}; the sum gate CSUM ∣c,t⟩=∣c,c+t⟩\lvert c,t\rangle=\lvert c,c+t\rangle and its inverse CSUB; and the quadratic phase gate

Pθ=diag⁡(eiθk2)k=0d−1.(1)P_\theta=\operatorname{diag}\left(e^{i\theta k^2}\right)_{k=0}^{d-1}. \tag*{(1)}

With L\mathcal{L} the ladder CSUM(BC→S1)(B_C\to S_1), CSUM(S1→S2)(S_1\to S_2), …\ldots, CSUM(Sn−1→Sn)(S_{n-1}\to S_n) and F=F⊗(n+1)\mathbb{F}=F^{\otimes(n+1)} on (BC,S1,…,Sn)(B_C,S_1,\ldots,S_n), the encoder is

Uenc=UXUZ,UZ=L†(Pθ on Sn)L,UX=F†UZF,(2)U_{\mathrm{enc}}=U_XU_Z,\qquad U_Z=\mathcal{L}^{\dagger}(P_\theta\text{ on }S_n)\mathcal{L},\qquad U_X=\mathbb{F}^{\dagger}U_Z\mathbb{F}, \tag*{(2)}

which costs 4n4n two-qudit gates, two phase gates and two Fourier layers (Figure 2a). For d=2d=2 it reduces to the qubit encoder of Ref. [26]; the Pauli-exponential form used there has no qudit counterpart, which is why we build the encoder as a circuit (Methods). Each SjS_j starts in the Bell state ∣Φ⟩SjKj=d−1/2∑k∣kk⟩\lvert\Phi\rangle_{S_jK_j}=d^{-1/2}\sum_k\lvert kk\rangle. Privacy of the unselected shares holds for every θ\theta, but exact recovery requires θ\theta to be a flat angle (below), the smallest of which is

Figure 2. Explicit encoding and decoding circuits. a, The encoder UencU_{\mathrm{enc}} of Eq. (2): a CSUM ladder accumulates the total register sum onto SnS_n, a single quadratic phase gate Pθ∗P_{\theta^*} is applied, and the ladder is inverted; the whole block is then repeated conjugated by F⊗(n+1)F^{\otimes(n+1)}. Key registers KjK_j are spectators throughout. b, The decoder Udec(B)U_{\mathrm{dec}}^{(B)} of Eq. (5), acting only on (SB,K1,…,Kn)(S_B,K_1,\ldots,K_n): CSUB(KB→SB)(K_B\to S_B) extracts the shift label onto SBS_B; CSUM(SB→Kj)(S_B\to K_j) restores each unselected pair to ∣Φ⟩\lvert\Phi\rangle; CSUM(Kj→KB)(K_j\to K_B) and CSUB(SB→KB)n(S_B\to K_B)^n collect the phase label onto KBK_B; the single-qudit gates Λ\Lambda and Γ\Gamma cancel the two Gauss-sum chirps; and JJ followed by CSUM(KB→SB)(K_B\to S_B) recombines. Total: 3n3n two-qudit and three single-qudit gates, independent of the d2(n+1)d^{2(n+1)} Hilbert-space dimension a generic recovery map would act on.

References

  1. [1]W. K. Wootters and W. H. Zurek, A single quantum cannot be cloned, Nature 299, 802 (1982).
  2. [2]M. A. Nielsen and I. L. Chuang, Programmable quantum gate arrays, Phys. Rev. Lett. 79, 321 (1997).DOI
  3. [3]G. Chiribella, G. M. D’Ariano, and P. Perinotti, Optimal cloning of unitary transformation, Phys. Rev. Lett. 101, 180504 (2008).pubmed.ncbi.nlm.nih.gov/18999806
  4. [4]D. Gotteman and I. L. Chuang, Demonstrating the viability of universal quantum computation using teleportation and single-qubit operations, Nature 402, 390 (1999).
  5. [5]J. Eisert, K. Jacobs, P. Papadopoulos, and M. B. Plenio, Optimal local implementation of nonlocal quantum gates, Phys. Rev. A 62, 052317 (2000).DOI
  6. [6]S. F. Huelga, J. A. Vaccaro, A. Chefles, and M. B. Plenio, Quantum remote control: Teleportation of unitary operations, Phys. Rev. A 63, 042303 (2001).DOI
  7. [7]H. J. Kimble, The quantum internet, Nature 453, 1023 (2008).DOI
  8. [8]S. Wehner, D. Elkouss, and R. Hanson, Quantum internet: a vision for the road ahead, Science 362, eaam9288 (2018).DOI
  9. [9]K. S. Chou, J. Z. Blumoff, C. S. Wang, P. C. Reinhold, C. J. Axline, Y. Y. Gao, L. Frunzio, M. H. Devoret, L. Jiang, and R. J. Schoelkopf, Deterministic teleportation of a quantum gate between two logical qubits, Nature 561, 368 (2018).DOI
  10. [10]Y. Wan, D. Kienzler, S. D. Erickson, K. H. Mayer, T. R. Tan, J. J. Wu, H. M. Vasconcelos, S. Glancy, E. Knill, D. J. Wineland, A. C. Wilson, and D. Leibfried, Quantum gate teleportation between separated qubits in a trapped-ion processor, Science 364, 875 (2019).DOI
  11. [11]D. Main, P. Dmrota, D. P. Nadlinger, E. M. Ainley, A. Agrawal, B. C. Nichol, R. Srinivas, G. Araneda, and D. M. Lucas, Distributed quantum computing across an optical network link, Nature 638, 383 (2025).DOI
  12. [12]A. Shamir, How to share a secret, Commun. ACM 22, 612 (1979).
  13. [13]M. Hillery, V. Bužek, and A. Berthiaume, Quantum secret sharing, Phys. Rev. A 59, 1829 (1999).DOI
  14. [14]R. Cleve, D. Gottesman, and H.-K. Lo, How to share a quantum secret, Phys. Rev. Lett. 83, 648 (1999).DOI
  15. [15]A. Karlsson, M. Koashi, and N. Imoto, Quantum entanglement for secret sharing and secret splitting, Phys. Rev. A 59, 162 (1999).DOI
  16. [16]A. Keet, B. Fortescue, D. Markham, and B. C. Sanders, Quantum secret sharing with qudit graph states, Phys. Rev. A 82, 062315 (2010).DOI
  17. [17]W. Helwig, W. Cui, J. J. Latorre, A. Riera, and H.-K. Lo, Absolute maximal entanglement and quantum secret sharing, Phys. Rev. A 86, 052335 (2012).DOI
  18. [18]B. A. Bell, D. Markham, D. A. Herrera-Martí, A. Marin, W. J. Wadsworth, J. G. Rarity, and M. S. Tame, Experimental demonstration of graph-state quantum secret sharing, Nat. Commun. 5, 5480 (2014).DOI
  19. [19][19*] J. Pinnell, I. Nape, M. de Oliveira, N. TabeBordbar, and A. Forbes, Experimental demonstration of 11- dimensional 10-party quantum secret sharing, Laser Photonics Rev. 14, 2000012 (2020).
  20. [20]Z.-j. Zhang and C.-Y. Cheung, Shared quantum remote control: quantum operation sharing, Journal of Physics B: Atomic, Molecular and Optical Physics 44, 165508 (2011).DOI
  21. [21]A. M. Wang, Combined and controlled remote implementations of partially unknown quantum operations of multiqubits using Greenberger-Horne-Zeilinger states, Phys. Rev. A 75, 062323 (2007).DOI
  22. [22]A. SaiToh, R. Rahimi, and M. Nakahara, Economical (k,m)-threshold controlled quantum teleportation, Phys. Rev. A 79, 062313 (2009).DOI
  23. [23]M. Christandl and S. Wehner, Quantum anonymous transmissions, in Advances in Cryptology – ASIACRYPT 2005, Lecture Notes in Computer Science, Vol. 3788 (Springer, 2005) pp. 217–235.DOI
  24. [24]C. Crépeau, D. Gottesman, and A. Smith, Secure multi-party quantum computation, in Proceedings of the 34th Annual ACM Symposium on Theory of Computing (STOC) (2002) pp. 643–652.
  25. [25]A. Broadbent, J. Fitzsimons, and E. Kashefi, Universal blind quantum computation, in 50th Annual IEEE Symposium on Foundations of Computer Science (FOCS) (2009) pp. 517–526.
  26. [26]K. Yamaguchi and A. Kempf, Encrypted qubits can be cloned, Phys. Rev. Lett. 136, 010801 (2026).DOI
  27. [27]F.-I. Ceară, Cloning encrypted quantum states in arbitrary dimensions, Phys. Rev. A 114, 022450 (2026).
  28. [28]Z. L. Lim and H.-K. Lo, Encrypted cloning, absolute maximal entanglement and quantum secret sharing (2026), arXiv:2605.26866 [quant-ph].arxiv.org/abs/2605.26866
  29. [29]M. Ringbauer, M. Meth, L. Postler, R. Stricker, R. Blatt, P. Schindler, and T. Monz, A universal qudit quantum processor with trapped ions, Nat. Phys. 18, 1053 (2022).DOI
  30. [30]T. Kim, T. Roy, X. You, A. C. Y. Li, H. Lamm, O. Pritchitev, M. Bal, S. Garattini, F. Crisa, D. Bafia, D. Kurkcuoglu, R. Pilipenko, P. Heidler, N. Bornman, D. van Zanten, S. Zorzetti, R. Harnik, A. Murthy, A. Lunin, S. Belomestnykh, S. Zhu, C. Wang, A. Vallieres, Z. Huang, J. Koch, A. Grassellino, S. Chakram, A. Romanenko, and Y. Lu, Ultracoherent superconducting cavity-based multidot platform with error-resilient control (2025), arXiv:2506.03286 [quant-ph].arxiv.org/abs/2506.03286
  31. [31]J. Huang, E. Kasaba, T. J. DiNapoli, T. Roy, and S. Chakram, Universal jaynes-cummings control of an oscillator (2026), arXiv:2605.18658 [quant-ph].
  32. [32]N. J. Cerf, M. Bourennane, A. Karlsson, and N. Gisin, Security of quantum key distribution using d-level systems, Phys. Rev. Lett. 88, 127902 (2002).
  33. [33]D. Gottesman, Fault-tolerant quantum computation with higher-dimensional systems, Chaos, Solitons & Fractals 10, 1749 (1999).
  34. [34]D. M. Appleby, Symmetric informationally complete–positive operator valued measures and the extended Clifford group, J. Math. Phys. 46, 052107 (2005).DOI
  35. [35]H. Barnum, C. Crépeau, D. Gottesman, A. Smith, and A. Tapp, Authentication of quantum messages, in Proceedings 43rd Annual IEEE Symposium on Foundations of Computer Science (IEEE, 2002) pp. 449–458.
  36. [36]M. N. Wegman and J. L. Carter, New hash functions and their use in authentication and set equality, Journal of Computer and System Sciences 22, 265 (1981).DOI
  37. [37]G. Gianini, O. Hasan, C. Mio, S. Cimato, and E. Damiani, Encrypted clones can leak: Classification of informative subsets in quantum encrypted cloning (2026), arXiv:2604.10155 [quant-ph].
  38. [38]G. Gianini, S. Cimato, J. Lin, O. Hasan, and E. Damiani, Full characterization of informative subsets in quantum encrypted cloning (2026), arXiv:2605.27421 [quant-ph].arxiv.org/abs/2605.27421
  39. [39]C.-M. Bai, X.-L. Zhou, and Y. Luo, Classification of informative subsets in quantum encrypted cloning on qudits (2026), arXiv:2605.11642 [quant-ph].arxiv.org/abs/2605.11642
  40. [40]T. Ogawa, A. Sasaki, M. Iwamoto, and H. Yamamoto, Quantum secret sharing schemes and reversibility of quantum operations, Phys. Rev. A 72, 032318 (2005).DOI
  41. [41]A. Ambainis, M. Mosca, A. Tapp, and D. W. Wolf, Private quantum channels, in Proceedings 41st Annual Symposium on Foundations of Computer Science (IEEE, 2000) pp. 547–553.
  42. [42]A. Broadbent, G. Gutoski, and D. Stebila, Quantum one-time programs, in Advances in Cryptology – CRYPTO 2013 (Springer Berlin Heidelberg, Berlin, Heidelberg, 2013) pp. 344–360.DOI
  43. [43]S. Aaronson, Quantum copy-protection and quantum money, in Proceedings of the 24th Annual IEEE Conference on Computational Complexity (2009) pp. 229–242.DOI
  44. [44]A. Broadbent and S. Lord, Uncloneable quantum encryption via oracles, in 15th Conference on the Theory of Quantum Computation, Communication and Cryptography (TQC 2020), LIPIcs, Vol. 158 (2020) pp. 4:1–4:22.DOI
  45. [45]B. Reznik, Y. Aharonov, and B. Groisman, Remote operations and interactions for systems of arbitrary-dimensional Hilbert space, Phys. Rev. A 65, 023312 (2002).DOI
  46. [46]N. Bornman, T. Roy, J. A. Job, N. Anand, G. N. Perdue, S. Zorzetti, and M. S. Alam, Benchmarking the performance of a high-q cavity qudit using random unitaries, Quantum Science and Technology 10, 025062 (2025).DOI
  47. [47]T. Roy, Systems and methods for deferred-recipient teleportation of a two-qudit gate, U.S. Provisional Patent Application No. 64/182,988 (2026).
  48. [48]V. Gheorghiu, Standard form of qudit stabilizer groups, Phys. Lett. A 378, 505 (2014).DOI

Paper details

vlnc:W_v9vwkmdy
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 2

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents