Controlled Secure Direct Quantum Communication

• Controlled secure direct quantum communication is a family of quantum cryptographic protocols that transmit secret messages directly while requiring controller cooperation for message recovery.
• These protocols utilize diverse quantum resources such as GHZ-like, Bell, cluster, and continuous-variable states to enforce control through entanglement, particle permutation, and measurement verification.
• They achieve secure, efficient, and scalable communication by integrating entanglement swapping, threshold control, and rigorous error testing against various quantum attacks.

Searching arXiv for recent and foundational papers on controlled secure direct quantum communication and closely related QSDC variants. Controlled secure direct quantum communication denotes a family of quantum cryptographic protocols in which secret messages are transmitted directly through a quantum channel, while successful recovery of the message requires the permission or cooperation of a controller. In the cited literature, this family includes controlled quantum secure direct communication (CQSDC), controlled deterministic secure quantum communication (CDSQC), controlled bidirectional quantum secure direct communication (CBQSDC), and controlled quantum dialogue (CQD). The controller may authorize a one-way transmission, simultaneous two-way exchange, or threshold access in which only a prescribed coalition of controllers enables recovery. Across the literature, control is implemented by withholding classical information about entangled resources, particle order, measurement outcomes, or decoding parameters, and the central design objective is to combine direct communication with security, efficiency, and enforceable authorization (Sarvaghad-Moghaddam, 2019, Patwardhan et al., 2015, Pathak, 2014).

1. Conceptual scope and protocol roles

The standard actors are a sender, a receiver, and a controller. In the unidirectional form, Alice sends a secret bit-string directly to Bob, and Bob can obtain the complete bit-string only with the permission of Charlie; in the bidirectional form, Alice and Bob exchange secret messages simultaneously under Charlie’s authorization (Patwardhan et al., 2015, Sarvaghad-Moghaddam, 2019). This distinguishes controlled direct communication from quantum key distribution, because the communicated object is the message itself rather than a key later used for ciphertext transmission.

The literature also uses closely related labels. CDSQC emphasizes deterministic recovery after the controller’s cooperation, while CQD emphasizes simultaneous or two-way exchange. The continuous-variable controlled-two-way scheme is explicitly described as very general in nature, since it can be reduced to controlled deterministic secure communication, quantum dialogue, and quantum key distribution depending on user actions (S et al., 2019). Threshold control extends the model further: in the Brown-state proposal, any two of three controllers are required to retrieve the complete message (Patwardhan et al., 2015).

A second axis of variation concerns the quantum capability of the communicating parties. In semi-quantum controlled deterministic secure communication, Charlie is fully quantum-capable, Bob is quantum-capable, and Alice is classical in the specific sense that she can measure and prepare qubits only in the computational basis or reflect them without disturbance (Shukla et al., 2017). This widens the protocol class without altering the basic control principle: decoding remains impossible until the controller releases indispensable information.

2. Quantum resources and physical mechanisms

Controlled direct communication has been realized, at the protocol level, with GHZ-like states, Bell states, Cluster states, Brown states, single-mode squeezed coherent states, discrete-time quantum walk states, and four-dimensional single-particle states. The resource choice determines both the control mechanism and the efficiency claims made for the protocol family.

Approach	Quantum resource	Distinctive feature
GHZ-like CQSDC	Three-qubit GHZ-like state	No qubits carrying secret messages are transmitted
Bell/EPR with PoP	Bell states	Control via permutation of particles
Cluster/Brown CQSDC	4-qubit Cluster; 5-qubit Brown	Single-controller and $2$-of-$3$ threshold control
CBQSDC by entanglement swapping	Bell states chosen from four Bell states	Simultaneous bidirectional direct communication
CV controlled two-way communication	Single-mode squeezed coherent states	No requirement for two-mode squeezed states
Higher-dimensional CQSDC	Four-dimensional single-particle states	Three-party decoding with collaborative unitary sequence

GHZ-like-state CQSDC uses the state

$$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$

and motivates this channel by the property that if a qubit is lost in the GHZ-like state the other two qubits are still entangled (Hassanpour et al., 2014). Bell-state approaches reduce the entanglement requirement to bipartite resources. In particular, the PoP-based CDSQC construction argues that Bell state is sufficient, whereas earlier protocols required at least tripartite entanglement (Pathak, 2014).

Entanglement swapping is one of the principal physical mechanisms in controlled direct communication. In the EPR-based bidirectional protocol, two initial Bell pairs are transformed by Bell-basis measurements so that distant qubits become entangled without directly interacting:

$$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$

This identity is the basis for protocols in which local operations and local Bell measurements replace transmission of message-carrying qubits after the initial distribution stage (Sarvaghad-Moghaddam, 2019).

Continuous-variable controlled communication replaces entangled two-mode resources with single-mode squeezed coherent states of the form $S(s)D(\gamma)|0\rangle$, while discrete-time-quantum-walk protocols replace qubit or Bell-pair carriers with states in a larger Hilbert space $H_p\otimes H_c$. The four-dimensional qudit proposal moves further by using a set of 16 maximally symmetric states and a collaborative unitary-sequence decoding paradigm (S et al., 2019, S et al., 2020, Lu et al., 17 Dec 2025).

3. Canonical protocol structure and enforcement of control

A representative controlled bidirectional protocol proceeds in five stages. First, the controller prepares $2N+c$ Bell states, randomly chosen from the four Bell states, distributes one qubit of each pair to Alice and the other to Bob, and receives acknowledgments of receipt. Second, Alice randomly selects $c$ checking particles, measures them in randomly chosen $X$ or $Z$ bases, broadcasts positions, bases, and results, and Charlie discloses the initial Bell states for the testing pairs. If correlations match the Bell-state predictions within thresholds, the protocol continues; otherwise it aborts. Third, Alice and Bob group the remaining $3$0 qubits in pairs and encode their two-bit messages by local Pauli operations. Fourth, both perform Bell measurements on their local pairs and publicly announce the outcomes. Fifth, Charlie either withholds or announces the classical information about the initial Bell states. Only with this disclosure can each user combine the initial state, local encoding, and the other user’s Bell result to infer the counterpart’s message (Sarvaghad-Moghaddam, 2019).

This structure clarifies how control is enforced. In the Bell-state bidirectional protocol, the controller’s announcement is not a mere administrative acknowledgment; it is the missing variable required to invert the entanglement-swapping correlations. The paper states that if Charlie refuses, communication cannot proceed, and mutual message extraction is impossible. The same logic appears in the GHZ-like three-party protocol, where Bob needs one classical bit from Alice and one from Charlie before applying the unitary operation and Bell-basis measurement that reveal Alice’s two-bit secret (Hassanpour et al., 2014).

Other families implement the same principle with different technical means. In the Cluster-state protocol, Alice, Bob, and Charlie all measure in the computational basis, and Bob reconstructs Alice’s two-bit message only after Charlie supplies his one-bit outcome; in the Brown-state protocol, any coalition of two out of three controllers is sufficient, while one controller alone is not (Patwardhan et al., 2015). In the semi-quantum protocol, Charlie prepares $3$1 copies of a GHZ-like channel, Alice encodes by measuring in the computational basis and preparing $3$2, and Bob remains unable to decode until Charlie announces his measurement outcome in the $3$3 basis (Shukla et al., 2017).

The 4D-qudit protocol makes the control mechanism especially explicit. Charlie’s authorization releases both the initial-state description $3$4 and the license key $3$5, enabling Bob to apply

$3$6

up to a global phase. Without this authorization, Bob cannot reconstruct the decoding sequence (Lu et al., 17 Dec 2025).

4. Security model, attack classes, and common misconceptions

A recurring security claim in this literature is that, in entanglement-swapping protocols, no qubits carrying secret messages are transmitted after the security check. This is stated explicitly for the GHZ-like CQSDC protocol and for the EPR-based CBQSDC protocol, and it underlies the claim that a single security check suffices in those constructions (Hassanpour et al., 2014, Sarvaghad-Moghaddam, 2019). The practical significance is narrow but important: the main exposure surface is the initial distribution of entangled states, not a later transmission of message-bearing qubits.

The attack models considered are broad. The entanglement-swapping bidirectional protocol analyzes intercept-and-resend, CNOT, and entangle-and-measure attacks, and argues that Bell-correlation testing or disturbance introduced during channel verification reveals these attacks (Sarvaghad-Moghaddam, 2019). The GHZ-like protocol gives an explicit undetected-escape probability for intercept-resend under its checking procedure and states that CNOT and entanglement-and-measure attacks also introduce detectable errors (Hassanpour et al., 2014). The discrete-time-quantum-walk protocols add denial-of-service and man-in-the-middle attacks, and claim that the CQD protocol is unconditionally secure against an untrusted service provider (S et al., 2020).

Continuous-variable controlled communication formulates security in information-theoretic terms. Against a Gaussian Quantum Cloning Machine attack, the criterion is

$3$7

For the QSDC protocol, the paper attributes unconditional security against GQCM to the randomization of squeezing $3$8 and displacements $3$9; for CQD, security is stated to hold against GQCM as long as transmittance $$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$0 and proper choice of squeezing is made (S et al., 2019). The 4D-qudit proposal adds decoy-photon authentication, with an undetected-escape probability $$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$1 for $$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$2 decoy photons under the specified attack model (Lu et al., 17 Dec 2025).

Two misconceptions are explicitly separated in the literature. First, controller access is not supposed to imply controller knowledge of the message. Several proposals present “blind control,” meaning that Charlie can authorize or deny communication without learning the secret bits because the private encodings remain local to Alice and Bob (Sarvaghad-Moghaddam, 2019, S et al., 2019). Second, an untrusted third party in a QSDC protocol is not necessarily a controller. The measurement-device-independent QSDC protocol delegates all measurements to an untrusted Charlie in the middle, but the paper states that it does not incorporate an explicit controller and that Charlie’s role is measurement rather than authorization (Zhou et al., 2018).

5. Efficiency, bidirectionality, and network generalization

Efficiency is a major comparative axis in controlled direct communication. Several papers use

$$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$3

or equivalent notation, where $$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$4 is the number of secret bits, $$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$5 the number of qubits used, and $$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$6 the number of classical bits exchanged for decoding (Hassanpour et al., 2014, Patwardhan et al., 2015). The Bell-state bidirectional protocol also gives $$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$7 and claims maximum efficiency for bidirectional controlled direct communication (Sarvaghad-Moghaddam, 2019).

The numerical comparisons reported in the literature vary substantially by resource model. The GHZ-like three-party protocol reports $$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$8, compared with $$|P_{abc}^{+}\rangle=\frac{1}{2}(|100\rangle+|010\rangle+|001\rangle+|111\rangle)_{abc},$$9 for the Dong et al. and Kao et al. protocols in its comparison table (Hassanpour et al., 2014). The Cluster-state CQSDC reports $$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$0 and $$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$1, while the Brown-state $$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$2-$$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$3-CQSDC reports $$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$4 and $$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$5 (Patwardhan et al., 2015). The PoP-based Bell-state CDSQC reports, with decoys, $$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$6 and $$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$7, and uses these values to argue that Bell-state control can outperform GHZ-like alternatives (Pathak, 2014). The four-dimensional-qudit protocol defines qudit efficiency as $$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$8 and reports $$|\phi^+\rangle_{12}|\phi^+\rangle_{34} = \frac{1}{2} \left( |\phi^+\rangle_{14}|\phi^+\rangle_{23} + |\phi^-\rangle_{14}|\phi^-\rangle_{23} + |\psi^+\rangle_{14}|\psi^+\rangle_{23} + |\psi^-\rangle_{14}|\psi^-\rangle_{23} \right).$$9, derived from $S(s)D(\gamma)|0\rangle$0 information bits sent with total quantum resources $S(s)D(\gamma)|0\rangle$1 (Lu et al., 17 Dec 2025).

Bidirectionality is another major differentiator. Earlier direct-communication schemes often required two protocol runs to emulate dialogue, whereas the entanglement-swapping CBQSDC protocol states that both users can transmit secret messages directly and simultaneously under the permission of the controller (Sarvaghad-Moghaddam, 2019). In the same paper, each EPR pair is said to carry two bits, one each direction, and a generalization to a CBQSDC-based network is presented using GHZ states. Continuous-variable CQD is likewise presented as controlled-two-way secure communication, and the paper states that the scheme can be reduced to controlled deterministic SDQC, quantum dialogue, and quantum key distribution (S et al., 2019).

Scalability claims are tied to the chosen resource. GHZ-state generalization supports $S(s)D(\gamma)|0\rangle$2-party networks in the Bell-state bidirectional line of work (Sarvaghad-Moghaddam, 2019). Threshold access appears in the Brown-state construction (Patwardhan et al., 2015). The semi-quantum proposals are explicitly motivated by settings such as e-commerce in which not all parties can be assumed to possess full quantum capability (Shukla et al., 2017). The 4D-qudit paper presents its protocol as a scalable architecture for future quantum networks requiring supervision or centralized gatekeeping (Lu et al., 17 Dec 2025).

6. Experimental status and adjacent research directions

The experimental record in the cited set is concentrated on QSDC rather than on an explicit controller layer, but these results define the practical substrate on which controlled variants would have to rely. The first experimental demonstration of QSDC with single photons, based on the DL04 protocol and frequency coding, reported a block transmission of 80 single photons, effectively 16 different values, equivalent to 4 bits of direct transmission in one block, with a demonstrated communication rate of 4 kbps (Hu et al., 2015). The emphasis on block transmission is directly relevant because controlled direct communication also depends on separating eavesdropping detection from message disclosure.

Quantum memory is another enabling component. The proof-of-principle demonstration of QSDC with atomic quantum memory reported entanglement-decoding fidelities of $S(s)D(\gamma)|0\rangle$3, $S(s)D(\gamma)|0\rangle$4, $S(s)D(\gamma)|0\rangle$5, and $S(s)D(\gamma)|0\rangle$6 for the four encoded Bell states, with an average fidelity of approximately $S(s)D(\gamma)|0\rangle$7, a bit rate of approximately $S(s)D(\gamma)|0\rangle$8, and storage times of 50 ns in MOT A and 120 ns in MOT B (Zhang et al., 2016). This suggests that controlled protocols requiring synchronization, buffering, or staged authorization would benefit directly from advances in memory-assisted QSDC.

Security hardening has also progressed in adjacent QSDC lines. Measurement-device-independent QSDC eliminates security loopholes associated with the measurement device and doubles the communication distance compared to protocols without the technique, but it does not itself realize controlled communication (Zhou et al., 2018). Continuous-variable direct secure communication removes the requirement for two-mode squeezed states in the cited schemes (S et al., 2019). More recent work on quantum semantic communication integrates QSDC with semantic encoding, experimentally transmits 3D point clouds over 50 km of standard single-mode fiber, maintains QBER at $S(s)D(\gamma)|0\rangle$9, reports a secure transmission rate of 37.36 kbps for raw quantum bits, and achieves a 46.30-fold efficiency gain over direct transmission (Wang et al., 11 Nov 2025). Although this is not a controlled protocol, it indicates that future controlled architectures may inherit the broader QSDC trajectory toward higher data efficiency, more sophisticated physical-layer security, and compatibility with quantum-network deployment.