TCG Protocol for High-Fidelity Quantum Control

• TCG is a quantum control protocol that leverages auxiliary energy levels to mediate conditional operations with reduced circuit depth.
• It employs composite pulse sequences and phase-modulated gates in both transmon and Rydberg systems to achieve high gate fidelities.
• The protocol simplifies scalable quantum circuit design by offering robust error suppression and time-optimal pulse strategies.

The Transition-Composite-Gate (TCG) protocol is a quantum control methodology for implementing scalable, high-fidelity conditional operations in both superconducting and neutral-atom platforms. By exploiting engineered transition pathways that traverse auxiliary energy levels beyond the usual computational subspace, TCG achieves conditional logic with reduced circuit depth, robust error characteristics, and practical pulse sequences. The protocol encompasses both composite pulse strategies in circuit QED and phase-modulated Rydberg gates, offering a unified framework for efficient digital quantum logic under realistic noise and control constraints (Zhang et al., 2024, Cole et al., 28 Dec 2025).

1. Theoretical Foundations

TCG protocols are grounded in systems with access to auxiliary states outside the computational basis, enabling temporally engineered population transfer through well-controlled transitions. The theory is formalized on two leading architectures:

• Superconducting Transmons (Circuit QED): The Hamiltonian includes a chain or pair of weakly anharmonic transmon qutrits, represented as

$$H_0 = \sum_{i=1,2}\left[\hbar\,\omega_{01}^{(i)}\,a_i^\dagger a_i - \frac{\hbar\alpha_i}{2}\,a_i^\dagger a_i^\dagger a_i a_i\right] + \hbar\,\omega_c\,c^\dagger c$$

with interaction

$$H_\mathrm{int} = \sum_{i=1,2}\hbar\,g_{i}(a_i c^\dagger + a_i^\dagger c)$$

After diagonalization, the computational subspace is $S_c = \{|0\rangle,|1\rangle\}^{\otimes2}$, and the non-computational subspace involves higher levels such as $|2\rangle$.

• Neutral Atom Rydberg Arrays: Atoms are modeled in an extended basis ($|0\rangle$, $|1\rangle$, $|r\rangle$), with coupled Hamiltonians for control and target atoms:

$H(t) = H_c(t) + H_t(t) + V|rr\rangle\langle rr|$

$H_c(t)$ and $H_t(t)$ describe resonant or detuned laser couplings and the Rydberg blockade shift $V$.

Both architectures utilize "transition" operations ($T$) to mediate population between computational and auxiliary states, interleaved with "internal" operations ($I$) fully contained within either subspace (Zhang et al., 2024, Cole et al., 28 Dec 2025).

2. TCG Circuit Construction and Gate Decompositions

Transition Pathways and Protocol Structure

The TCG protocol connects computational input and output states via a chain of $N$ transitions $T_i$, interleaved with $N-1$ internal operations $I_i$. The symmetric (unitary) TCG sequence is:

$$U_\textrm{TCG} = \prod_{i=1}^{N-1}(T_i I_i)\,T_N\,\prod_{i=1}^{N-1}(I_{N-i} T_{N-i})$$

For state preparation or unidirectional tasks, the short-path (SPTCG) variant truncates the sequence:

$U_\textrm{SPTCG} = \prod_{i=1}^{N-1}(T_i I_i)\,T_N$

Exemplary Case: Controlled-Unitary (CU) Gate

In superconducting devices, a two-qutrit TCG path yields the full controlled-unitary (CU) family:

• $T_1 = \sqrt{\rm CZ}_{(11\leftrightarrow 20)}$ (a diabatic resonance between $|11\rangle$ and $|20\rangle$),
• $I_1 = I_0\otimes X^{12}_{\theta,\phi}$ (single-qutrit $X^{12}$ rotation on the target qubit),
• $T_2 = \sqrt{\rm CZ}_{(11\leftrightarrow 20)}$.

The net unitary on $S_c$ is

$$U_\textrm{CU} = \sqrt{\rm CZ}\;[I\otimes X^{12}_{\theta,\phi}]\;\sqrt{\rm CZ}$$

This decomposition provides access to controlled-$R_x(\theta)$, controlled-$R_y(\theta)$, controlled-Hadamard, or CNOT gates in a three-pulse sequence by varying $(\theta, \phi)$ (Zhang et al., 2024).

Pulse Sequence Realization

For transmon-based TCG:

• The $\sqrt{\rm CZ}$ is implemented by tuning a coupler frequency to execute a $30\,\mathrm{ns}$ flat-top Gaussian flux pulse,
• The $X^{12}$ operation is a $30\,\mathrm{ns}$ Gaussian DRAG pulse on the $|1\rangle \leftrightarrow |2\rangle$ transition, with DRAG optimized for leakage suppression (Zhang et al., 2024).

For neutral atom TCG:

• The protocol follows a three-pulse “π–2π–π” sequence: a resonant $\pi$-pulse on control, a detuned $2\pi$-pulse on target (with $\Delta=V/2$), and a second $\pi$-pulse on control. These pulses are parameterized for time-optimal return, population restoration, and desired controlled phase by adjusting $\Omega$, $\Delta$, and durations (Cole et al., 28 Dec 2025).

3. Performance Metrics and Error Analysis

TCG protocols achieve substantial gate and circuit-level advantages:

Metric	Superconducting TCG	Rydberg TCG
Two-qubit gate fidelity	95.2%–99.0% (CNOT 97.5%)	$1.7$–$2.4\times$ lower limit set by $1/(V\tau)$
Three-qubit GHZ fidelity	96.8% (SPTCG)	n/a
Circuit depth reduction	40–44% (GHZ, W states)	n/a
Limiting errors	Decoherence, DRAG leakage	Rydberg decay

Gate errors for transmons are dominated by sequence decoherence ($T_1, T_2^* \sim 10\;\mu$s vs. sequence $\sim90$ ns) and DRAG pulse imperfections, particularly for larger $X^{12}$ rotations. Calibration protocols with closed-loop quantum process tomography (QPT) are used to optimize process fidelity $$F_\mathrm{proc} = \mathrm{Tr}[\chi_\mathrm{exp}\chi_\mathrm{ideal}]$$ (Zhang et al., 2024).

For Rydberg TCG, the irreducible infidelity is due to finite Rydberg lifetime. An analytic evaluation yields

$$\epsilon_{\rm TCG} = \left(\frac{11}{8}+\frac{1}{\sqrt{3}}\right)\frac{\pi}{V\tau} \approx 2.39\,\epsilon_{\mathrm{DDP}}$$

where $\epsilon_{\mathrm{DDP}}$ is the fundamental limit for blockade-based entangling gates. In fully asymmetric configurations ($p \to \infty$), this limit decreases to $1.68\,\epsilon_{\mathrm{DDP}}$. The protocol is robust for blockade strengths $V$ down to $\Omega$ and retains zero coherent rotation error (Cole et al., 28 Dec 2025).

4. Applications to Quantum Circuits and Algorithms

TCG directly enables resource-efficient quantum circuit implementation for complex conditional logic:

Entangled State Preparation

For $m$-qubit GHZ and W state circuits, use of TCG-based CU gates (and SPTCG for state prep) reduces depth by $40\%$ (GHZ) and $44\%$ (W at $m=3$) relative to CZ-based decompositions. Three-qubit GHZ and W preparation achieves fidelities $>$96% (Zhang et al., 2024). Example circuits are explicitly constructed using three-pulse or short-path TCGs.

Quantum Comparators and Multifunctional Gates

TCG provides a scalable approach for constructing comparators (multi-control conditional logic) with linearly scaling resource count. Each $n$-control unitary (e.g., CCU, CCCU) is realized via a symmetric TCG path of $2N-1$ operations, leveraging auxiliary-level “bridge” transitions (Zhang et al., 2024). This approach is particularly advantageous for quantum algorithms with intensive conditional branching.

Generalization to Arbitrary Controlled Phases

The Rydberg-TCG scheme generalizes to arbitrary controlled-phase gates by selecting pulse parameters $(\Omega, \Delta, t_2)$ such that both the “blocked” and “unblocked” populations undergo integer multiples of $2\pi$ rotation, allowing the controlled phase $\theta$ to span $[0, 2\pi]$ continuously (Cole et al., 28 Dec 2025).

5. Time-Optimality and Robust Control Strategies

Analytic and numerical analyses establish that the canonical three-pulse TCG is time-optimal among phase-modulated, constant-amplitude solutions for a fixed $\Omega$ and tunable $V$, with a minimal total gate time in the $p\to\infty$ limit converging to $2\pi/V$ (Cole et al., 28 Dec 2025).

For robust operation under miscalibrations ($\delta\Omega/\Omega_0$, $\delta V/V_0$), phase modulation during the central pulse is optimized using a GRAPE search. Robust pulses are $1.5$–$2\times$ longer than non-robust gates but suppress worst-case error by factors of $3$–$7$ under static drift. When spontaneous decay is included, the robust protocol yields an optimal tradeoff between gate time and fidelity (Cole et al., 28 Dec 2025).

6. Practical Implementation and Calibration

In superconducting transmon arrays, three adjacent qubits on a $72$-qubit device (Q1–Q3) serve as experimental testbed for TCG; detailed frequencies, anharmonicities, and coherence times (e.g., $T_1$ and $T_2^*$ both 10–11 $\mu$s) are reported (Zhang et al., 2024). The effective two-qubit coupling $g_{qq}$ is tuned to $10$–$15$ MHz for the $\sqrt{\rm CZ}$. Pulse calibration includes fine-tuning Gaussian DRAG parameters for leakage suppression and maximizing $P_{11}\leftrightarrow P_{20}$ swap probabilities during $\sqrt{\rm CZ}$, with QPT iteratively verifying fidelity.

In neutral atom experiments, square $π$- and $2π$-pulses are applied with parameters chosen according to the analytic detuning and duration formulas, and phase correction Z-rotations are applied to correct final logic phases (Cole et al., 28 Dec 2025).

---

The TCG framework, as established in (Zhang et al., 2024) and (Cole et al., 28 Dec 2025), constitutes a generalizable and experimentally validated protocol for scalable, conditional quantum operations. Its architecture-agnostic utilization of auxiliary-level “fly-by” transitions, composite-pulse design, and controllable constructive interference achieves circuit simplification, high-fidelity conditional logic, and robustness to hardware non-idealities—key features for NISQ and near-future quantum processors.