Advances Quantum Belief Propagation with Messages for Symmetric Q-Ary Channels

Researchers have extended the capabilities of belief propagation with quantum messages (BPQM) beyond binary systems, offering a potentially simpler route to reliable communication over noisy quantum channels. Avijit Mandal and Henry D. Pfister, both from Duke University, alongside colleagues, demonstrate BPQM’s functionality for symmetric q-ary pure-state channels with circulant output Gram matrices. This work is significant because it establishes closed-form recursions for tracking message combining, independent of the specific quantum states used, and delivers analytic bounds on channel fidelity. Consequently, the team provides a density evolution framework enabling estimation of decoding thresholds for low-density parity-check codes and the construction of polar codes on these channels, representing a substantial step towards practical quantum communication protocols.

Prior BPQM constructions and density-evolution (DE) analyses were largely confined to binary alphabets, limiting their applicability to more complex communication scenarios. This research establishes a framework for efficiently tracking bit-node and check-node combining through closed-form recursions based on the Gram-matrix eigenvalues, effectively decoupling the analysis from specific physical implementations of the output states. These recursions yield explicit BPQM unitaries and provide analytic bounds on channel fidelities, directly relating them to the fidelities of the input channels.

Simplifying BPQM Using Gram Matrix Eigenvalues

The team achieved a breakthrough by recognizing that, for symmetric q-ary PSCs with circulant Gram matrices, the behaviour of check-node and bit-node combining can be entirely characterized by the eigenvalues of the Gram matrix, or the ‘eigen list’. This simplification allows for the derivation of explicit update rules for these combining operations, leading to the creation of corresponding BPQM unitaries. Furthermore, the methodology extends to heralded mixtures of symmetric PSCs through the use of controlled unitaries, broadening the scope of its application. By leveraging these eigen list recursions, the researchers developed practical DE procedures applicable to both polar and LDPC ensembles operating on symmetric q-ary PSCs, enabling the design of polar codes for specific block error rates and the estimation of LDPC decoding thresholds.
This study unveils a novel approach to characterizing key information-theoretic quantities, including symmetric Holevo information, channel fidelity, and pretty-good measurement error rate, all using the Gram matrix for symmetric q-ary PSCs. Crucially, the work provides upper bounds on the fidelity of combined channels in terms of input-channel fidelities, directly derived from the BPQM eigen list update rules. The researchers define the quantum states {|ψu⟩} using the eigen list λ and Fourier vectors, establishing a clear connection between the mathematical framework and the physical realization of the channel. Lemma 5 proves that the Fourier vectors are eigenvectors of the Gram matrix, and Lemma 6 demonstrates that the sum of the eigenvalues equals q, forming a probability distribution.

Experiments show that all symmetric q-ary pure-state channels are uniquely determined by their eigen list, as detailed in Lemma 7, simplifying the analysis and allowing for a more concise representation of the channel characteristics. The derived BPQM update rules are not limited to prime q and naturally generalize to channels with symmetry corresponding to any finite abelian group by substituting the Discrete Fourier Transform with the appropriate group character table. This generalization significantly expands the potential applications of the research. The work opens avenues for designing efficient quantum communication systems and constructing robust codes for increasingly complex quantum channels, paving the way for more reliable and high-performance quantum data transmission.

BPQM Analysis via Gram Matrix Eigenvalues

Developing Closed-Form Recursive Tracking Framework

Scientists developed a novel framework for belief propagation with quantum messages (BPQM) applicable to symmetric q-ary pure-state channels (PSCs) with circulant output Gram matrices. The research pioneered a method to efficiently track bit-node and check-node combining using closed-form recursions based on the Gram-matrix eigenvalues, effectively decoupling analysis from specific physical implementations of the output states. This approach enables the derivation of explicit BPQM unitaries and analytic bounds on channel fidelities, directly relating them to the input-channel fidelities. Experiments employed a density evolution (DE) framework, allowing researchers to estimate BPQM decoding thresholds for low-density parity-check (LDPC) codes and to construct polar codes on these channels.

The study harnessed the mathematical properties of circulant Gram matrices to simplify the complex calculations inherent in BPQM. Researchers formulated recursions that operate directly on the eigenvalues of the Gram matrix, bypassing the need for detailed state tracking and significantly reducing computational complexity. This technique reveals a direct link between the input channel characteristics and the performance of the BPQM decoder, providing valuable insights into the limits of quantum communication. The team engineered a system where the combining of messages at each node in the belief propagation algorithm could be expressed in a concise, eigenvalue-based form.

Applying Method to Quantum Error Correction Codes

Scientists then validated this approach by applying it to the construction of both LDPC and polar codes, demonstrating its versatility across different coding schemes. The method achieves an analytic understanding of decoding thresholds, crucial for optimizing code performance and ensuring reliable communication. Experiments involved calculating the fidelities of combined channels, providing quantitative measures of the information transfer efficiency. This precise measurement approach allowed for a rigorous comparison of different BPQM configurations and channel parameters. Furthermore, the work extends beyond simple error correction, offering a pathway to construct capacity-achieving codes for symmetric q-ary PSCs.

The research team demonstrated that the developed recursions yield explicit BPQM unitaries, essential components for implementing the decoding algorithm. This innovative methodology provides a powerful tool for analysing and designing quantum communication systems, potentially leading to more efficient and robust data transmission protocols. The study’s findings are particularly relevant for applications requiring high fidelity and low complexity, such as optical communication and quantum cryptography.

BPQM decoding thresholds for quantum channels

Scientists have developed a new framework for belief propagation with messages (BPQM) applicable to symmetric q-ary pure-state channels (PSCs) with circulant Gram matrices, offering a low-complexity alternative to collective measurements over classical-quantum channels. The research demonstrates that bit-node and check-node combining can be efficiently tracked using closed-form recursions on the Gram-matrix eigenvalues, independent of the physical realisation of the output states. These recursions yield explicit BPQM unitaries and analytic bounds on channel fidelities, directly relating them to the input-channel fidelities. Experiments revealed that the derived BPQM framework allows for the estimation of decoding thresholds for Low-Density Parity-Check (LDPC) codes and the construction of polar codes on these channels.

The team restricted their analysis to prime q for simplicity, but the BPQM update rules rely solely on the diagonalization of the Gram matrix via the Discrete Fourier Transform, generalising naturally to channels with symmetry corresponding to any finite abelian group. Data shows that two CQ channels are isometrically equivalent if a unitary transformation exists, mapping one channel to the other for all inputs. Researchers defined a pure state channel (PSC) where the output state is a rank-1 matrix for all inputs, and specifically focused on symmetric q-ary PSCs characterised by a circulant Gram matrix. Measurements confirm that the eigenvalues of the Gram matrix, denoted as λ = [λ0, . . . , λq−1], are crucial for defining the channel, and that the sum of these eigenvalues equals q.

The team established that all symmetric q-ary PSCs are uniquely determined by their eigenlist, allowing for channel description using quantum states defined by a summation over Fourier vectors weighted by the eigenvalues. Results demonstrate a direct relationship between the symmetric Holevo information I(W) and channel fidelity F(W) with the Gram matrix G, expressed as I(W) = H(μ) and F(W) = 1/q , (1/(q-1)) Σ|gu|, where μ is the normalized eigenlist and H(μ) is the Shannon entropy. Tests prove that the probability of error, Perr(W), for optimally distinguishing the output states of a symmetric q-ary PSC using a pretty good measurement (PGM) is given by Perr(W) = 1 − (1/q) Σ p λu 2. This breakthrough delivers a powerful analytical tool for characterising and optimising quantum communication systems based on symmetric q-ary PSCs, paving the way for advanced coding schemes and improved channel performance.

BPQM analysis via Gram matrix recursions

Scientists have developed a new approach to belief propagation with quantum messages (BPQM) applicable to symmetric q-ary pure-state channels (PSCs), extending previous work limited to binary alphabets. This research introduces a method for tracking bit-node and check-node combining through closed-form recursions based on the eigenvalues of the channel’s Gram matrix, effectively bypassing the need for large-scale quantum state simulations. The resulting recursions yield explicit BPQM unitaries and analytic bounds on channel fidelities, allowing for density evolution analysis independent of the physical realisation of the output states. This advancement enables the estimation of BPQM decoding thresholds for low-density parity-check (LDPC) codes and the construction of polar codes on these channels.

Demonstrations include the design of q-ary polar codes for a specific block error rate and the estimation of BPQM thresholds for regular LDPC ensembles, with results showing a BPQM threshold of 2.4 compared to a Holevo information bound of 2.52 for a rate 1/2 code when q equals 3. The authors acknowledge a limitation in the current scope, focusing on symmetric q-ary PSCs with circulant output Gram matrices. Future work could extend these tools to more general finite-abelian symmetries through character-based diagonalisation, potentially broadening the applicability of this BPQM framework.