Quantum LDPC Decoding via Belief Propagation Polynomial Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional silicon-based LDPC decoders face limitations in communication throughput and computation due to serial message passing iterations, leading to bit errors from both noise in channels and computational processes.
Innovation Solution
Implementing a quantum annealer that reduces the belief propagation algorithm of LDPC codes into a quadratic polynomial and embeds it onto quantum bits (qubits) for decoding, leveraging the parallel structure of LDPC codes and the capabilities of quantum annealing to improve decoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional silicon-based LDPC decoders use serial message passing iterations, then decoding can be performed, but communication throughput is limited and bit error rates increase due to computational errors
Solution Approach 1:
The patent replaces the conventional silicon-based serial message passing algorithm with a quantum annealing system. The belief propagation algorithm is reduced to a quadratic polynomial and embedded onto a quantum annealer, substituting the mechanical/electrical serial processing with quantum parallel processing. This substitution enables simultaneous exploration of multiple decoding paths, improving both throughput and reliability by avoiding iterative computational errors.
2Reliability
If LDPC codes are used for error correction, then error correction capability is improved, but processing requirements increase with complexity and size
Solution Approach 1:
The patent changes the fundamental processing parameter from serial iterative message passing to quantum parallel annealing. By reducing the belief propagation algorithm to a quadratic polynomial form suitable for quantum annealing, the system maintains strong error correction capability while dramatically reducing processing complexity. The quantum annealer naturally handles the combinatorial optimization problem without requiring multiple iterative passes.
Solution Approach 2:
The patent substitutes the complex serial processing architecture with a quantum annealing system that inherently handles the optimization problem in parallel. This replacement reduces device complexity by eliminating the need for multiple iterative processing stages and intermediate storage, while maintaining or improving error correction performance.
3Measurement precision
If more iterations are performed in serial message passing, then decoding accuracy may improve, but communication throughput decreases due to time consumption
Solution Approach 1:
The patent replaces the time-consuming serial iterative process with quantum annealing that performs the equivalent computation in parallel. The quantum system explores the solution space simultaneously through quantum tunneling and superposition, achieving high decoding accuracy without the iterative time penalty that limits throughput in conventional systems.
Solution Approach 2:
The patent transitions from sequential time-based iteration to quantum parallel processing, effectively adding a dimension of simultaneity to the computation. Multiple decoding possibilities are evaluated in parallel across the quantum state space, achieving both high accuracy and throughput by eliminating the sequential time constraint.
Data Source
AI summary
Systems and methods herein provide for error correction via Low Density Parity Check (LDPC) coding. In one embodiment, a system includes a data buffer operable to receive a block of Low Density Parity Check (LDPC) encoded data. The system also includes a processor operable to reduce a belief propagation algorithm used to encode the LDPC encoded data into a quadratic polynomial, to embed the quadratic polynomial onto a plurality of quantum bits (qubits), and to decode the block of LDPC encoded data via the qubits.