Running large, useful quantum algorithms on a quantum computer requires high-quality quantum bits (qubits), with error rates far below what is physically achievable with a single qubit. By using a quantum error correction code (QECC), where a logical qubit is encoded by combining multiple physical qubits, errors on the physical qubits can be detected – by measuring only some of them – and corrected, leading to a logical qubit with a lower error rate than any of its constituent physical qubits.

Decoding a QECC is a classical computing problem where the classical, binary measurement data are used to derive an operation that corrects the errors and maintains the state of the logical qubit. While various decoder algorithms have been proposed, this project extends current active work on a novel decoder based on hyper-dimensional computing (HDC) where all decoding information is encoded in the relations between very large binary vectors (thousands of bits), so-called hypervectors.

In the current state, the decoding accuracy of the constructed hypervectors is limited. The decoding accuracy can be improved by considering the relations between more combinations of hypervectors, however this leads to a massive increase in problem size, requiring advanced optimization techniques.

The goal of this project is to develop a method for constructing hypervectors for popular QECCs such that they result in a high-accuracy decoder that can compete with popular alternatives at a much higher decoding speed.

This project requires an interest in unconventional computing paradigms and solving large optimization problems.

For more information, contact S.Feld@tudelft.nl