Current quantum computing devices have different strengths and weaknesses depending on their architectures. This means that flexible approaches to circuit design are necessary. We address this task by introducing a novel space-efficient quantum optimization algorithm for the graph coloring problem. Our circuits are deeper than the ones of the standard approach. However, the number of required qubits is exponentially reduced in the number of colors. We present extensive numerical simulations demonstrating the performance of our approach. Furthermore, to explore currently available alternatives, we also perform a study of random graph coloring on a quantum annealer to test the limiting factors of that approach, too.
Current quantum computing devices have different strengths and weaknesses depending on their architectures. This means that flexible approaches to circuit design are necessary. We address this task by introducing a novel space-efficient quantum optimization algorithm for the graph coloring problem. Our circuits are deeper than the ones of the standard approach. However, the number of required qubits is exponentially reduced in the number of colors. We present extensive numerical simulations demonstrating the performance of our approach. Furthermore, to explore currently available alternatives, we also perform a study of random graph coloring on a quantum annealer to test the limiting factors of that approach, too.
Current quantum computer prototypes exhibit various strengths and weaknesses based on their individual architectures. Consequently, the development of flexible circuit design approaches becomes a critical tool. To tackle this challenge in the case of clique cover optimization, we propose a novel embedding method of this problem, efficiently reducing the width of the required quantum circuit exponentially in the number of cliques. We compare this new embedding based on a Polynomial Unconstrained Binary Optimization reformulation to the traditional one-hot encoded quadratic problem and present numerical simulations of the Quantum Approximate Optimization Algorithm for both methods, studying in particular the solution quality. The results obtained from the numerical experiments provide valuable insights into the performance gains achieved by employing our space-efficient embedding method, verified through both analytical and stochastic simulation benchmarks. Our results highlight the effectiveness of the space-efficient embedding method for the clique cover problem and extend the set of problems formulated with this technique.
Graph coloring is a well-known NP-complete problem with applications in scheduling, register allocation, frequency assignment, and network optimization. Quantum computing offers the potential for polynomial or even super-polynomial speedups in certain problem instances, yet practical quantum graph coloring methods must carefully balance qubit resources, circuit depth, and constraint enforcement. We proposed a solver agnostic quantum framework for K-coloring. Along with a novel encoding and efficient constraint implementation for its exact coloring exploiting Grover search and almost optimal coloring by the Quantum Approximate Optimization Algorithm \emph{(QAOA)}, and \emph{Quantum Annealing}. Unlike using O(NK) qubits as in \emph{SOTA}, we reduced the qubit requirements to O(Nlog2K), along with optimized comparator circuits enforcing adjacency constraints. Further symmetry-based graph reduction is incorporated as an optional preprocessing step to further reduce the instance size before quantum execution.
Genetic algorithms (GA) are computational methods for solving optimization problems inspired by natural selection. Because we can simulate the quantum circuits that implement GA in different highly configurable noise models and even run GA on actual quantum computers, we can analyze this class of heuristic methods in the quantum context for NP-hard problems. This paper proposes an instantiation of the Reduced Quantum Genetic Algorithm (RQGA) that solves the NP-hard graph coloring problem in O(N 1/2 ). The proposed implementation solves both vertex and edge coloring and can also determine the chromatic number ( i.e ., the minimum number of colors required to color the graph). We examine the results, analyze the algorithm convergence, and measure the algorithm's performance using the Qiskit simulation environment. Our Reduced Quantum Genetic Algorithm (RQGA) circuit implementation and the graph coloring results show that quantum heuristics can tackle complex computational problems more efficiently than their conventional counterparts.
Quantum computing holds the potential to efficiently solve certain problems believed to be intractable for classical computers. Realizing this potential, however, requires substantial physical resources, including large qubit counts, sufficient connectivity, and long coherence times. As current quantum devices are still far from this regime, the efficient compilation of quantum algorithms into hardware-compatible instructions remains a central challenge in quantum computing, especially for near-term applications. A representative example of such applications is combinatorial optimization, where exponentially large search spaces motivate quantum algorithms that leverage interference to approximate solutions more efficiently than classical heuristics, even at modest system sizes. Two prominent paradigms in this context are quantum annealing and the Quantum Approximate Optimization Algorithm (QAOA): Quantum annealing realizes optimization via continuous-time quantum dynamics implemented on analog hardware, while QAOA is a variational, gate-based approach that builds parameterized quantum circuits from problem-specific Hamiltonians. Both paradigms require compilation strategies tailored to the constraints of their respective hardware architectures. Most existing compilation strategies are problem-agnostic and therefore broadly applicable, yet often suboptimal for specific tasks. This thesis uses the graph coloring problem as a case study to investigate a problem-driven approach to quantum compilation, where structural features of the problem class are exploited to design tailored compilation schemes that offer improved efficiency and scalability for the targeted application. To address hardware connectivity constraints, the presented approaches adopt the parity framework, which encodes long-range interactions into single-qubit terms. This thesis develops two novel problem-driven compilation methods for the graph coloring problem, the Bandwidth-Optimized LHZ layout and the 2D-Twine circuit. The former leverages bandwidth minimization to optimize parity encodings for analog quantum annealers, while the latter constructs QAOA unitaries for gate-based optimization using parallelized Parity Twine circuits. Both approaches are benchmarked against general-purpose methods and demonstrate substantial improvements in qubit usage or circuit complexity. These results illustrate the advantages of incorporating problem structure into the compilation process and provide a proof-of-concept for the effectiveness of problem-driven compilation strategies.
Abstract In the era of Noisy Intermediate-Scale Quantum (NISQ) computers it is crucial to design quantum algorithms which do not require many qubits or deep circuits. Unfortunately, most of the well-known quantum algorithms are too demanding to be run on currently available quantum devices. Moreover, even the state-of-the-art algorithms developed for the NISQ era often suffer from high space complexity requirements for particular problem classes. In this paper, we show that it is possible to greatly reduce the number of qubits needed for the Travelling Salesman Problem (TSP), a paradigmatic optimization task, at the cost of having deeper variational circuits. While the focus is on this particular problem, we claim that the approach can be generalized for other problems where the standard bit-encoding is highly inefficient. Finally, we also propose encoding schemes which smoothly interpolate between the qubit-efficient and the circuit depth-efficient models. All the proposed encodings have the same volume up to polylogarithmic factors and remain efficient to implement within the Quantum Approximate Optimization Algorithm framework.
Abstract Graph coloring with minimum colors is a well-known NP-complete problem in computer science and mathematics, having diverse real-life applications such as scheduling, register allocation, and frequency assignment. However, a classical computer requires exponential time to solve the graph coloring problem with the minimum number of colors. Therefore, classical approaches become infeasible for solving this problem on large graphs. In contrast, quantum computing offers the potential for speedup on specific problems, although solutions are typically obtained with a certain probability of correctness. This work proposes a quantum–classical method for solving the graph coloring problem and finding the chromatic number of a graph with guaranteed correctness, and also reduces the time and qubit complexities. The proposed method is inspired by Randomized Grover’s search algorithm, and the classical binary search algorithm. Here, Randomized Grover’s search algorithm provides quadratic speedup and this algorithm is useful in real situations when the number of solutions is not known in advance in contrast to the Grover’s search algorithm. The binary search algorithm helps to achieve logarithmic speedup, which reduces overall complexity. The experimental results strongly validate our claims and demonstrate the superiority of the approach. Furthermore, a comprehensive mathematical analysis is carried out to examine its complexities, as well as to compare its computational efficiency with existing methods.