N-Queens CSP Solver

A Spring Boot web app that solves the N-Queens problem using three CSP algorithms — Backtracking, Forward Checking, and MAC — with optional MRV, Degree, and LCV heuristics. Includes an interactive frontend for visualizing and comparing results across algorithms.

---

Tech Stack

• Backend: Java 17, Spring Boot 3.2
• Build: Maven
• Frontend: HTML, CSS, JavaScript
• Libraries: AOS (scroll animations)
• Deployment: Railway

---

Features

• Three CSP Algorithms: Backtracking (baseline), Forward Checking (domain pruning after each placement), and MAC (arc consistency via AC-3 at every step).
• Three Heuristics: MRV (choose the most constrained variable), Degree (break ties by most-constrained), and LCV (order values by least impact on neighbors) — toggleable per run.
• Random Start: Optional random initialization to diversify search paths, enabling board sizes up to N=32.
• Interactive Board Visualizer: Renders the solved chessboard per algorithm with tabbed switching between results.
• Metrics Comparison: Displays constraint checks and solve time for each algorithm side-by-side.
• Preset Board Sizes: One-click presets (4, 8, 16, 20 — and up to 32 with Random Start enabled).

---

Process

1. Modeled N-Queens as a CSP: variables = columns, domains = rows, constraints = no two queens share a row, column, or diagonal.
2. Implemented Backtracking as the baseline recursive search.
3. Extended with Forward Checking to prune domain sizes after each queen placement.
4. Implemented AC-3 for MAC to propagate arc consistency more aggressively at each step.
5. Layered MRV, Degree, and LCV heuristics as optional enhancements on top of all three algorithms.
6. Exposed a /api/solve REST endpoint via Spring Boot and built a vanilla JS frontend to drive it.
7. Deployed to Railway.

---

Running the Project

mvn spring-boot:run

• Live at: n-queens.up.railway.app
• Or locally at http://localhost:8080