Advanced Hashing Models

A research and benchmarking suite for advanced hash function models, including perfect hashing, cryptographic hashing, and hybrid approaches. This project provides high-performance C++ implementations, benchmarking tools, and result visualization.

---

Project Structure

├── include/         # Header files for all hash models and utilities
├── src/             # C++ source files for hash models, main, and benchmarks
├── build/           # Build output (executables, intermediate files, results)
├── CMakeLists.txt   # CMake build configuration
├── benchmark_results.csv # Output: benchmark results (CSV)

Key Directories & Files

• include/: All hash model headers (see below for details)
• src/: Implementations, main entry, and benchmarking logic
• build/: Output directory after building (executables, results)
• CMakeLists.txt: CMake build script

---

Building the Project

Prerequisites

• C++17 compatible compiler (MSVC, GCC, or Clang)
• CMake 3.15+

Build Instructions (Windows, Linux, macOS)

1. Clone the repository

   git clone <repo-url>
   cd hash_project

2. Configure and build with CMake

   mkdir build
   cd build
   cmake ..
   cmake --build . --config Release

   • Executables will be in build/Release/ (Windows) or build/ (Linux/macOS)

---

Running Benchmarks

• Run all benchmarks:

  # From build directory
  ./hash_project         # (Linux/macOS)
  .\Release\hash_project.exe  # (Windows)

• Run throughput benchmark:

  ./throughput_benchmark         # (Linux/macOS)
  .\Release\throughput_benchmark.exe  # (Windows)

• Results:
  • Benchmark results are saved as CSV in build/ (e.g., benchmark_results.csv)

---

Hash Model Overview

Perfect Hashing

• BasicPerfectHash (FKS): Two-level perfect hashing (Fredman-Komlós-Szemerédi)
• BDZ_MPHF: Minimal perfect hash (Belazzougui-Botelho-Dietzfelbinger)
• CuckooPerfectHash: Cuckoo-based perfect hashing

Cryptographic Hashing

• BasicCryptoHash: SHA-256 baseline implementation
• SipHash: Fast keyed hash (DoS-resistant)
• BLAKE3Hash: High-performance cryptographic hash
• MurmurHash3: Fast, non-cryptographic hash

Hybrid & Advanced Models

• UltimateHybridHash: Multi-stage (SipHash → BDZ → BLAKE3 → cache-aware)
• AdaptiveSecurityHybrid: Runtime-tunable cryptographic verification
• ParallelVerificationHybrid: Latency hiding via parallel verification
• CachePartitionedHybrid: Hardware-aware, Bloom filter-optimized
• UltraLowMemoryHybrid: Minimal storage, on-demand verification
• TwoPathHybrid: Dual-lane (fast/secure) hybrid
• HybridCryptoPerfect: SipHash preprocessing + BDZ MPHF
• HybridPerfectVerify: Cuckoo perfect index + BLAKE3 verification

---

File Descriptions

include/

• base_hash.hpp — Abstract base class for all hash models
• benchmark_suite.hpp — Benchmarking utilities and result structures
• basic_perfect_hash.hpp — FKS two-level perfect hashing
• bdz_mphf.hpp — BDZ minimal perfect hash
• cuckoo_perfect_hash.hpp — Cuckoo-based perfect hash
• basic_crypto_hash.hpp — SHA-256 implementation
• siphash.hpp — SipHash-2-4 (keyed hash)
• blake3_hash.hpp — BLAKE3-inspired hash
• murmur_hash.hpp — MurmurHash3 (fast, non-crypto)
• ultimate_hybrid.hpp — Multi-stage hybrid (SipHash, BDZ, BLAKE3)
• adaptive_security_hybrid.hpp — Security-level-tunable hybrid
• parallel_verification_hybrid.hpp — Parallelized hybrid
• cache_partitioned_hybrid.hpp — Cache/Bloom filter hybrid
• ultra_low_memory_hybrid.hpp — Minimal memory hybrid
• two_path_hybrid.hpp — Dual-lane hybrid
• hybrid_crypto_perfect.hpp — SipHash + BDZ hybrid
• hybrid_perfect_verify.hpp — Cuckoo + BLAKE3 hybrid

src/

• main.cpp — Main entry, runs all benchmarks
• benchmark_runner.cpp — Factory and runner for all models
• throughput_benchmark.cpp — Throughput-focused benchmark
• *.cpp — Implementations for each hash model

Other

• benchmark_results.csv — Output: benchmark results

---

Credits

• Authors: Mahir Rešidović and Nedim Bečić
• Research: Faculty of Electrical Engineering Sarajevo

---

Contact

For questions, suggestions, or contributions, open an issue or contact the maintainers:

• mresidovic1@etf.unsa.ba
• nbecic1@etf.unsa