probabilistic-packet-filter

Project Bi-Weekly Update: Implementing Bloom Filters & DDoS Attack Simulation

Student: Jonathan Ami Date: Feb 23, 2025

Planned Activities:

Implement a Bloom Filter for improved probabilistic packet filtering.
Simulate a DDoS attack to evaluate performance under high load.
Benchmark false positive rates and filtering efficiency.
Optimize filtering parameters to minimize overhead.

Progress Update:

Implementation of Bloom Filter:

In this update, we have integrated a Bloom Filter into our probabilistic filtering system to enhance accuracy and reduce false positives. Key modifications include:

Multi-hash function support for improved distribution.
Optimized bit-array size selection to balance memory use and false positive rates.
Support for dynamic insertions without excessive memory expansion.

These changes were inspired by recent updates in the Probabilistic Packet Filter Repository, where similar filtering optimizations were applied for enhanced performance.

DDoS Attack Simulation & Testing:

A large-scale DDoS attack simulation was implemented to stress-test the filtering system:

Attack packets and normal traffic were generated at high volume.
Packets were screened using the Bloom Filter.
Results were logged to measure accuracy, efficiency, and false positives.

Changes in the Repository:

Refactored probabilistic filter implementation to use Bloom filtering.
Added new hashing functions for improved accuracy.
Integrated logging and benchmarking for in-depth performance analysis.
Updated test cases to validate Bloom Filter efficiency.

Next Steps:

Fine-tune filter parameters for adaptive threshold adjustments.
Benchmark bloom filter against regular hashsets
Extend real-world testing with live network traffic datasets.
Explore parallelized filtering for higher throughput scenarios.
Update GitHub Pages documentation with latest results and insights.