4. Bundle Detection

4.1 Coordinated Trading Activities

High-volume memecoins are particularly susceptible to coordinated pump-and-dump schemes. Our system detects bundling by:

  • Monitoring trading intervals for abnormally high trade frequencies.

  • Linking wallet addresses that repeatedly transact in unison.

  • Statistical correlation analysis to measure the relationship between price movements and group trading behavior.

4.2 Sniper Group Identification Algorithms

Sniper bots aim to front-run or back-run trades at lightning speed. We employ:

  • Kalman Filtering to track real-time price fluctuations and estimate state vectors of suspected sniper wallets.

  • A specialized Sniper Detection Metric SSS given by:

S=∑j=1m(γj×Δtj×Wj)mS = \frac{\sum_{j=1}^{m} \bigl(\gamma_j \times \Delta t_j \times W_j \bigr)}{m}S=m∑j=1m​(γj​×Δtj​×Wj​)​

Where:

  • Δtj\Delta t_jΔtj​ is the inter-transaction latency.

  • WjW_jWj​ is the weighted magnitude of trades.

  • γj\gamma_jγj​ is a factor quantifying historical success rates of similar strategies.

A higher SSS indicates a higher likelihood of sniper activity.

Last updated