Graph Neural Network for Early DDoS Detection: Evaluating the Impact of Progressive Node Reduction in Flow Graphs
Bibliographic record
Abstract
Distributed Denial of Service (DDoS) attacks represent a significant threat to network intrusion, often overwhelming systems by flooding them with traffic. Graph Neural Networks (GNNs) have grown as powerful tools in detecting cyberattacks due to their ability to effectively model and analyze network traffic as graphs. This thesis explores how progressively reducing the number of nodes in flow graphs influences the detection capabilities of a GNN in identifying DDoS attacks. The CICIDS2017 dataset provided by the Canadian Institute for Cybersecurity, widely utilized in cybersecurity research, was employed for experiments. Due to limitations in the dataset’s aggregated format, packet flows between IP addresses were artificially simulated. Endpoint graphs were constructed by sequentially reducing nodes to represent earlier stages of network interactions, simulating scenarios where only partial traffic flows are observable. A Graph Isomorphism Network (GIN) was chosen for its known effectiveness in capturing complex graph structures. Experimental results revealed that the accuracy of the GNN decreased progressively from 93.50\% (full graph) down to 80.27\% when only 10\% of the nodes were retained. This finding highlights a significant relationship between graph completeness and detection accuracy, demonstrating that GNNs become progressively less effective when provided with fewer nodes from the early stages of network communication. Furthermore, the study identifies a limitation related to the structural simplicity among generated graphs, suggesting that increased variation and complexity could lead to higher accuracy.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.003 | 0.002 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".