Network Digital Twin-Generated Dataset for Machine Learning-based Detection of Traffic Congestion Problems
Authors/Creators
-
1.
Universidad Politécnica de Madrid
Description
Overview
This record contains a set of synthetic datasets generated for the analysis and detection of congestion problems in realistic virtualized networks. The dataset collects metrics obtained via SNMP from network elements, enabling the capture of detailed information on device operational status and performance. These data are essential for training machine learning models aimed at anticipating, detecting, and preventing congestion episodes.
To obtain this information without impacting the performance of the scenario, specific probes were deployed within the Network Digital Twin (NDT), capable of collecting metrics in a non-intrusive manner. In addition, data flows emulating congestion or degradation conditions were generated on the scenario, enriching the dataset and providing greater realism to the modeled situations.
To address this challenge, a Network Digital Twin (NDT) approach was employed to emulate realistic network conditions and traffic patterns, enabling the automated generation of labeled data suitable for advanced analysis and early detection of congestion problems.
Feature Set:
📌 General Information
-
Interfaces: list of monitored interfaces
-
Host Name: device name
-
Timestamp: time marker
-
Number of detected interfaces
📌 TCP Metrics
-
tcpOutRsts -
tcpInSegs -
tcpOutSegs -
tcpPassiveOpens -
tcpRetransSegs -
tcpCurrEstab -
tcpEstabResets -
tcpActiveOpens
📌 UDP Metrics
-
udpInDatagrams -
udpOutDatagrams -
udpInErrors
📌 Per-Interface Metrics (suffixes: *.3, *.4, *.5, .6)
-
Traffic in bytes:
ifInOctets,ifOutOctets -
Traffic in packets:
ifInUcastPkts,ifOutUcastPkts,ifInNUcastPkts,ifOutNUcastPkts -
Discards:
ifInDiscards,ifOutDiscards -
Host statistics:
hostInPkts,hostOutPkts -
Errors and collisions:
etherStatsCollisions,etherStatsCRCAlignErrors -
Abnormal packets:
etherStatsUndersizePkts,etherStatsOversizePkts,etherStatsFragments,etherStatsJabbers -
Volume and distribution:
etherStatsOctets,etherStatsPkts,etherStatsBroadcastPkts,etherStatsMulticastPkts
(each repeated per interface: .3, .4, .5, .6)
📌 Bandwidth Metrics per Interface
-
bw_rxandbw_tx(per interface .3, .4, .5, .6)
📌 Dataset Label
-
LABEL0/1/2/3/4
Dataset Variations:
To accommodate diverse research needs and scenarios, the dataset is provided in the following variations. Traffic bandwidth ranges emulated are as follows:
-
Range 0: 1–10 Mbps
-
Range 1: 11–40 Mbps
-
Range 2: 41–70 Mbps
-
Range 3: 71–90 Mbps
-
Range 4: Greater than 91 Mbps
dataset_01_TC31_05062025_labeled_final.csv- The bandwidth range distribution is approximately homogeneous, with around 20% for each generated range (ranges 0, 1, 2, 3, and 4).
dataset_02_TC31_07062025_labeled_final.csv- The bandwidth range distribution is approximately homogeneous, with around 50% for each generated range (ranges 3 and 4)
dataset_03_TC31_10062025_labeled_final.csv- In this case, a higher proportion of ranges 0, 1, and 2 has been enforced, while ranges 3 and 4 are reserved for emulating specific bursts exceeding 91 Mbps, up to 180 Mbps.
- For ranges 0, 1, 2, and 3, the duration is between 60 and 120 seconds, whereas for burst type 4, the duration ranges between 5 and 40 seconds.