NetSecBed : A Container-Native Testbed for Reproducible Cybersecurity Experimentation

Leonardo Bitzki 1 2 , Diego Kreutz 2 , Tiago Heinrich 3 , Douglas Fideles 2 , Leandro Bertholdo 1 , Silvio Quincozes 2 , Angelo Diniz 2 · 2026

Read NetSecBed : A Container-Native Testbed for Reproducible Cybersecurity Experimentation by Leonardo Bitzki 1 2 , Diego Kreutz 2 , Tiago Heinrich 3 , Douglas Fideles 2 , Leandro Bertholdo 1 , Silvio Quincozes 2 , Angelo Diniz 2 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-04121.txt.

Introduction

Leonardo Bitzki and team built NetSecBed, a box-native testbed for same-again security tests. Public data sets usually sit static, with weak support for set re-runs and tracking across rules. It packs 60 attack cases, 9 target services, and benign traffic makers into single-purpose boxes with plain-text specs. Its line does runs, packet capture, log grabs, probing, feature pulls, and data set merging, cutting day-to-day bias.

Catalog record

CREATORLeonardo Bitzki 1 2 , Diego Kreutz 2 , Tiago Heinrich 3 , Douglas Fideles 2 , Leandro Bertholdo 1 , Silvio Quincozes 2 , Angelo Diniz 2
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE31924
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.04121

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

NetSecBed : A Container-Native Testbed for Reproducible Cybersecurity Experimentation Leonardo Bitzki 1 2 , Diego Kreutz 2 , Tiago Heinrich 3 , Douglas Fideles 2 , Leandro Bertholdo 1 , Silvio Quincozes 2 , Angelo Diniz 2 Abstract Resumo Cybersecurity research increasingly depends on reproducible evidence, such as traffic traces, logs, and labeled datasets, yet most public datasets remain static and offer limited support for controlled re-execution and traceability, especially in heterogeneous multi-protocol environments. This paper presents NetSecBed , a container-native, scenario-oriented testbed for reproducible generation of network traffic evidence and execution artifacts under controlled conditions, particularly suitable for IoT, IIoT, and pervasive multi-protocol environments. The framework integrates 60 attack scenarios, 9 target services, and benign traffic generators as single-purpose containers, enabling plug-and-play extensibility and traceability through declarative specifications. Its pipeline automates parametrized execution, packet capture, log collection, service probing, feature extraction, and dataset consolidation. The main contribution is a repeatable, auditable, and extensible framework for cybersecurity experimentation that reduces operational bias and supports continuous dataset generation. I Introduction Experimental cybersecurity research increasingly depends on reproducible evidence, such as traffic traces, logs, telemetry, and labeled datasets, to enable rigorous and comparable evaluation of detection, classification, and incident response mechanisms. Yet generating such evidence remains an open technical and methodological challenge. In modern environments, experimental fidelity is tightly coupled with ecosystem heterogeneity, spanning traditional services (e.g., HTTP, SSH, and SMB), IoT protocols (e.g., MQTT and CoAP), and emerging communication frameworks (e.g., Zenoh and XRCE-DDS), where even minor variations in software versions, timing, configuration, and background noise may significantly alter observed artifacts and compromise cross-experiment comparability. This creates a recurring trade-off between realism, operational cost, and reproducibility: higher realism typically increases deployment, instrumentation, and auditing complexity. This challenge is further aggravated by the growing obsolescence of public cybersecurity datasets. Continuous changes in service behavior, protocol stacks, and adversarial strategies progressively reduce the temporal validity of static collections, limiting both scientific evaluation and model training. Consequently, the central problem is no longer merely how to obtain data, but how to systematically generate experimental evidence in a repeatable, auditable, and extensible manner, preserving full traceability of origin, parameters, temporal context, and execution conditions. The state of the art remains fragmented. Existing testbeds often optimize isolated dimensions, such as physical realism, protocol specificity, low-cost deployment, or fidelity in specialized domains (e.g., IoT, IIoT, SCADA, and CPS), while still exhibiting recurring limitations in (i) attack coverage, (ii) protocol heterogeneity, (iii) experimental scalability, and (iv) support for systematic re-execution and continuous dataset generation. In practice, many approaches still rely on static collections, domain-coupled scenarios, or partially manual pipelines, which hinder comparability, auditing, and incremental evolution of the experimental environment. To address these limitations, we propose NetSecBed 1 1 1 NetSecBed is available at https://github.com/ANONIMIZADO_PARA_REVISAO . , a container-native, scenario-oriented, and natively extensible testbed for reproducible evidence generation in heterogeneous network environments. The current architecture integrates 60 attack scenarios, 9 target services, and parameterizable benign traffic models, all encapsulated as single-purpose containers and described through declarative metadata specifications. Unlike approaches centered on static datasets, NetSecBed provides a complete and auditable experimental pipeline, including scenario selection, parametrized execution, packet capture, log collection, service probing, feature extraction, and automatic dataset consolidation. Our central methodological hypothesis is that explicitly separating declarative scenario specification from orchestrated execution reduces operational variability, improves traceability, and enables incremental scalability of the attack catalog without coupling extensions to the system core. This design allows new attacks, services, and benign profiles to be incorporated through YAML specifications and containerized artifacts, supporting continuous sample renewal and improving the external validity of experimental results. This makes NetSecBed particularly relevant for cybersecurity experimentation in ubiquitous and pervasive computing environments, where heterogeneous devices, protocols, and continuously evolving traffic patterns are intrinsic characteristics. The main contributions of this work are: (i) an observable, container-native architecture for reproducible cybersecurity experimentation; (ii) a structured and extensible catalog of 60 attacks with support for multiple protocols and services; (iii) an automated pipeline for continuous artifact and dataset generation; and (iv) a methodological framework that reduces operational bias and improves comparability across heterogeneous experiments. II Related Work Recent cybersecurity testbeds span a broad spectrum of objectives, ranging from IoT vulnerability assessment and traffic generation to cyber-physical attack impact analysis. However, as summarized in Table I , the current state of the art remains fragmented across four critical methodological dimensions: scalability, protocol heterogeneity, attack coverage, and reproducibility. Tabela I: State-of-the-art comparison of cybersecurity testbeds. Testbed Scientific Contributions Scalability Protocols #Atk Target Services Main Limitations Gotham Reproducible IoT testbed; multi-protocol support; real dataset generation \sim 140 nodes / machine MQTT, CoAP, RTSP 18–25 MQTT broker, CoAP, RTSP Restricted service scope; limited device diversity; no MITRE ATT&CK mapping RESLab Security experimentation and dataset collection \sim 50 substations DNP3 4 DNP3 masters / outstations Very limited attack diversity; single-protocol scope; no MITRE mapping IoT Security Testbed Open-source automated vulnerability assessment platform for IoT devices Low (1 device / run) HTTP(S), UDP, FTP, SSH, Telnet 13 Camera, smart bulb, IoT web services Few attacks; minimal scalability; no native IoT protocols or MITRE mapping LoRaWAN MitM LPWAN security testbed for real over-the-air experiments Very low LoRaWAN 1.0.2 2 LoRa end devices, gateway Very limited attack coverage; radio-dependent; unrealistic assumptions CPS Microgrid Real-time co-simulation for CPS attack impact analysis Very low Modbus, IEC 61850, DNP3 3 IEDs, RTUs, PMUs Low scalability; protocol limitations; proprietary hardware dependence QR-GridEx HIL-based large-scale coordinated attack simulation for smart grids Medium DNP3, IEC 61850, C37.118 10 SCADA/WAMS, substations Limited attack catalog; low reproducibility; energy-specific scope Lightweight SCADA Traffic generation and attack collection for SCADA security datasets Lightweight Modbus 2 SCADA-LTS, Conpot Only two attacks; single-protocol scope; no quantitative metrics or MITRE mapping NetSecBed (ours) Reproducible attack pipeline; structured attack catalog; feature extraction; observability and extensibility by design High (container-native) Web, MySQL, SSH, Telnet, SMB, MQTT, CoAP, Zenoh, XRCE-DDS 60 Web, DB, SSH, SMB, IoT/IIoT brokers External threat focus; no firmware/CVE-specific or industrial protocol attacks Recent survey studies have reinforced the role of cybersecurity testbeds as scientific infrastructures for reproducible experimentation, dataset generation, and systematic benchmarking. In IoT and IIoT domains, recent reviews consistently highlight persistent challenges in scalability, protocol heterogeneity, reproducibility, and experimental realism [ 5 , 1 ] . Similar limitations have also been reported in evaluation frameworks for anomaly detection in built environments, further emphasizing the need for large-scale, repeatable, and well-instrumented infrastructures for trustworthy security research [ 2 ] . Within this landscape, representative testbeds typically optimize specific methodological dimensions. In the IoT domain, the Gotham Testbed [ 4 ] provides a reproducible architecture with multi-protocol support (MQTT, CoAP, and RTSP) and moderate scalability, reaching approximately 140 nodes per machine. However, it remains restricted to a narrow set of IoT services, limited device diversity, and lacks explicit alignment with standardized adversarial taxonomies such as MITRE ATT&CK. Similarly, the IoT Security Testbed [ 14 ] advances automated vulnerability assessment for real devices, but offers only 13 attacks, low scalability, and no support for native IoT messaging protocols such as MQTT or CoAP, limiting its representativeness in modern IoT/IIoT scenarios. A second line of work focuses on industrial control systems and SCADA experimentation. RESLab [ 17 ] and the Lightweight SCADA Testbed [ 8 ] provide relevant

Read the complete text