Expert Analysis

A core feature of the platform, 5G network slicing technology is leveraged to dynamically allocate logically isolated dedicated networks for distinct training cohorts. Each slice constitutes a complete end-to-end 5G subnetwork with independent identifiers (S-NSSAI), resource allocations, and control policies. This ensures non-interference between concurrent training tasks and authentically replicates future 5G industry-specific private network deployments, allowing for isolated attack-defense scenarios.

For example, a 'Basic Protocol Analysis' slice might contain streamlined 5G core network functions, while an 'Advanced Offense/Defense Confrontation' slice includes a complete 5G core, multiple base station simulators, and pre-configured virtual network functions with known vulnerabilities.

The platform integrates a powerful security toolchain including network vulnerability scanners, Intrusion Detection/Prevention Systems (IDS/IPS), mimicry firewalls, Web Application Firewalls (WAFs), and diverse honeypot systems. These tools serve as objects for analysis and as defensive mechanisms during adversarial exercises. A dynamic threat intelligence module continuously monitors industry threat feeds, updating the attack simulation library with new attack vectors. Future plans include applying artificial intelligence (AI) for intelligent attack variant generation and automated security log analysis.

The platform features a systematic, story-driven attack script design covering the complete attack chain, from peripheral penetration to core compromise. Scenarios are divided into Access and Network Element Layer Attacks (e.g., jamming, fake gNB), Network and Service Layer Attacks (e.g., slice management misuse, API abuse), and Application and Data Layer Attacks (e.g., industrial control data tampering, privacy theft). The built-in attack simulation engine automates execution and dynamically updates scripts based on real-world threat intelligence.

Multi-round comparative studies across three academic semesters demonstrated statistically significant superiority of the experimental group using the platform over the control group. Metrics like task completion rates, vulnerability identification capabilities, and emergency response latency showed marked improvements. The platform's high-fidelity environment, combined with gamified adversarial elements, significantly elevated learners' intrinsic motivation and exploratory engagement, providing objective foundations for formative assessment.