Welcome to the tea party my Sony🐭☕️

I decided to make you a compilation of 5G security material 📡
Here will be some articles and increments on attacks and detection

A little more about detection.

Since 5g has a tightly implemented interaction with neural networks, attacks on it are also detected. That's why some of the repositories I provided on 5G attack detection will be mostly related to them. 

If you don't know much about them, here's a pretty good book as an introduction to neural networks + in the future I can make a gist about them 🧬


5GAD

This is a 5G attack detection dataset for machine learning tools 

It contains two types of intercepted network packets-normal packets of non-malicious traffic and attack packets from 5G kernel attacks implemented with free5GC.
The captures were collected using tshark or Wireshark on 4 different network interfaces in the 5G kernel.
The files beginning with "allcap" contain packets recorded simultaneously on all four interfaces, the other *.pcapng files represent the same data that was split on one of the four interfaces.
If you are building your own neuronics this might be useful for training


5G_security_malicious_traffic_detection

This project is from NSW Hackathon 2020. 
It introduces the concept of computing for security by taking streaming traffic from the RAN and looking for malicious traffic on the mobile edge segment side to quickly recognize threats. 

In this way, it differs from solutions that must store data from the underlying network and assess possible threats in a second step.
Time window feature aggregation will be evt. Implemented with Spark.
ASD: Anomaly Symptom Detection. It is located in the RAN and is designed to quickly find any traces or signs of anomaly in the network traffic generated by UEs connected to the RAN.
NAD: Network Anomaly Detection. This is a collector of timestamped and RAN-related symptoms generated and forwarded by the ASD, where a central process, such as EPC, analyzes the timeline and relationship of these symptoms to detect any network anomaly. As soon as an anomaly occurs, it is immediately forwarded to the monitoring and diagnostics module.
Server diagnostics and policy. A GUI environment that creates graphs of interconnected infected devices and, in the case of botnets, also botmasters. This gives the expert the ability to look deeper into the flow of packets and initiate preventive actions. Only graph generation is implemented.
You will be interested if you want to understand the attack detection


Distributed-Slice-mobility-attack-detection

An article about how inter-segment abundance in 5G networks enables mobility of user sessions from one network segment to another. 

It will mention a new attack on 5G network segments using user-initiated mobility between segments. The performance and economic damage from a DSM attack is higher than from attacks like DDOS and Yo-Yo. It will also describe building a deep learning model to effectively detect such attacks.
#ss7 #mobile #network #5g #neural_networks #attacks #distinction