چارچوب پیش بینی حملات و شکست ها برای یک شبکه تلفن همراه 5G مشترک
خلاصه
1. مقدمه
2 اثر مرتبط
3 چارچوب پیش بینی حملات و شکست ها
3.1 ویژگی های ماژول مهندسی
3.2 ماژول پیش بینی
3.3 ماژول تحریک
4 AFPA برای برشهای انعطاف پذیر شبکه 5G
4.1 مرحله استقرار و آموزش اولیه
4.2 فاز زمان اجرا
4.3 AFPA برای یک شبکه چند فناوری
5 آزمایش عددی
5.1 مدل ARIMA برای پیش بینی حملات
5.2 مدل پیش بینی شکست ها
5.3 پیش بینی ناهنجاری مدل ML
6. نتیجه گیری
منابع
Abstract
1 Introduction
2 Related works
3 Attacks and failures prediction framework
3.1 Features engineering module
3.2 Prediction module
3.3 Actuation module
4 AFPA for resilient 5G network slices
4.1 Deployment and initial training phase
4.2 Run-time phase
4.3 AFPA for a multi-technology network
5 Numerical experimentation
5.1 ARIMA model for attacks prediction
5.2 Failures prediction ML model
5.3 Anomaly prediction ML model
6 Conclusion
References
Abstract
Although mobile technologies keep evolving through years, Fault management and cyber-security management in mobile networks are still treated as separated notions with different blocks and different approaches whereas in practice, they are highly correlated. In this paper, we propose a framework that takes into account the correlation between these two management systems. The framework is based on several prediction agents where each agent is composed of a security predictor, a fault predictor and a generic anomaly detection model. A re-enforcement process allows to enhance the reliability of the machine learning training and prediction phases of the different predictors. Besides, each agent can collaborate with its neighborhood for a more resilient network. An application of this framework to 5G architecture is proposed by mapping the components of our framework with network slices. Finally, an experimentation is held over a testbed that we set up on openstack in order to forecast future anomalies related to proxy overload, latency violation in call session network functions and to excessive usage of memory. The training is achieved with ARIMA and deep learning models with promising results.
1 Introduction
As mobile network technologies evolve, new services are offered and more sophisticated networks are needed. The increasing number of Internet users leads to a redesign of network architecture, forcing designers to take into account new parameters such as the need of global coverage combined with low latency, as well as a high reliability and security level. Additionally, new networking experiences are added, such as Internetof-Things (IoT), which promise to offer new services and facilities to people’s daily lives. In this demanding environment, 5G technology is emerging, playing a decisive role in the implementation of new visions and promising to deliver solutions. A major innovation introduced by 5G technology [1] is the scalability. 5G architectures take into account the possible need of extending the capabilities of the network, both at the level of user traffic growth and at the level of new services input from providers. Slicing could be the ideal solution for such networks, offering scalability as well as flexibility in managing a giant network. Network Slicing is set to be a prominent feature of 5G to allow connectivity and data processing tailored to specific customers requirements. Mobile communications provided by smart networks will enhance the efficiency and productivity of business processes and will open up opportunities for network operators to address the Business-to-Business segment more effectively.