In the next decade, communication technologies will be dominated by the fifth generation (5G) of cellular networks, which is currently under deployment in many countries. Although at the moment 5G consists of an improvement in the performance of 4G, the future versions of 5G are expected to revolutionize wireless communications.
One of the technologies that will make the 5G revolution possible is the Multi-access Edge Computing (MEC). The 5G MEC consists of a cloud platform that is deployed close to the 5G end user. In this way, the 5G MEC will make possible the future 5G applications that require low delays, with a high impact on use cases such as eHealth, Smart City, Industry 4.0, and automotive.
Motivated by the important role of security and dependability in the maturity of a technology, 5G-MODaNeI aims to improve both these aspects in 5G MEC. To achieve this, 5G-MODaNeI focuses on allocating together the network resources of 5G and the data resources of MEC.
5G-MODaNeI consists of three main parts. In the first part, the risks and the vulnerabilities for attacks and failures in 5G MEC will be identified. The impact of joint data and network resource allocation on security and dependability in 5G MEC will be determined. In the second part, innovative and intelligent solutions will be created. These solutions will allocate the network and data resources to maximize the security and dependability in 5G MEC. In the third part, a 5G-MEC experiment in a vehicular scenario will be developed and used to test the proposed solutions.
In the first years of the project, the standards of MEC provided by the European Telecommunications Standards Institute (ETSI) have been studied. The study also focused on the integration of the MEC architecture with the 5G architecture and the use of other general virtualization technologies, such as Network Function Virtualization, in implementing MEC. The project has continued by evaluating the security, dependability, and performance aspects of the 5G-MEC architecture. The state of the art has been analyzed and the challenges of jointly addressing the three aspects have been discussed.
A first evaluation of the dependability of a 5G-MEC system has been performed by proposing a model of the system availability, i.e., the readiness of providing a service with the desired requirements. The results show the importance of having proper redundancy and dependable software in the management elements of the 5G and MEC. Afterwards, the model has been extended to include security attacks and connectivity between the elements of the 5G-MEC system. The results highlight the importance of the attack detection and recovery. A new work will evaluate the impact of simultaneous attacks and failures of multiple elements of the 5G-MEC system on the overall system availability.
Novel solutions to allocate data and network resources in a 5G-MEC scenario have been proposed. These solutions are based on various techniques: heuristic, mathematical optimization, and Machine Learning (ML). One set of the works is also addressing the security and dependability of ML-based solutions, which may be attacked by adversaries or subjected to failures. Another work proposes a ML-based solution for the MEC application migration, i.e., moving the MEC application from one MEC server to another. The MEC application migration is due to the mobility of the user that uses the MEC application or the failure of the MEC server where the MEC application is running. This work will be extended to also consider privacy in the MEC application migration. Another work proposes a ML-based solution for the selection of the path between a user and the related MEC server by optimizing multiple performance metrics.
A hybrid 5G-MEC testbed has been developed. The testbed integrates various 5G and MEC simulators and emulators, includes actual MEC servers, and proposes a controller. The controller allows the experimentation on MEC application migration and path selection. The controller interacts with the MEC framework by using standard ETSI MEC interfaces and allows the use of complex migration solutions.
The fifth generation (5G) of cellular networks is currently consisting in a huge improvement of performance with respect to 4G, but the future releases of 5G are going to revolutionize the wireless communications. One of the technologies that will make the 5G revolution possible is the Multi-access Edge Computing (MEC). The MEC in 5G consists in the provision of computing and storage resources within the edge of the Radio Access Network and is enabled by virtualization and softwarization technologies, such as Software-Defined Networking (SDN), Network Function Virtualization (NFV), and Network Slicing. The 5G MEC will effectively enable real-time 5G applications in use cases such as eHealth, Smart City, Industry 4.0, and automotive.
The project focuses on security and dependability in the orchestration of data and network resources in 5G MEC. The project has three main objectives: investigate and characterize security and dependability in the orchestration of data and network resources in 5G MEC; develop and analyze intelligent solutions for jointly orchestrating data and network resources for dependability and security in 5G MEC; develop a 5G-MEC testbed in automotive and verify the proposed solutions.
In the project, models will identify the risks and the vulnerabilities for attacks and failures in 5G MEC. The impact of joint data and network resource allocation on security and dependability in 5G MEC will be determined. Moreover, after formulating the problem of resource allocation for security and dependability in 5G MEC, innovative and intelligent solutions will be proposed. The solutions will allow both long-term design and real-time adaptation for maximizing security and dependability in 5G MEC and the future 5G applications. Finally, a 5G MEC testbed in automotive will be developed and it will be used to analyze the performance of the proposed solution in an experimental scenario.