Testing at the edge: the challenges of Secure Mobile Edge Computing
A New Approach to Network Architecture
Fulfilling the potential of 5G will bring the need not only for new air interfaces and frequency bands, but also for an innovative and more dynamic approach to network architecture. 5G networks will have to be capable of providing flexible allocation of computing resources in order to enable all the proposed use cases to co-exist. Network Functions Virtualization (NFV) and Centralized Radio Access Network (C-RAN) will both form a key part of the new networks, but moving vast amounts of data around the cloud introduces latency and can begin to consume significant amounts of power, both of which would conflict with the broader aims of 5G communications.
Mobile Edge Computing (MEC) – where certain processing functions are located close to the user – is gaining recognition as one of the key enabling technologies for 5G, and will be particularly relevant for critical low-latency Machine Type Communications such as connected cars and virtual healthcare, as well as for speeding up data-hungry services such as video streaming and packet brokering. As the push for all IP networking was a 4G milestone, locating intelligence at the network edge enables services and content to be activated closer to the subscriber is key to next generation mobile network. This reduces latency because it minimises the amount of data that has to be moved between the user and the core, and also reduces the traffic load on the network.
MEC Technology to Open Up New Business Segments
MEC is being promoted by ETSI, which has an Industry Specification Group (ISG) dedicated to developing MEC technology as a means of opening up new vertical business segments and services for consumers and enterprise customers – use cases are expected to include the Internet of Things (IoT), location-based services, augmented reality (AR) and the distribution of local content. It is also seen as an enabler towards the convergence of fixed and mobile communications. The main participants in the ISG include Huawei, IBM, Intel, Nokia Networks, NTT DOCOMO and Vodafone. ETSI estimates that deploying MEC could result in backhaul and transport savings of up to 35%, and an improved QoE benchmarked by a 20% improvement in the time to load a web page. However, this must not come at the cost of security and MEC must encapsulate the role of the Security Gateway (SeGW).
Although it is slated as an enabler for 5G, MEC is already starting to be adopted in 4G networks for local caching of on-demand video content, which improves throughput and hence also user experience. As the network becomes more software-defined and the content delivery network (CDN) moves closer to the user, resources can be flexibly increased or reduced according to demand. Cloud applications can then access real-time radio network information identifying the user and their context, allowing the delivery of personalized services like AR or localized product promotions.
MEC will also mean that operators will be able to open up their Radio Access Network (RAN) edge to authorized third-parties, further enabling the rapid and flexible deployment of new applications and services.
Centralizing RAN Functions?
At first glance, centralizing RAN functions may appear at odds with MEC, but virtualization is all about making sure that each network function is deployed where it has the best effect, whether this is at the customer premise, network edge or at the core. For many applications, the mobile edge is the most appropriate location for additional processing, and this is consistent with the principles of NFV, where the processing requirements of distributed SeGW must not diminish the gains of the MEC. Encryption slows down application traffic and the virtual SeGW must be deployed with varying capacity, based on the local demands of the cell-site.
As a result, the Secure MEC requires thorough and rigorous testing. As it forms part of the broader software-defined and virtualization ecosystem, the most efficient way to validate MEC is to utilise the same type of virtual technologies, and TeraVM offers the ideal way of achieving this. TeraVM provides validation solutions for Physical and Virtual SeGW functions that are deployed in the core and edge, measuring the QoE for OTT Video, VoLTE and data applications ensuring the benefits of Secure MEC does not outweigh the performance it promises.