Over the years large network operators saw a need to seamlessly merge their fixed and mobile solutions. The problem is, considering the difference in technologies, this convergence is implemented at the service level (ie IP multimedia subsystem). Actual physical infrastructure convergence is needed to help mobile operators lower CAPEX, OPEX and seamlessly integrate mobile edge computing technologies, pushing them one step closer to 5G viability.
“Mobile” to become “Multi-access” in ETSI’s Mobile Edge Computing
Several edge computing initiatives exist, the most prominent being: the open edge computing group, the Openfog consortium, and of course ETSI’s (European Telecom Standardizations Institute) MEC. Last year ETSI decided to extend the reach of its edge computing architecture to encompass not only mobile access networks, but fixed access networks as well (DSL, DOCSIS, Wi-Fi). These changes are set to take place early this year.
This strategic shift is a move that envisions network convergence as a means to ultimately lower both CAPEX and OPEX for network operators. Instead of focusing solely on accelerating mobile RAN’s, ETSI’s goal is to apply edge computing uniformly across all access networks.
Why it’s needed
Over time , average internet bandwidth usage has exploded, driven mostly by video consumption. Due to the much higher cost of mobile bandwidth compared to using fixed networks (i.e Wi-Fi- or ethernet), many users have grown accustomed to switching between fixed/mobile networks to cater to their needs. Considering MEC’s promise of accelerating these services and alleviating other sections of the network, leaving fixed networks without edge computing capabilities would ostracize this massive and untapped market.
Broadening the specifications and enabling large fixed-mobile network operators -like Verizon, AT&T- to maximize cost-effectiveness and will go a long way towards a future-proof and efficient network.
MEC and 5G will accelerate MNO’s evolution into Fixed-Mobile Network Convergence
Due to the coverage limitations of high-speed millimeter-wavelength technology 5G RAN’s will require distributed approach, what this essentially means is that 5G’s distributed small cell architecture goes hand-in-hand with many of the same concepts employed by MEC. As large mobile & fixed network operators work to optimize their networks and leverage benefits from datacenter economics, their network infrastructure deployments will eventually converge to maximize MEC’s effectiveness.
All of these things sound great in theory, but a proper implementation requires a complete rethink on how a fixed and mobile network should be architected and orchestrated. As fate would have it, the largest telecoms in the US have banded together with many open source MEC, NFV and SDN initiatives in an attempt to accomplish just that.
Designing a future-ready edge computing network from the ground-up
Efficiently shifting computing and storage resources from the core to the edge will require extensive use of not only NFV (Network Functions Virtualization), but Software-defined networking as well. CORD and specifically M-CORD (Mobile Central Office Rearchitected as a Datacenter) are being worked on closely by the open source community and telecom operators Verizon, AT&T, SK-Telecom and several others to create next-generation networking infrastructure that is built from the ground-up with the fundamentals of SDN, NFV and MEC taken to heart.
The m-CORD project brings the many benefits of datacenter economics, the agility flexibility of cloud computing, and leverages the speed of MEC by pushing everything much closer to the edge – the network operator’s central office. Not only does it bring all of the aforementioned benefits in efficiency, cost-reduction and capability, it also eases orchestration in a centralized location using SDN controllers. This opens up many MEC related applications at a reduced cost for network operators, as only few extreme applications require >10ms latency. Deploying such an elastic software-defined network is a major step forward towards fulfilling the ~1ms at the the very edge of the network.
Multi-Access Edge Computing key applications
Machine 2 Machine Communications: One of the most latency sensitive applications, MEC will eventually enable driverless vehicles to use localized decision making in real-time. In healthcare time is a most critical asset that can mean the difference between life and death, as we rely more and more on machines for patient care MEC will help meet the stringent requirements.
Network Security: as the resources are stored at the edge and only sent to the cloud as needed, the possibility of attacks through the network can be significantly reduced if properly implemented.
Caching: heavy static content like videos and images can be cached at the edge, and even dynamic applications can be served from the edge, creating an environment that maximizes efficiency in latency, capacity and efficiency.
It’s becoming clear that SDN, NFV and MEC are the way forward in the next upgrade to our networks, and as titans in the industry get closer to finishing the technologies the first next-generation network will start to appear.