August 6, 2020 – With 5G technologies driving the advancement of low-latency and high-capacity network applications, the development of decentralized MEC (Multi-access Edge Computing or Mobile Edge Computing) has become a future trend. Striving to optimize the utilization of sizable edge computing resources, meanwhile addressing the more dispersed and stringent deployment environment, Lanner and Taiwan National Chiao-Tung University (NCTU) joint-handedly, announced the research outcomes of “High scalable multi-access edge computing structure” and “Next-gen Fanless Cooling System” projects. Five years ago, Lanner established its Telecommunications Department, dedicated to research and development on SD-WAN/5G relevant edge computing applications. Through the industry-academia-research collaboration with NCTU, Lanner has become the network appliance supplier to Verizon’s uCPE 2.0 platform in 2019 and continues to enhance R&D competencies in edge computing technologies in order to stay both in line with international norms and on top of the latest 5G specifications.
A Scalable and Flexible Automated Deployment Edge Computing Platform Based on Open Architecture
Edge computing has given birth to diverse low-latency 5G applications, such as the autonomous vehicles, the Augmented Reality (AR) and Virtual Reality (VR). The implementation of a massive edge computing platform makes it possible to distribute the services over various locations including the Small Cells (this term encompasses Picocells, Microcells and Femtocells), the Radio Access Network (EAN) and the Core Network (CN). Hence comes the mutual focus of future development for all telecom operators – How to achieve the optimal allocation of distributed edge computing resources without having any impact on user experience. The cooperative research of over seven months between Lanner and NCTU eventually carried out the Phase I proof of concept (PoC) for a scalable MEC platform. This implementation employed Lanner’s telecom network appliance HTCA-6200 as the hardware solution, running the open software platforms OpenStack and Intel OpenNESS respectively to create a simulated scenario where the number of clients, computing requirements and throughput all surge concurrently, and succeeded in auto-scaling the requests to those idle MEC platforms including Small Cells, Access Networks (AN) and Core Networks (CN). Not only can this implementation equip the entire 5G network architecture with flexibility and scalability, the CapEx (Capital expenditure) and OpEx (Operational expenditure) are also substantially lowered.
The Brand-new High-Performance Cooling Chassis Designed for Fanless Environment
As the key of ultra-low 5G latency application lies in Small Cell base station implementation, dissipating the unwanted by-product of high-performance computing (HPC) – heat, in critical conditions, has become the primary issue. The next-gen cooling fin design built for fanless edge computing systems – the eight-month collaborative work of Lanner’s mechanical engineering team and NCTU, has gone through repeated tests and been proved to create sufficient extra paths for airflow, drawing in more air to pass through the fins and consequently effectively enhancing the overall heat dissipation performance by resolving the thermal convection issues accompanying high-perforce processors.
Lanner Joint Forces with Partners in North America to Build Programmable MEC Solutions
Through participating in the latest 5G conformance tests and collaborating with telecom operators/software partners, Lanner research team, debuted at the beginning of March its programmable MEC solution; this solution integrates Lanner MEC server HTCA-6600, which carries up to twelve Intel® Xeon®scalable compute modules, P4-programmable Barefoot Tofino Switch modules, NoviFlow SDN collaborative management platform (Orchestration), and Fortinet Security Virtual Network Functions (VNF). The programmable MEC solution offers an optimized highly-integrated hardware platform for 5G applications, streamlining the data flow, throughput and computing resources while considerably reducing latency. With this model going on the market in Q3 2020, Lanner has forged new bonding with telecom equipment suppliers/operators in North America and geared up for the future 5G massive distributed edge computing applications.
High Availability Chassis 2U Telecom Network Appliance with 2 x86 CPU Blades and 2 I/O Blades
|CPU||Intel® Xeon® processor E5-2600 v3/v4 Series|
|Chipset||Intel C612 Chipset|