A cloud native solution for dynamic auto scaling of MME in LTE

Abstract

Due to rapid growth in the use of mobile devices and as a vital carrier of IoT traffic, mobile networks need to undergo infrastructure wide revisions to meet explosive traffic demand. In addition to data traffic, there has been a significant rise in the control signaling overhead due to dense deployment of small cells and IoT devices. Adoption of technologies like cloud computing, Software Defined Networking (SDN) and Network Functions Virtualization (NFV) is impressively successful in mitigating the existing challenges and driving the path towards 5G evolution. However, issues pertaining to scalability, ease of use, service resiliency, and high availability need considerable study for successful roll out of production grade 5G solutions in cloud. In this work, we propose a scalable Cloud Native Solution for Mobility Management Entity (CNS-MME) of mobile core in a production data center based on micro service architecture. The micro services are lightweight MME functionalities, in contrast to monolithic MME in Long Term Evolution (LTE). The proposed architecture is highly available and supports auto-scaling to dynamically scale-up and scale-down required micro services for load balancing. The performance of proposed CNS-MME architecture is evaluated against monolithic MME in terms of scalability, auto scaling of the service, resource utilization of MME, and efficient load balancing features. We observed that, compared to monolithic MME architecture, CNS-MME provides 7% higher MME throughput and also reduces the processing resource consumption by 26%.

SDN-Based Network Slicing Mechanism for a Scalable 4G/5G Core Network: A Kubernetes Approach

“…In [ 18 ], a solution to migrate the MME component into a cloud-native one, in order to implement dynamically auto-scaling, was proposed. The motivation for the paper was to minimize the overhead due to control signaling from IoT applications implemented over 5G networks.…”

Nervion

“…The core is a key mobile network component that handles the essential control functionality such as mobility, session and security management, and also bridges data communication between end devices and external networks like the Internet via the radio access network (RAN). Driven by the limitations of the traditional mobile network architecture and the wider trends in the mobile network industry, such as embracing network softwarization and supporting diverse use cases, mobile core networks have received a great deal of attention from the standardization bodies [1][2][3][4] as well as research [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21] and open source [22][23][24][25][26] communities.…”

“…RAN emulation has become the commonly used mobile core system evaluation methodology, as reflected by [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21], compared to other alternatives like testbed-based experimental evaluation, simulation and mathematical analysis. The general idea behind RAN emulation is to emulate the RAN elements -base stations and end devices (UEs) -from the perspective of the core system under test.…”

SABER: An Infrastructure-Agnostic Benchmark Tool for Elasticity Evaluation on Cloud Foundry Based Platforms

“…Cloud native is a new approach that has emerged as a solution to extract the maximum potential from cloud computing [7] [8] [9]. Contrary to a monolithic application, it is decomposed into smaller modules called microservices, and each microservice runs in a container in order to optimize resource utilization [10] [11]. Cloud-native applications are designed to run on cloud-native platforms, which runs on top of a cloud computing infrastructure.…”