Low Cost MEC Server Placement and Association in 5G Networks

With high demands for delay-sensitive services such as ultra low latency communication (URLLC), newly developed 5G networks support mobile edge computing (MEC), in which service servers are located nearby its users to minimize backhaul delay. In this paper, we study MEC server placement and associat...

Full description

Saved in:
Bibliographic Details
Published in2019 International Conference on Information and Communication Technology Convergence (ICTC) pp. 879 - 882
Main Authors Lee, Seungyeol, Lee, Soohwan, Shin, Myung-Ki
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.10.2019
Subjects
5G cellular networks
5G mobile communication
Clustering algorithms
Delays
Edge server placement
MEC
Propagation delay
Servers
Telecommunications
Ultra reliable low latency communication
URLLC
Online AccessGet full text

Cover

More Information
Summary:With high demands for delay-sensitive services such as ultra low latency communication (URLLC), newly developed 5G networks support mobile edge computing (MEC), in which service servers are located nearby its users to minimize backhaul delay. In this paper, we study MEC server placement and association algorithm that promises a certain end-to-end service latency with the minimum MEC servers to be placed. We model this as a capacitated clustering problem which determines the minimum number of clusters and the associated elements for each cluster within given constraint. Since, however, capacitated clustering problem is well known as an NP-Complete problem, we propose a practical algorithm, called LowMEP, with some heuristics to tackle such challenges. Furthermore, we demonstrate that LowMEP requires the minimum number of MEC servers to assure a certain latency budget through numerical simulations based on a real dataset provided by a telecom operator.
DOI:10.1109/ICTC46691.2019.8939566