Flexible DRX Optimization for LTE and 5G

Research output: Contribution to journalArticle

Standard

Harvard

APA

CBE

MLA

Vancouver

Author

RIS

TY - JOUR

T1 - Flexible DRX Optimization for LTE and 5G

AU - Moradi, Farnaz

AU - Fitzgerald, Emma

AU - Pioro, Michal

AU - Landfeldt, Björn

PY - 2020

Y1 - 2020

N2 - With the advancement of the next generation of cellular systems, flexible mechanisms for Discontinuous Reception (DRX) are needed in order to save energy. 5G will bring heterogeneous packet sizes and traffic types, as well as an increasing need for energy efficiency. The current static DRX mechanism is inadequate to meet these needs. In this paper we exploit channel prediction to develop integer programming models. We aim to minimize the energy usage of user devices while streaming video, as well as to create extended sleep opportunities, while simultaneously preventing buffer underflows. We also develop an online algorithm to obtain an efficient solution robust to prediction errors. Our results show that using a variable DRX cycle length can reduce the energy usage by up to 60 percent and 40 percent, in the offline and online cases, respectively, compared with a static DRX configuration. Our proposed online algorithm can also reduce the number of buffer underflows by up to 97 percent compared to the offline case. Both our online and offline solutions can provide extended DRX opportunities, which is required in 5G scenarios.

AB - With the advancement of the next generation of cellular systems, flexible mechanisms for Discontinuous Reception (DRX) are needed in order to save energy. 5G will bring heterogeneous packet sizes and traffic types, as well as an increasing need for energy efficiency. The current static DRX mechanism is inadequate to meet these needs. In this paper we exploit channel prediction to develop integer programming models. We aim to minimize the energy usage of user devices while streaming video, as well as to create extended sleep opportunities, while simultaneously preventing buffer underflows. We also develop an online algorithm to obtain an efficient solution robust to prediction errors. Our results show that using a variable DRX cycle length can reduce the energy usage by up to 60 percent and 40 percent, in the offline and online cases, respectively, compared with a static DRX configuration. Our proposed online algorithm can also reduce the number of buffer underflows by up to 97 percent compared to the offline case. Both our online and offline solutions can provide extended DRX opportunities, which is required in 5G scenarios.

KW - 5G mobile communication

KW - channel capacity

KW - streaming media

KW - optimization

KW - prediction algorithms

KW - DRX

KW - Integer programming

KW - resource allocation

KW - video streaming

KW - energy efficiency

U2 - 10.1109/TVT.2019.2952251

DO - 10.1109/TVT.2019.2952251

M3 - Article

VL - 69

SP - 607

EP - 621

JO - IEEE Transactions on Vehicular Technology

JF - IEEE Transactions on Vehicular Technology

SN - 1939-9359

IS - 1

ER -