Device-Controlled Traffic Steering in Mobile Networks

TY - JOUR

T1 - Device-Controlled Traffic Steering in Mobile Networks

AU - Jiang, Menglan

PY - 2016/1/5

Y1 - 2016/1/5

N2 - User equipments (UEs) such as smartphones and tablets are widely used in heterogeneous wireless networks which include different technologies like WiFi and Long Term Evolution (LTE) systems. Implementing traffic steering effectively for user equipments to maximize their quality of service has become a key challenge these years. Therefore, in this paper, we propose a device controlled mechanism. We mainly focus on history records of each available networks and distributively consider multi-criteria like Received Signal Strength (RSS) and energy consumption of battery as our performance matrics from UE side. Device controlled mechanism is a fully distributed traffic steering approach that runs at user equipments. At the same time, to benefit from the network knowledge, we propose to use network analytic mechanism to further enhance certain performance metrics. Our proposed approach and analytic mechanism can help user equipments to make efficient decisions and minimize battery power consumption. This is particularly important to distribute traffic load across different radio access with less energy consumption for user equipments. The traffic steering of radio accesses in this paper is modelled based on Reinforcement Learning (RL) mechanism which considers past experiences of user equipements and help them to improve their quality of satisfactions. Through extensive simulation scenarios, we demonstrate how such device controlled mechanism with multi-criteria metrics can improve received throughput values of user equipments and reduce energy consumption of equipments' batteries.

AB - User equipments (UEs) such as smartphones and tablets are widely used in heterogeneous wireless networks which include different technologies like WiFi and Long Term Evolution (LTE) systems. Implementing traffic steering effectively for user equipments to maximize their quality of service has become a key challenge these years. Therefore, in this paper, we propose a device controlled mechanism. We mainly focus on history records of each available networks and distributively consider multi-criteria like Received Signal Strength (RSS) and energy consumption of battery as our performance matrics from UE side. Device controlled mechanism is a fully distributed traffic steering approach that runs at user equipments. At the same time, to benefit from the network knowledge, we propose to use network analytic mechanism to further enhance certain performance metrics. Our proposed approach and analytic mechanism can help user equipments to make efficient decisions and minimize battery power consumption. This is particularly important to distribute traffic load across different radio access with less energy consumption for user equipments. The traffic steering of radio accesses in this paper is modelled based on Reinforcement Learning (RL) mechanism which considers past experiences of user equipements and help them to improve their quality of satisfactions. Through extensive simulation scenarios, we demonstrate how such device controlled mechanism with multi-criteria metrics can improve received throughput values of user equipments and reduce energy consumption of equipments' batteries.

KW - battery lifetime

KW - device-centric network assisted selection algorithm

KW - heterogeneous mobile networks

KW - Received Signal Strength

UR - http://www.scopus.com/inward/record.url?scp=84964584463&partnerID=8YFLogxK

U2 - 10.1109/NGMAST.2015.13

DO - 10.1109/NGMAST.2015.13

M3 - Conference paper

AN - SCOPUS:84964584463

SP - 7

EP - 12

JO - Next Generation Mobile Applications, Services and Technologies, 2015 9th International Conference on 9-11 Sept. 2015

JF - Next Generation Mobile Applications, Services and Technologies, 2015 9th International Conference on 9-11 Sept. 2015

T2 - 9th International Conference on Next Generation Mobile Applications, Services and Technologies, NGMAST 2015

Y2 - 9 September 2015 through 11 September 2015

ER -