Mobility-Aware Cooperative Service Caching for Mobile Augmented Reality Services in Mobile Edge Computing
Qingyang Fan, Weizhe Zhang, Chen Ling, Rahul Yadav, Desheng Wang, Hui He
Abstract
Mobile edge computing (MEC) plays a significant role in reducing network delay for Mobile Augmented Reality (MAR) services by caching these services close to the User Equipments (UEs). These MAR services collect UEs' network traffic and orientation information, and generate the service results back to UEs. However, the UE's mobility features change network traffic and orientation, negatively impacting MAR services' access frequencies and service preferences. Moreover, the changed access frequencies also influence the workload of cached MAR services, resulting in the uneven workload of edge servers. Therefore, this paper formalizes cooperative service caching based on UEs' location and orientation to optimize network delay and response fairness in MEC environments. To solve the problem, we propose a <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S</b>ervice <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</b>aching strategy based on <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R</b>egional <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">M</b>obility features <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</b>wareness (SCRMA) algorithm, which consists of two stages. Firstly, the <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R</b>egional <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">M</b>obility features <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</b>wareness (RMA) algorithm perceives the user mobility features and service preferences, which provides a prerequisite for determining service caching strategy. Then, a <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S</b>ervice <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</b>aching strategy based on a <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">G</b>enetic <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</b>lgorithm (SCGA) is proposed to optimize network delay and response fairness. The simulation experiment on a real dataset shows that our service caching strategy averagely reduces network delay, fairness factor, and total cost by 11.49%, 33.24%, and 17.86% compared with the existing algorithms, respectively.