2024 : 5 : 5
Farhad Pakdaman

Farhad Pakdaman

Academic rank: Assistant Professor
ORCID:
Education: PhD.
ScopusId:
Faculty: Faculty of Technology and Engineering
Address:
Phone: 011-35302903

Research

Title
SARA-SDN: state aware resource allocation in SDN to improve QoE in cloud gaming
Type
Presentation
Keywords
Software Defined Network (SDN), Game State, Cloud gaming, Delay sensitivity, Quality of Experience (QoE)
Year
2020
Researchers Farhad Pakdaman ، Mehrdad Hosseinnejad Yami ، Mahmoud Reza Hashemi

Abstract

Cloud Gaming (CG) solutions run games on cloud servers and stream the compressed video of the game over a network. While this has many benefits for both users and game companies, CG imposes two major challenges: low delay and high bandwidth are essential to ensure QoE. While the tolerable delay varies based on state of the game, previous researches do not fully exploit this concept to enhance QoE. This paper demonstrates the importance of the game-state as a criterion for game requirements. Accordingly, a game-State Aware Resource Allocation for Software-Defined Network (SDN)-based cloud gaming (SARA-SDN) is presented, which incorporates the game state into decision making. First, an SDN architecture is designed to facilitate communication of game information and control commands between game-server, SDN controller, and network switches. Second, a resource allocation algorithm is proposed that dynamically considers game-specific requirements such as delay, bandwidth, and game-state. Finally, SARA-SDN is implemented and evaluated in a simulated environment, where real users played in real time, and different aspects of QoE was measured via extensive subjective tests. Results show the superiority of SARA-SDN over existing methods (on average 1.26, 1.18, 1.21, and 1.12 increase in QoE, responsiveness, controllability, and immediate feedback, respectively, in a scale of 0-5).