Deep learning based offloading for mobile augmented reality application in 6G. (October 2021)

• Record Type: Journal Article
• Authors: Chakrabarti, Koyela
• Is Part Of: Computers & electrical engineering. Volume 95(2021)
• Publication Date: 2021-10
• Subjects: MAR -- DRL -- 6G -- transmission latency -- offloading -- Frame Processing Rate (FPR); Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2021.107381
• Languages: English
• ISSNs: 0045-7906

Abstract

Highlights[1] : Mobile Augmented Reality (MAR) applications will increasingly be integrated in time constrained fields like healthcare, military, education. But for the given network conditions suffer from delay in rendering output. MAR is typically characterised as a processing and data intensive application, requiring to process high quality video data at around 60 frames per second on average. With the ultra-high bandwidth provided by 6G, MAR applications can be processed with low turnaround time making it suitable for using in delay-sensitive mission critical applications. Taking advantage of increased network speed, Deep Reinforcement Learning (DRL) can be used to divide and distribute the processing tasks of an MAR application to nearby devices. Abstract: Mobile Augmented Reality (MAR) applications are fast becoming popular with the growth in use of smartphones and smart wearable devices. Apart from gaming, MAR finds useful application in any field for attractive visualization of the environment. The computer vision algorithms used in MAR applications are both data and computation extensive which renders them difficult to use in delay sensitive applications, given the present network scenario. But the network standard 6G expected to be deployed around 2030 is supposed to operate at a GHz to THz frequency. This will increase the bandwidth of the network in manifolds and can support the seamless real time transfer of the multimedia data. The article proposes to divide the various phases of an MAR application into sequential and parallel tasks and attempts to offload the task to the nearby devices with the help of Deep Reinforcement Algorithm (DRL) depending on transmission, task and energy constraints. Graphical abstract: Image, graphical abstract