Article Preview
Top1. Introduction
Since the inception in early 1990, video streaming applications have experienced explosive growth and transformations. Video streaming via Online is one such transformation and a novel application experienced by the people today (Kim & Chung, 2019; Lusilao Zodi et al., 2019; Song et al., 2018). Streaming is the continuous transmission of audio or video files from a server to a client. In simpler terms, streaming is what happens when consumers watch TV or listen to podcasts on Internet-connected devices (Park et al., 2019; Wang et al., 2018). Currently, consumer video is dominated by High Definition (HD), but higher resolutions, such as 4K are gaining mainstream popularity, with up to 10% market penetration in the US alone. Furthermore, there are an increasing number of video streaming devices and platforms being added globally every day (Gatimu et al., 2020). Streaming content providers, such as Netflix, Hulu, and YouTube, strive to offer high-quality video content that is viewed by millions of subscribers under very diverse circumstances, using a plethora of devices, under varying viewing resolutions and network conditions (Christos Bampis & Li, 2018; Wahab et al., 2020). An important point regarding the last mile of the end-to-end communication path is the heterogeneity and variability of the elements composing it, e.g., different available radio-transmission technologies, state of the access network, end-user equipment, etc.
Therefore, video transport algorithms must permit their adjustment in realtime to this wide variety of aspects that heavily determine the quality perceived by the end-user quality of user experience (QoE) (Bermudez et al., 2019; Vega et al., 2018). Providing satisfactory QoE to all clients by handling such huge volumes of data over a limited available bandwidth, is a very challenging task to be exclusively conducted in the network, irrespective of the radio-resource allocation scheme adopted (Lekharu et al., 2020; Qiu, 2018; Shabrina et al., 2020). Hence, a client-side bit-rate prediction mechanism is required, which will be able to provide customized bit-rate adaptation decision of client by taking into account client-specific factors, such as the size of video segments received in the past, current play-out buffer status, and history of actually received bandwidth (Al-Issa et al., 2019; Bentaleb et al., 2018).
However, it is difficult for the user to accurately estimate available network bandwidth because the users cannot specify whether a downloaded segment is obtained from a content server or an intermediate router (Arya & Sharma, 2018; Paul et al., 2016; Sharma, 2018). So, the main challenge in many-to-one video streaming is that the wireless network has limited available bandwidth, which changes dynamically and should be estimated accurately and distributed efficiently among various video sources to maximize the quality of the video streams received by the proxy station (Mohammad Alsmirat & Nabil, 2018). The most previous approaches based on the estimation of available link bandwidth or fullness of media buffer tends to become ineffective due to the variability of IP traffic patterns (Kumar, 2018). But, it is not possible to see how the video quality changes throughout a streaming session with the average viewport peak signal-to-noise ratio (PSNR) (Duc Nguyen & Huyen, 2019; Venkatesh & Nithiyanandam, 2019). Therefore, this paper proposed an efficient technique to estimate the available bandwidth for effective video streaming services.
The remainder of this paper is organized as follows: Section II provides state-of-art of the bandwidth estimation for video streaming services. In section III, the system discusses the proposed methodology. In section IV, the system describes the evaluation and its results. Finally, the system summarizes the conclusions and future scope in section V.