Browsing by Author "Mongi, Alex Frank"

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Analyzing the influence of smart-device visual features, viewing distance, and content factors on video streaming QoE
    (AJOL, 2022) Mongi, Alex Frank
    Quality of experience (QoE) over wireless networks has attracted attention from industry and academia due to an increase in video streaming applications. Several researchers have attempted to understand the factors affecting QoE and design appropriate quality control strategies. Normally, video streaming is initiated by a user who accesses video contents over wireless networks using a smart device held at various viewing distances. Each aforementioned factor has the potential to affect QoE of the viewed session. However, several studies explore the behavior of wireless networks on video streaming QoE. To understand the effects of other factors on QoE, this paper investigates the influence of the device's visual features, viewing distance, and content factors on video streaming. The study adopted an emulation technique to conduct multi-factor experiments designed using the Taguchi method. The 5-ways ANOVA analysis revealed that the effects of smart-device visual features, viewing distance, and content types are significant on video streaming QoE at p<0.05. Moreover, smart devices with a pixel density index of more than 200 ppi produce high QoE, with the viewing distance limited to 45 cm. Lastly, the video bitrate greater than 1024 kbps produced a good QoE regardless of the frame rates
  • Thumbnail Image
    Item
    Developing video streaming quality of experience prediction model in wireless broadband networks
    (The University of Dodoma, 2017) Mongi, Alex Frank
    Mobile video streaming has grown tremendously in recent years due to technological advancement of both network and end-devices. Unlike the voice services which run on the background traffic mode, the network impairments on video streaming traffic affect users’ perceived quality directly which results into high rate of churn. For that reason, measuring users’ quality of experience (QoE) has become the major challenge to network operators and service providers. This study therefore, aimed at developing a mathematical model for predicting video streaming QoE in wireless broadband networks. The systematic literature review and survey methods were used to identify the variables affecting user QoE through frequency counts and correlation analysis. Furthermore, experiments were conducted using emulation technique over a wireless broadband network test-bed to investigate the effects of identified variables on video streaming QoE. The data extracted were analysed by using the Taguchi method, 3-way Anova and 5-way Anova. This study found that, pixel density index and viewing distance of smart devices’ screen had significant effect on video streaming QoE. Moreover, bit rate, frame rate and content type significantly affected video streaming QoE. The network impairments due to delay and jitter were also found to be more destructive on video streaming QoE compared to packet loss. Eventually, the nonlinear mathematical model was developed by basing on the combined effects of content type, bit rate, delay, jitter and pixel density index. It achieved high level of prediction accuracy in three content types grouped into slow, medium and fast moving contents.
  • Thumbnail Image
    Item
    Modeling video streaming quality of experience using Taguchi and fuzzy logic methods
    (The University of Dar Es Salaam, 2023) Mongi, Alex Frank
    The popularity of mobile video streaming has increased significantly in recent years, and is expected to account for two-thirds of global internet traffic in the near future. However, determining accurately end-users' satisfaction based on network parameters remains a challenge. Existing research often uses network parameters, such as packet loss, delay, and jitter, to estimate users' Quality of Experience (QoE). However, most models present QoE estimates in Mean Opinion Scores (MoS), which are not easily understood by the customers. In this study, we used the Taguchi approach to conduct QoE experiments over a wireless tested. We investigated the simultaneous effects of packet loss, corruption, delay, and jitter on video streaming QoE, as well as their interaction effects. Furthermore, we developed a Fuzzy logic model in MatlabR2016a to establish the relationship between input variables and video streaming QoE. The model presents the results in an easily understandable linguistic terms such as excellent, good, average, bad, and poor. Additionally, the proposed model achieves a correlation of 0.875 between the predicted and user scores, with a Root Mean Square Error of 0.344