Research On Wireless Video Transmission Optimization And Its Quality Accessment Methods

In recent years, with rapid development of broadband wireless mobile communication technology, the communication systems are envisioned to provide high bit rate transmission and QoS guarantees. The OFDM is such an advanced communication technique, which has widely adopted in the broadband wireless access technologies, such as WiMAX and LTE. It can provide great flexibility for subcarrier assignment to maximize the system capacity and the spectral efficiency. Thus, the problem of video transmission scheme over downlink OFDMA network is focused on how to design effective video packet scheduling and subcarrier allocation algorithm for improving the end-to-end video quality. However, these methods which are available for data transmission cannot be directly applied to video applications. Therefore, it is necessary to study video transmission optimization schemes which accord with human’s subjective perception.Videos are subject to a wide variety of distortions during transmission, so it’s always need to access its quality. In practice, however, subjective evaluation is usually time-consuming and expensive. The traditional objective methods such as MSE has low correlation with the perceptual visual quality. SSIM has good correlation with perceived video quality, but it has a high computational complexity. It is significant to propose a system that has competitive performance with normal SSIM and much lower computational complexity for assessing video quality.This paper investigates the problem of multiuser video transmission over downlink OFDMA networks. A scheme named quality-maximizing based on MOS is proposed, which is to maximize the video quality without fairness constraint. The simulation results show that the proposed algorithm outperforms dynamic subcarrier allocation and stepwise allocation algorithm.On the other hand, a fast video quality assessment system is proposed. The experimental results show that the FSSIM index algorithm not only has competitive performance with normal SSIM for assessing video quality, but also has much lower computational complexity.