Research On WLAN Streaming Media System Based On Application-Level FEC

In recent years, with the rapid development of wireless technology and mobile internet devices like smartphones and Pads, wireless streaming media technology has been widely used. In the mobile Internet era demand for wireless streaming media has surged dramatically. To improve the reliability and stability of the streaming media transmission without excessive occupancy of wireless bandwidth, it is essential to research on ensuring the Qo S of streaming media in a complex wireless network.The number of redundant packet in traditional application-level FEC algorithm is often determined beforehand. When the network packet loss rate is low, redundant packet consumes too many bandwidth resources. As the bandwidth resource is quite limited it would be easy to cause streaming delay jitter. When the network packet loss rate is too high and the transmission network environment has become increasingly severe, fixed redundant packet is not enough to restore the right source, which would be easy to cause the phenomenon of streaming media pause. In the thesis, two kinds of adaptive FEC algorithm are studied and an adaptive FEC algorithm based on Kalman Filter is proposed. According to the posterior estimation of FEC redundancy packet, the adaptive FEC algorithm based on Kalman Filter we bring forward makes more accurate estimation and adjustment of the FEC coding redundancy. In a simple WLAN transmission simulation environment, the adaptive FEC algorithm based on Kalman Filter keep the balance of guaranteeing the streaming media quality effectively and saving the wireless bandwidth resources. Then we implement a WLAN streaming media simulation system based on my Evalvid framework and compare the three kinds of adaptive FEC algorithm in a complex WLAN transmission simulation environment. The result shows that adaptive FEC algorithm based on Kalman Filter is slightly better than the other two kinds of adaptive FEC algorithm in decoding frame rate and PSNR performance.At last, a wireless streaming media server based on S3C6410 processor is implemented, and an adaptive FEC decode module is added to a streaming media client based on FFmpeg. The adaptive FEC algorithm is tested and evaluated in a real WLAN test environment. The result shows that adaptive FEC algorithm based on Kalman Filter has achieved good effect in real-time transmission and image display. The packet loss rate and jitter delayed are controlled effectively within a reasonable scope and the guaranteed Qo S of WLAN streaming media transmission is provided.