raniita/videochat-qos

[4 GIT] Proyecto final de laboratorio de contenidos digitales

RTP over UDP implementation for A/V multicast-streaming

Objetive

• Analyze the impact on the network by the A/V streaming. Calculate delay, jitter and losses.
• Implementation of RTP Protocol. Spec of RTP
• Chat-multicast over UDP
• Webcam USB multicast-streaming (jpeg compression)
• Microphone multicast-streaming (encoded with a-law[g711])

Disclaimer: The sound only reproduce if the videochat-client form is focus.

Wireshark RTP decoding

If we capture the traffic of videochat-server with wireshark, we can decode the UDP packets with the RTP spec. In case of JPEG frames, we find a RTP Stream and in other case, the audio streaming, we find too a diferent RTP Stream

Usage of wireshark for RTP statistics

Pics of a example traffic capture!

Analyze

On the client, we implement the delay, jitter and losses for every RTP received. We calculate this values for each 1sg

Pics of client!

Pics of server!

Spec of RTP. Appendix for jitter calc

Analysis of RTP Packet Delay and Loss

File explanation

• ALaw. Teacher files to encode and decode audio with ALaw
• RTPStream. My implementation of RTP spec. Create the RTP header, create the packet and send it over the network.
• videochat-client. Client that connect to the net, receiving the incoming RTP A/V packets and show that to the user (reproduce and display).
• videochat-server. Server to wrap the audio and the video, encapsulate and send over the net.

External references

• WebCamLib + WebCamWrapper => C++ Lib that teacher give to wrap USB cameras
• DirectX DirectSound => To capture the sound
• RTPStream => My implementation of RTP Spec
• ALaw => To encode/decode PCM.