Video streaming has gained its popularity in both academia and industry. With the prevalence of mobile computing, it poses new challenges and opportunities. A video encoder that adapts to the instant channel conditions has the potential to provide better video quality even under hostile transmission environments. The motivation of MCN Streaming is to provide a complete video streaming platform that can ease the experimenting of various adaptive video encoding algorithms over the real-world networks, instead of merely depending on simulations.

To summarize, MCN Streaming is:

• a complete video streaming software, including the encoder/sender and receiver/decoder.
• a platform for experimenting adaptive video coding algorithms.
• flexible and extensible.

Above figure shows the high level overview of the MCN Streaming architecture. MCN Streaming consists of two components: a sender and a receiver. Sender is responsible for video encoding and transport layer packet encoding, while receiver is responsible for transport layer packet decoding and video decoding. Encoded packets travel through the Internet. Receiver also sends feedback to receiver via RTCP packets. The feedback (QoS information) is used by video encoder to adjust its encoding parameters on-the-fly. By doing this, video encoder adapts to the instant channel conditions and better video quality is expected at the receiver side.