Research in multimedia with computer network mainly involved streaming services. Streaming services are gaining popularity every day. Few examples are YouTube, Netflix, Skype. There are three types of online streaming services: 1) On-demand (YouTube), ii) Live (YouTube Live) and iii) Interactive (Skype). These three streaming systems have a different kind of problems. Research problem related to video streaming system can be categorized in two ways in computer networking viewpoint. They are i) media delivery system/protocol ii) traffic pattern analysis. In our research group (CNeRG), Currently, there are few projects going on. These projects are:
- Underlying Traffic pattern analysis (lead by Satadal Sengupta):
Description: Traffic pattern analysis involves identifying the origin of traffic. When different types of multimedia traffic are going through a single network link, it may require identifying critical traffic to give priority. Most of these video applications use HTTP/HTTPS tunneling making it challenging to apply port based or packet data based identification of flows. This makes it challenging for a network administrator to enforce application-specific prioritization.In this project, we try to identify the flow based on traffic pattern.
- Candid with YouTube: Adaptive Streaming Behavior and Implications on Data Consumption. Abhijit Mondal, Satadal Sengupta, B.R. Reddy, M.J.V. Koundinya, Chander G., Pradipta De, Niloy Ganguly, Sandip Chakraborty. NOSSDAV’17 – ACM SIGMM Workshop on Network and Operating Systems Support for Digital Audio and Video
- MoViDiff: Enabling Service Differentiation for Mobile Video Apps. Satadal Sengupta, Vinay Kumar Yadav, Yash Saraf, Harshit Gupta, Niloy Ganguly, Sandip Chakraborty, Pradipta De. IM’17 – 15th IFIP/IEEE International Symposium on Integrated Network Management
- Video delivery system (Lead by Abhijit Mondal):
Description: In this project, we analyse efficient way to deliver video content to the user. Currently, almost all video service provider use Dynamic Adaptive Streaming over HTTP (DASH) to deliver video to the user/viewer. DASH is efficient in reducing video stall due to network fluctuation. As demands for online video streaming is increasing, it is the time to think about an alternative to save server and network resource utilization by other means. We are trying to decrease network utilization by involving peer-to-peer (P2P) network and scalable video coding (SVC) in streaming system.
With the help of P2P network, a user can share downloaded video segment to other users in the system. Thus, those users don’t need to download segment from the server.
Scalable video coding provides a mechanism to create an embedded bitstream from which different representations can be extracted by partially decoding the compressed data. The more data is obtained, the higher the quality. In practice, the embedded bitstream features a layered structure that includes a base layer and additional enhancement layers. The base layer corresponds to the data that can be decoded independently of the other layers, providing the lowest supported quality. The key advantage of layered coding is that the embedded bitstreams can be tailored to match the desired bandwidth by just selecting some layers and dropping others without transcoding or re-encoding the content.
- Publications: Not Yet