Seminar on Internet Technologies (Winter 2018/2019): Difference between revisions

no edit summary
No edit summary
No edit summary
Line 85: Line 85:
| [https://arxiv.org/pdf/1711.04710.pdf]
| [https://arxiv.org/pdf/1711.04710.pdf]
|-
|-
| ''' Mobile spatial and temporal data based user profiles identification'''
| ''' HTTP/2 Push-Based Streaming Over Mobile Networks With Stream Termination'''
| When people using their APPs in smartphones, the communication base station will catch the URL requests from each customer. Each record is assigned to a specific user ID, and includes a time stamp and an URL request. Moreover, most frequently used APPs need location service provided by assembled GPS sensor in the smartphone, and some coordinates are contained in the URLs. With the information above, each user could be modeled by a set of spatial-temporal data. Different users could show different patterns, which includes the features like POI preference, trajectory type, usage frequency at different time and etc. We aim to cluster the users in different categories according to their spatial-temporal data.
| HTTP adaptive streaming (HAS) has become popular for delivering videos over the Internet. However, most HAS services are currently based on the traditional HTTP/1.1, which has limitations for live streaming. Meanwhile, the new HTTP/2 was standardized in 2015 with many improvements. Therefore, a novel HTTP/2-based adaptation scheme for low-delay live streaming over mobile networks is required in nowadays.


| Basic data science knowledge, Python programming and data science related libraries.  
| Basic networking knowledge, Python, C/C++ or other programming skills.  
| Jiaquan Zhang (jzhang@cs.uni-goettingen.de)
| Yali Yuan (yali.yuan@informatik.uni-goettingen.de)
| [https://arxiv.org/pdf/1711.04710.pdf]
| [https://ieeexplore.ieee.org/abstract/document/8396293]
|}
|}