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

Jump to navigation Jump to search

Seminar on Internet Technologies (Winter 2018/2019) (view source)

Revision as of 19:04, 23 September 2018

7,069 bytes removed ,  23 September 2018
→‎Topics
Line 50: Line 50:
|{{Hl2}} |'''Initial Readings'''
|{{Hl2}} |'''Initial Readings'''
|-
|-
| '''Reinforcement Mechanism Design''' 
| '''The application of Recurrent neural network in Human mobility modeling (assigned to ?)'''
| Mechanism design is a modeling and algorithmic framework  to  design  and  optimize  mechanisms  in  dynamic  industrial  environments  where  a  designer can  make  use  of  the  data  generated  in  the  process to automatically improve future design.  Reinforcement mechanism design is  rooted  in  game  theory  but  incorporates  recent AI  techniques  to  get  rid  of  nonrealistic modeling assumptions and to make automated optimization feasible.  The framework can be applied on many key application scenarios, such as Baidu and Taobao, two of the largest mobile app companies in China. For the Taobao case, the framework automatically designs mechanisms that allocate buyer impressions for the e-commerce website; for the Baidu case, the frame-work automatically designs dynamic reserve pricing schemes of advertisement auctions of the search engine. Experiments show that the solutions outperform the state-of-the-art alternatives and those currently deployed, under both scenarios.
| RNN/LSTM is one of the most widely used deep learning method, especially good at sequence modeling. Human mobility is a typical class of sequence data. So until now, what and how is RNN working in human mobility area, especially in online check-ins? Understanding this question will help you know the basic concept, general problems and important approahces in this field.
| Basic knowledge of machine learning, deep learning and big data analysis. Familar with mechanism design.
|Bo Zhao (bo.zhao@gwdg.de)
|[https://pdfs.semanticscholar.org/5ec9/ed5c9936c069fa554603aa773d3ee036b4ac.pdf][https://arxiv.org/abs/1708.07607][http://iiis.tsinghua.edu.cn/~kenshin/rmd_ec.pdf]
|-
| '''Understanding and implementing web crawler'''
| Web crawlers collect information such the URL of the website, the meta tag information, the Web page content, the links in the webpage and the destinations leading from those links, the web page title and any other relevant information. Web crawler can be used in data mining, wherein pages are analyzed for different properties like statistics, and data analytics are then performed on them. The topic has two main tasks: 1) explore and summarize existing python-based web crawling frameworks; 2) design and implement a simple web crawler with Python, e.g., crawl some information of a simple webpage in Twitter.
| The student interested in this topic should be familiar with Python, and ideally has some practical experience with Python.
| [http://www.net.informatik.uni-goettingen.de/people/tao_zhao Tao Zhao]
| [https://en.wikipedia.org/wiki/Web_crawler]
|-
| '''Getting a Practical Understanding of Segment Routing'''
| Segment Routing (SR) is a new data-plane paradigm that employs source routing and software defined networking (SDN) to present an ease to manage network. The main key advantage compared to openflow based SDN is that SR embeds the path in the packet header and hence no states are kept in the core network. Your task is to understand in practical and theory how this main key affects the way SR is doing relative network tasks such as traffic enigneering and service function chaining. If you are not sure about your time schedule during this semester, please choose another topic.
| The student should be at least familiar with one programming language (eg. Java or Python), basic open-flow SDN and basic linux skills.
| [http://www.net.informatik.uni-goettingen.de/people/osamah_barakat Osamah Barakat]
| [http://www.segment-routing.net/][http://www.segment-routing.org/]
|-
| '''Hands-on study of Vector Packet Processing (VPP)'''
| The VPP platform is an extensible framework that provides out-of-the-box production quality switch/router functionality. It is the open source version of Cisco's Vector Packet Processing (VPP) technology: a high performance, packet-processing stack that can run on commodity CPUs. The benefits of this implementation of VPP are its high performance, proven technology, its modularity and flexibility, and rich feature set.
The student's task will be acquiring hands-on experience with the VPP platform (included source code). The expected outcome would be a detailed report on how to use VPP and how to create a custom plug-in (for this part a demo application should be implemented).
 
| Basic networking knowledge, C/C++ programming, Unix/Linux administration. Knowledge of Data Plane Development Kit (DPDK) would be beneficial (but not mandatory).
| [http://www.net.informatik.uni-goettingen.de/people/jacopo_de%20benedetto Jacopo De Benedetto]
| [https://wiki.fd.io/view/VPP]
|-
 
| '''Deep Learning and Smart City (assigned to Jihan Munkar)'''
| Smart city is a hot topic in recent years. And deep learning is another hot topic. However, the application of DL in smart city areas is largely overlooked by the reserach community. The student picking this topic need to read several recent papers about connecting DL to smart city questions. In this process, you will know basic concept, general problems and important approahces in this field.
| Basic machine learning knowledge
| Basic machine learning knowledge
| [Shichang Ding--shichang.ding@informatik.uni-goettingen.de]
| [Shichang Ding--sding@gwdg.de]
| [https://aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14501]
| [https://www.ijcai.org/proceedings/2018/0477.pdf]
|-
 
| ''' Research and implementation of an OPC UA application in ICN'''
| THE OPC foundation is a consortium of industry partners that is responsible for creating and maintaining industry standards. There most recent standard is called Open Platform Communications Unified Architecture (OPC UA). OPC UA brings a significant enhancement to the existing OPC framework, especially making is platform independent, and turning it into a service oriented architecture. It is an open source architecture and in this topic you will be required to perform a research on the OPC UA standard and the communication protocols offered in this architecture and implement an application based on this architecture in Information Centric Networking (ICN).
 
| Basic networking knowledge, C/C++ programming, Unix/Linux, Java and Information Centric Networking (ICN).
| Sripriya Adhatarao (adhatarao@cs.uni-goettingen.de)
| [https://opcfoundation.org/]
|-
 
|'''NFV Frameworks - Insight and Assessment'''
| Focus of this topic is to present a comprehensive study and evaluate at least 2 of the NFV platforms (either the Industry and academic works targeted towards deployment of NFV in Telecommunications, Data Center and Enterprise networks). Understand and Analyze the key aspects of the predominant NFV frameworks, and characterize them in terms of the adopted standards, resource requirements, deployment factors and constraints, performance metrics, support for service function chaining, etc. for deployment of Middleboxes and Network Functions in Telco/ISP/Data Center Networks. Evaluate the key deployment metrics accounting both qualitative and quantitative aspects in deploying and building the NFV solutions.
| Familiarity with SDN, NFV and DPDK (students who have already taken SDN course are preferred). Must have Basic networking knowledge, C/C++ programming, Linux administration.
| [http://www.net.informatik.uni-goettingen.de/people/sameer_kulkarni Sameer Kulkarni]
| [https://www.opnfv.org/] [https://www.usenix.org/system/files/conference/nsdi14/nsdi14-paper-martins.pdf] [http://dl.acm.org/ft_gateway.cfm?id=2940155&ftid=1754642&dwn=1&CFID=919487200&CFTOKEN=31286219] [http://superfluidity.eu/about/research-description] [https://fd.io/]
|-
|-
| '''Online Pricing Mechanisms for Cloud and Edge Computing'''
| The goal is to present at least two online pricing schemes in cloud computing. Currently, cloud resources are billed hourly basis. These online pricing schemes charge customers based on supply demand at current time instance.
| Familiarity with basic optimization (linear programming) and probability.
| [https://www.net.informatik.uni-goettingen.de/people/abhinandan_s_prasad Abhinandan S Prasad]
|[https://dl.acm.org/citation.cfm?id=3086574] [https://dl.acm.org/citation.cfm?id=3084460]
|-
|}
|}


Sding
308

edits

Retrieved from "https://wiki.net.informatik.uni-goettingen.de/wiki/Special:MobileDiff/5563"

Navigation menu