Periodic Reporting for period 3 - ICN2020 (ICN2020: Advancing ICN towards real-world deployment through research, innovative applications, and global scale experimentation)
Période du rapport: 2018-07-01 au 2019-06-30
Information access on Internet is exploding, primary use has shifted to multimedia, social networking and Internet of Everything (IoE). Cellular networks are constantly advancing to meet current needs of users and are moving to next generation. Networking technology is also shifting towards virtualization (SDN, NFV). This will have a major impact in changing infrastructure landscape. The cloud is also transforming the Internet into a network of data centers, computer-to-cloud-to-computer communication models. Security concerns are increasing and are leading to encryption of all traffic, wreaking havoc on network. With such dramatic growth and evolution, where abstractions and interfaces become fundamental, Information Centric Networking (ICN) is the perfect solution. In a time when 5G networks are being designed, with foreseen unprecedented flexibility due to virtualization and slicing, development of compelling demonstrations of ICN for real-world use-cases will encourage critical industry investment of resources. ICN2020 project with its ambitious and practical goal is aligned to provide the best technology and value-added services to the citizens of Europe and Japan. The ICN2020 project will help bridge the gap between day-to-day services and exceptions by making use of ICN features to enhance 5G and other future networking architectures and provide a very critical service to the society. The key objective of ICN2020 is to facilitate real-world deployment of ICN through Research, Innovation and large scale experiments.
In Y1, we focused our effort to establish global-scale ICN testbeds specifically for large deployments. For local testbeds, first, CUTEi, a novel framework to build unified testbeds for ICN with linux containers to share testbed’s hardware resources among different users is deployed, while the second tool vICN is designed as general-purpose management and orchestration tool for ICN testbed and is part of CICN project. The experience in deploying CUTEi and vICN is reported. Experience with NDN testbed is reported & some of its limitations are shown. GEÁNT Testbed Service (GTS) is proposed to both create an independent and isolated global-scale ICN testbed or to extend functionalities of existing testbeds (e.g. NDN testbed). We mainly accumulated experience with these tools and platforms and executed some early experiments. In particular, the implementation of a congestion feedback mechanism for NDN is shown and tested on a local CUTEi-based testbed. For details, please look at public deliverables D2.1 D3.1 and D4.1. In Y2, we developed live video streaming application comprising of edge computing and ICN. We provide a new methods on evolutionary deployable ICN and provide a new architecture named NEO and support this further with testbed validations. We also have three new solutions: a keyword based ICN platform, exploit pub/sub for data retrieval and fCOPSS for using name based pub/sub for IoT. We also have a lightweight named object method for programming and managing IoT devices. we have significantly improved NeMoI for network mobility and provide there new solutions namely “MAP-Me” for managing anchor-less producer mobility in ICN, “Track” for tracking system for moving objects and “EdgRepo” for mobile data repositories at the edge for supporting producer mobility. We provide four new solutions that propose in-device proxy re-encryption service for access control in ICN, route advertisement scheme for region control on ICN, provide emergency message delivery with source location verification and securing outsourcing of computations. A new testbed tool for CICN has been devised as an update of LURCH and Grid 5000. We also deployed a new local testbed infrastructure at URO to carry out experiments on high speed ICN routers, IoT, and ICN NoSQL databases. We evaluated the inter-operability feature in cefore that is newly supported with ICN platform in CUTEi testbed. We also designed, deployed and tested a EU-JP federated testbed made of a vertical integration between CUTEi testbed (JP) and Geant Testbed Service (EU), with the support of a gateway node deployed in URO. We designed, developed and tested an ICN 360 video streaming application with edge transcoding also presented at CEBIT exhibition. For further details, please take a look at the publicly available deliverable D2.2 D3.2 and D4.2. In Y3, we further enhanced our video and crowd-sensing applications. In the video application, research and tests based enhancements to our e360 application was made. Edge computing support was provided to the e360 applicaiton in order to enhance performance. The crowd sensing application was enhanced with an object identification mechanism which leveraged machine learning, edge computing and pub/sub features in order to take in support from people in a certain area to search for a missing object/person. We ran tests both on the local as well as the integrated testbed. For further details, please take a look at the publicly available deliverable D2.3 D3.3 and D4.3.
In Y1, we developed an NDN based video streaming application which enhances user QoE and works friendly with ICN in-network caches. We proposed pub/sub communication for IoT and emergency services, a light-weight pub/sub protocol for IoT, ICN based mobile edge computing for vehicular environments and a verifiable and flexible data sharing mechanism for IoT, a FoG based gateway to integrate sensor networks with Internet to realize IoT, services for NoSQL database. In Y2, we achieved solutions beyond the state of the art for the 360 degree video streaming. 360-degree panoramic video streaming are becoming very popular, due to the affordable price of cameras and the natural experience given by specific visors like smartphones. We presented the e360 (edge 360) application at the CeBIT 2018 in Hannover, the most important IT exposition in Europe. The most widely accepted ICN solution, Named-Data Networking (NDN) is based on hierarchical names, which easily lend themselves to naming aggregates of data. However, an application that wants to gather all data from temperature sensors, would prefer organizing names by sensor type (e.g. /temperature/...). Instead of defining multiple hierarchical names for the same data, we proposed using a keyword-based naming system to define IoT data available within an IoT domain (e.g. a company, a building, a city). In Y3, we further enhanced our applications and the tests that we performed. We developed a crowdsourcing prototype and demonstrated it at a festival of Osaka University, Ichosai. The demonstartion was introduced to more than 50 participants, including high school students, undergradudate students, technical junior college students, and their parents. Many students had an interest to our demonstration. We also introduced the technologies developed in the ICN2020 project at the the Wireless technology park (WTP) 2019, an exhibition held by National Institute of Information and Communications Technology (NICT) in Japan. WTP is one of the largest exhibitions in the area of communications and more than 50 thousand attendees came to WTP this year.