End of March, I visited the Internet of Things 2008 conference in Zurich. The conference was an interesting mix of business and research people, discussing the Internet of Things around RFID and beyond.
I was invited to talk about the challenges ahead for the Internet at large, driven by the increasing cross-value chain collaboration that is driven forward not only but in particular by RFID.
In this talk, I tried to make the point that the discussion around Internet of Things, Internet of Services, and Internet of whatever is counterproductive and an indication of today’s still prevailing ignorance and inability to consider the Internet as a large-scale system that requires system solutions. I put forward a suggestion for overcoming this silo-ed thinking through an architectural vision called tussle networking, which has been presented earlier in this blog. This vision envisages a collaborative environment in which concerns of participants are captured (through intelligent methods found in, e.g., context awareness) and expressed through explicit policies, executed by a system centered around information, its provisioning, representation and reasoning over.
It was certainly a different view in this conference, creating interesting debate at the end of the presentation. You can find the slides of my presentation here.
7 users commented in " Internet of Things 2008 "
Follow-up comment rss or Leave a Trackback[...] unknown wrote an interesting post today onHere’s a quick excerptEnd of March, I visited the Internet of Things 2008 conference in Zurich. The conference was an interesting mix of business and research people, discussing the Internet of Things around RFID and beyond. … [...]
[...] Internet Marketing Center wrote an interesting post today onHere’s a quick excerptEnd of March, I visited the Internet of Things 2008 conference in Zurich. The conference was an interesting mix of business and research people, discussing the Internet of Things around RFID and beyond. … [...]
[...] can read the rest of this blog post by going to the original source, here [...]
Hi Dirk,
IMHO, looking towards an information-centric future Internet is the right step given the actual use of the Internet as a content/information creation and delivery platform.
REgarding the tussle networking concept, I observed in the slides describing the architectural implementation the notion of Knowledge Base present in the concerns, policy and semantic planes. Could you elaborate a little more on this concept? Is it a mean to generalize every piece of information as a knowledge resource?
Warm regards and keep posting!
Christian
Hi Christian,
Thanks for the encouragement!
The knowledge plane refers to a proposition made originally by D. Clark et al, published at SIGCOMM 2003. It proposed an intermediary plane that hosted knowledge of the Internet resources (such as topologies and more) in order to reason about failures; a sort of new network management.
The reference in my slides is meant to point out that the tussle networking vision goes beyond the network management aspects of the knowledge plane. The focus on policies, expressing concerns and allowing for deliminating tussles, goes beyond pure management towards runtime system configuration.
The commonality comes with the necessity for a large-scale provisioning system for the required information to reason about. The difference to the knowledge plane lies in the proposition that I’m making (in the architectural implementation slide) that EVERYTHING is treated as information on the internetworking layer, not only management-related information.
This rather low level concept of information allows for building up powerful information and eventually knowledge concepts (like ontologies), even to the point to embed policies and context rather low level through representing these as metadata to information (metadata which in itself is simply information). It is some kind of bootstrapping complex information systems right on the internetworking layer, something quite different from today’s (IP) Internet.
Best,
Dirk
Hi Dirk,
thanks for the explanations! I think it becomes even clearer in your next post reference to the available PSIRP vision document with some use cases.
I had read the MIT “tussles in cyberspace” paper a time ago, however I was not aware of the follow-up paper you pointed out. Though conceptually very interesting, it is still difficult for me to understand how to translate the concepts into real-word scenarios, something you have tried in your “architectural implementation” design, but to me it is still too abstract…
I have been doing some research due to some R&D project inquiries and I found interesting similarities between the upper layer PSIRP vision and what is going on in the FP6 IST SPICE EU project, where there is also the notion of “Knowledge Layer” (see quote below from SPICE Unified Architecture Phase 2 – http://www.ist-spice.org/documents/SPICE_WP1_unified_architecture_Phase%202.pdf).
Of course, in the SPICE case it is heavily telco-driven and relates to telco service architectures (e.g. IMS) but nevertheless they share somehow the same spirit, isn’t this another sign of convergence??
Best regards,
Christian
”
Knowledge Layer
The Knowledge Layer provides service platform solutions for intelligent service behavior, user profile management, and pro-active service adaptation. These solutions are grouped in different enabler families that are all built on top of a common framework.
The Knowledge Layer supports the publication, discovery, delivery, and transformation of various information such as context, user profile and presence information. In SPICE, this information is generally referred as knowledge, where knowledge is described as expertise, skills acquired by a person through experience or education, the theoretical or practical understanding of a subject, what is known in a particular field or in total, and facts and information or awareness or familiarity gained by experience of a fact or situation. In the context of SPICE it is distinguished between lower and higher order knowledge, Lower order knowledge is propositional knowledge obtained by experience or sensorial information. That is, lower order knowledge is defined as information collected from different sensors, devices and data sources. By contrast, higher order knowledge is derived recommendations, predictions, activities and goals.
The Knowledge Layer is realized as a collection of distributed knowledge sources, sinks, and brokers. Knowledge
sources produce information, which is then consumed by the knowledge sinks. Information is either delivered directly or by employing the publish/subscribe pattern. Brokers are responsible for mediating and transforming information, and providing interfaces for knowledge discovery. Discovery is performed either using direct query to a
broker or by using the publish/subscribe pattern. The knowledge enrichment algorithms that are used in the knowledge layer, such as learning and prediction algorithms, are pluggable and new techniques can be added to the system.
“
[...] of semantics in the new Internet, leading to the notion of a A Knowledge Plane for the Internet. I observed this tendency only recently in a post from Dirk Trossen on the architectural vision called tussle [...]