Whitepapers and State of the Art Papers have been written mostly by the
postgraduate researchers. All of the whitepapers have been evaluated in an
M-Zones internal peer review before they have been accepted for a deliverable.
All whitepapers are part of one deliverable and are published on the M-Zones
web server.
The internal review process has evolved over time. Starting
with a very strict review, lasting for nearly three months for the
state-of-the-art papers of Deliverable 1.1 and finishing with a more lax review
for the Deliverable 2.3 & 3.3. This evolution took place due to two
developments that happened within the first two years of the M-Zones programme:
·
Postgraduate students became more mature in their
writing
·
Many whitepapers of the second half of this period
have been submitted and accepted by peer reviewed conferences, thus minimising
the effort that we had to spend for an internal review.
The following statistical information can be provided:
·
Overall 43 whitepapers have been written
·
3 whitepapers have been jointly written by researchers
from two partners
·
1 whitepaper has been written jointly with external
researchers
·
The overall rates are CIT: 14 papers, TCD 16 papers
and WIT 15 papers.
The deliverables D1.1 (State of the Art), D2.1 & D3.1 and D2.3 &
D3.3 have evaluated the quality of the whitepapers from the perspective of work
packages. D1.1 provides the state-of-the-art paper as a collection of
knowledge, classified into a number of areas with an introduction for each
area. This deliverable and its papers represent the starting point of the
research work of the programme.
The deliverables D2.1 and 3.1 evaluate the early stage research work
with given criteria’s for the quality of the papers. The deliverables conclude
that the papers cover a wide range of subjects and mostly domain specific
solutions. All papers address the specific targets of the work packages, and
all specific targets are dealt with within at least one white paper.
Furthermore, D2.1 and D3.1 phrase some interesting and essential questions as
result of the evaluation of the white papers:
|
D#
|
Question
|
|
D2.1
|
How can
commercial mobile service providers use the integration of quality of service
and novel applications to attract subscribers to use the unlicensed band in
the face of free access competition?
|
|
How does
peer-to-peer and ad hoc networking impact on the design of application
services?
|
|
How can
runtime bindings between services and different network types be effectively
configured?
|
|
How can
knowledge be easily shared between individual developers and between software
entities to avoid the constraints of slowly developed standards while
encourage enough conceptual convergence to avoid a chaotic body of reusable
knowledge?
|
|
What level
of commonality is possible or useful between knowledge-based representations
of services, shared information and behavioural rules?
|
|
D3.1
|
What is
the underlying metaphor (paradigm) of a smart space that needs to be
addressed by engineering managed zones?
|
|
Can we
identify a meta model for loosely coupled components that enables
interoperability, scalability and portability of these components?
|
|
To what
extend do we need to incorporate semantic information into object models,
repositories and formal notations?
|
|
What are
the novel management functions that are needed to use, operate, control,
administer and maintain a smart space and a group of smart spaces?
|
|
What
services, protocols and formats must be provided by the underlying technical
environment to enable managed zones?
|
Furthermore, the deliverables raised some questions about the quality
evaluation process and about the research work within the programme:
·
Refinement of quality criteria for the work packages
·
Identification of a common layout and style for white
papers (and general research papers)
·
Agreement on a common terminology for managed zones
·
Integration of (currently) individual research
activities on institutional level as well as on the M-Zones programme level
·
Definition of interfaces between the work packages for
cooperation and to work package 1 for reuse and new input
At the next level, the deliverables D2.3 and D3.3 (beside
evaluating the white papers against the developed criteria’s) specifically
addressed the raised questions and focused on a qualitative evaluation of the
whitepapers regarding these questions:
|
D#
|
Question
|
|
D2.3
|
How can commercial mobile service providers use the
integration of quality of service and novel applications to attract
subscribers to use the unlicensed band in the face of free access
competition?
This
question has to some extent already been bypassed by events, with a rapidly
growing mixture of free, subscription and local payment WLAN access services
being made available – so the market is now in a position to deliver its
verdict.
|
|
How does peer-to-peer and ad hoc networking impact
on the design of application services?
This
question is of increasing importance as we examine ad hoc knowledge based
routing for the communication of application-level information, and as noted
above further work is required to determine if there can be fruitful reuse of
ad hoc networking techniques, e.g. route caching, to the knowledge-based
network domain. The impact of ad hoc networking on request-response style
service is still not clear, though the clear need for dynamic service
discovery in such situations has already been addressed within the project.
|
|
How can runtime bindings between services and
different network types be effectively configured?
This has
been addressed by (Murray & Pesch 2004b) in
that broad classes of applications can be factored into the decisions made
about inter-access network handover. In addition, the increasing assumption
of TCP/IP transport makes the choice of wireless access technology less
crucial to QoS unaware services. The separation of abstract service
specification from concrete protocol grounding as defined for the OWL-S service
ontology offers a more general mechanism for addressing this problem, as this
language become more mature and widely used across the programme.
|
|
How can knowledge be easily shared between
individual developers and between software entities to avoid the constraints
of slowly developed standards while encourage enough conceptual convergence
to avoid a chaotic body of reusable knowledge?
An
increasing focus on service and resource meta-data semantics, captured either
in XML schema or ontologies, offers a mechanism for such sharing, but
specific understanding of suitable guidelines are yet to emerge within the
project. The exchange of schema for testbed
integration between partners in WP4 may offer some practical experience in
these issues.
|
|
What level of commonality is possible or useful
between knowledge-based representations of services, shared information and
behavioural rules?
This is
still an open issue, but as the discussion of architectural issues above
reveals, solutions would seem to lie in models of individual components that
bring together segments of service, resources and policy rules centred on a
deployable and reusable smart space element.
|
|
D3.3
|
What is the underlying metaphor (paradigm) of a
smart space that needs to be addressed by engineering managed zones?
The
complexity of Smart Space management is understood within different research
domains and with specific view points. It seams to be necessary to look at
common characteristics of these complexities in order to fully understand the
managed area.
|
|
Can we identify a meta model for loosely coupled
components that enables interoperability, scalability and portability of
these components?
There is
no agreement on a common meta-model. Scalability is rarely addressed.
Portability is becoming an interesting issue, by means of mechanisms to
enable mobile policies and management activities. Interoperability is mainly
addressed by the exchange of information supported by ontologies.
|
|
To what extend do we need to incorporate semantic
information into object models, repositories and formal notations?
This is
still an important research question. It should be addressed keeping in mind
that semantic information can also be found in the modelling and design tools
employed in the process. It is commonly understood that this information are
essential for an effective management of Smart Spaces.
|
|
What are the novel management functions that are
needed to use, operate, control, administer and maintain a smart space and a
group of smart spaces?
Management
functions have to be portable across domains. This impacts business modelling
and makes global (at least inter-domain) naming and addressing of resources
necessary.
|
|
What services, protocols and formats must be
provided by the underlying technical environment to enable managed zones?
The
underlying technologies already provide services, protocols and formats. This
question has to be partially re-phrased towards how existing services,
protocols and formats can be accessed by M-Zones in order to fully exploit network
capabilities.
|
Based on the evaluation of whitepapers from early 2003 and early 2004,
both deliverables agree that these discussions raise further issues that would
be fruitful to be addressed in the programme’s ongoing research activities:
|
D#
|
Question
|
|
D2.3
|
Can a
useful common meta-model of a smart space component be agreed? What aspects
should such a model encompass and what level of intelligence should such
components embody?
|
|
Can ad hoc
routing principles be applied to knowledge-based routing and distributed
context querying?
|
|
Can the
range of policy-based management approaches addressed across the project be
addressed in a single approach covering user, smart space operator and
network service provider requirements?
|
|
Is dynamic
service composition sufficient for adaptive application behaviour in smart
spaces and is it suitably manageable?
|
|
D3.3
|
Will
policy-based management be the dominant mechanism for managing Smart Spaces?
|
|
What
characterises ‘adaptive policies’, what mechanisms are used to adapt rules
and what algorithms are important?
|
|
How is the
transformation across different levels of abstraction (ontology, object
model, technical environment) organised?
|
The following sub-sections give a complete list of all whitepapers.
