
January 2004
Wireless, Mobile, and Always Best Connected
By Máirtín S. O'Droma
Always Best Connected" (ABC), a fixed and mobile wireless access
vision, is an integral and challenging dimension in the evolving
paradigm of fourth-generation (4G) networks. It goes beyond the
"always-on" slogan and seeks a firm foundation in already existing,
expected, proposed, and yet to be conceived networks. The vision is
perceived as a strategic goal to inform and define the significant
advances happening and forecast in technologies, networks, user
terminals, services, and service delivery, and in the future business
models encompassing these, which will help crystallize and realize
the 4G paradigm. It presents great research challenges, with road
maps already beginning to take shape and evolve. Taking the ABC
concept as an underlying theme, the aim of this workshop is to
identify key issues, present 4G type scenarios, concepts, and
solutions, and speculate and foster debate on the evolution of the 4G
paradigm for future fixed and mobile wireless networks.
Next generation, Beyond 3G, and 4G are just some of the terms being
applied to this further evolution of mobile and wireless
communications beyond the 3G world. Whether the 3G world conceived as
the logical step from 2G and then 2.5G will ever come to be is a
separate question. Basically, through the provision of more
bandwidth, its promise is to be the first purpose-built wireless
network integration of voice, data (Internet), and video services. 4G
paradigms and visions being debated [1, 2] do not follow this logic
(more bandwidth, more integrated services) and in fact do not depend
on 3G being a success or even coming into existence. Rather, as
portrayed by Dr. Jorge Pereira of the EU Commission in his
presentation "Optimising Spectrum Efficiency in a Heterogeneous
Network Environment," 4G is conceived as encompassing all existing,
planned, and future mobile and fixed wireless networks, both
terrestrial and satellite. The goal is to sell comprehensive
integrated services in a more open, accessible, competitive
environment to a great common market of consumers in overlapping
local, regional, and global domains. Terms that characterize this 4G
debate include ABC, application/service adaptability and dynamic
quality of service (QoS) provisioning, and network and terminal
dynamic reconfigurability. Proximate challenges this 4G paradigm
raises are convergence, integration, and/or open interworking of
existing and emerging terrestrial (wired and wireless) and satellite
networks.
As may be gleaned from papers presented at this workshop, in
"all-IP" networking the ABC vision should include the capability of
flexible management of the all important QoS requirements, which are
largely realized in the lower protocol layers, but the management and
request generation of which range over all protocol layers. In
environments containing multiple wireless (and fixed) networks, ABC
includes open access to all of these networks under certain
conditions, and the ability to create and update a mix and match of
desired services based on acceptable price/performance ratios
all responding to requirements set down by users in their user
profiles. Whether in a mobile or fixed context, it also includes a
capacity to advertise, discover, and learn about new networks (and
nodes/users in ad hoc networks), new network and service options and
price/performance offerings, and dynamically change access, without
losing service connection, in accordance with user preferences, or
network provider or service provider obligations to meet per-service
or per-connection service level agreements, or service-dependent
application service provider (ASP)-desired QoS performance. The list
could go on. Largely the above is coming at an ABC definition from a
user's viewpoint. There are other viewpoints such as those of the
ASP, Internet service providers (ISPs), and network providers.
Clearly the implications for the reconfigurability of networks and
terminals, network interworking, application adaptability, business
models, and so forth will raise major new and exciting challenges. An
example of the early stages of the trend may be seen in present
commercial activity to create wireless network interworking involving
2.5G cellular systems, 802.11 wireless LANs (WLANs), and 802.15
Zigbee to deliver services over a type of always-on or
always-connectable wireless network access connection service at
attractive location-dependent price/performance ratios [3, 4].
In this first international workshop organized by ANWIRE [5], three
of the invited speakers, Prof. Dipankar Raychaudhuri, Director of
WINLAB, Rutgers University, United States, Prof. Barry Evans,
Director of the Centre for Communication Systems Research, University
of Surrey, United Kingdom, and Prof. Bernhard Walke, Director,
Communication Networks Laboratory, University of Aachen, Germany,
propose and discuss futuristic though realistic networks and
interworking scenarios.
Speculative Networks
Prof. Raychaudhuri in "Topics in 4G Wireless Networks: Ad Hoc Nets,
Adaptive Services and QoS" addressed novel concepts and scenarios. He
argued the case for promoting opportunistic and adaptive formation of
wireless network infrastructures as a way of overcoming critical
technical and business model limitations of today's cellular and WLAN
networks. He proposed novel ideas for creating a wireless "network of
networks" together with techniques for support of heterogeneous
radios, protocols for self-organization and ad hoc routing, and
services that adapt to dynamic and opportunistic exploitation of
available or expected resources to meet application needs. Within
this context he also discussed software architectures suitable for
adaptive service creation and QoS management. Supporting his ideas,
Prof. Raychaudhuri presented some initial results from prototyping of
candidate 4G wireless network technologies, such as infostations, ad
hoc 802.11 networks, and sensor nets.
Satellites have a unique role and great importance in wireless
communications. Despite their daunting upfront capital investment
demands, seeing their amazing capacity to be successfully and
flexibly involved in virtually all communication services markets, it
was considered important that they be included in a workshop on the
evolution of ABC 4G wireless and mobile networks. This is in contrast
to most cellular mobile conference traditions! As Prof. Barry Evans
pointed out, "The mention of satellites in any gathering of mobile
communications specialists tends to engender either disinterest or
outright hostility. This in itself is an interesting reaction and
symptomatic of the classical and still pervasive division between the
terrestrial and satellite camps which has been rife in the standards
bodies since time immemorial." Almost by way of response to this
unattractive division, in his paper "Satellite Networks in 3G and
Beyond Systems" Prof. Evans provided an incisive and useful review of
fixed, broadcast, and mobile satellite services: their contributions
to date in effectively and distinctively serving commercial
communication needs, and proposed ways their unique attributes will
contribute and complement the evolution of 4G network infrastructures
and service delivery. Among the examples on which he expounded, he
highlighted the advantages satellites have in delivering broadcast
and multicast services to the mass market relative to the cost and
complexity of their delivery through 3G networks. Much of this theme
was echoed and reaffirmed from a quite different perspective by Prof.
Bernhard Walke in his wireless media system (WMS) proposals for 4G
infrastructural environments, including the idea that with
sufficiently large storage capacity in user terminals, unidirectional
point-to-multipoint (p-t-m) services are able to provide on-demand
and interactive applications because push and store mechanisms make
the p-t-m relationship transparent to users.
In illustrating the development of these ideas with pilot and
prototype field trials, Prof. Evans spoke about the EU project SATIN
where a geostationary Earth orbit (GEO) satellite was being
complemented by terrestrial repeaters called intermediate module
repeaters to provide mobile network operators with cost-effective
solutions of increased p-t-m transfer capacity over their whole MNW
coverage. The system uses 3G Partnership Project (3GPP) technology to
prevent any cost impact on the terminal. Another EU project, MoDiS,
provided a demonstrator in 2003 using the 3G network in Monaco. The
system will be an example of a truly integrated satellite/cellular
system for use by mobile operators and is the way forward for similar
collaborations toward 4G.
In a groundbreaking invited paper Prof. Bernhard Walke and his
co-authors, Ian Herwono, Ralf Pabst, and Daniel Schultz, at the
Communication Networks Laboratories, Aachen University (RWTH),
Germany, and Joerg Habetha of Philips Research, have drawn on a
wealth of research experience in evolving low-power wireless
networking concepts to propose practical integrating and
interoperating networking systems. Targeting goals of highest
efficiency and lowest possible transmit power, their broadband WMS
proposal will use 1 W air interfaces for pico base stations and fixed
wireless routers (FWRs). The latter are incorporated to extend radio
coverage range of the high capacity available from a pico base
station. Connectivity relies essentially on multihop communication
across wireless bridges/routers and to some extent also on ad hoc
networking. Both base stations and FWRs are small enough in size to
allow the system to be spoken of as an "invisible infrastructure."
The low RF radiating power, while maintaining high bandwidth and high
connectivity, naturally has the attraction of low radio exposure of
users.
Prof. Walke also described ways this broadband WMS concept combines
broadcast, multicast, and single-cast services to minimize the number
of transmissions required, thereby again reducing radio exposure of
humans, to provide the contents requested by user terminals. The
attraction and potential of integrating telecommunications satellite
and HAPs into the WMS is also addressed here.
For any of these speculative networks to realistically set the
roadmaps or trends toward 4G, appropriate standards, as always, are
needed. In his paper "Internet Protocols for Wireless Internet,"
Prof. Kimmo Raatikainen of the University of Helsinki, Finland,
reviews the protocols developed in the Internet Engineering Task
Force (IETF) that may provide the foundation for a ubiquitous ABC
wireless Internet. He also discusses recent developments in the Open
Mobile Alliance that address some middleware issues in the wireless
Internet.
QoS-Related Issues
The first of several papers to address QoS issues was that by Dr.
Yevgeni Koucheryavy, Tampere University of Technology, Finland,
entitled "On Quality of Service and Performance Evaluation in 4G
All-IP Networks." The theme taken up in his paper is that
reconfigurable mobile networks (including ad hoc networks) are
envisioned as being highly dynamic self-organizing and
self-configuring networks, which will not necessarily require fixed
infrastructure and may take on the characteristics of an autonomous
system in the Internet. Within this view, and in a context of all-IP
networking, meeting QoS requirements and creating traffic engineering
enablers will be defining challenges.
In "Physical Layer Strategies and Issues for ABC Vision," Dr. Carlos
Bader and Dr. Carolina Pinart of Telecommunications Technological
Center, Catalonia, Spain, discussed the interesting topic of support
for reconfigurability through cross-layer protocols at the physical
and link layers. They proposed a classification of information
exchange in the cross-layer consisting of channel state information,
physical layer resources, and QoS parameters, and suggested the use
of fuzzy logic to assist in the reconfigurability decision process.
Seeing 4G as a movement toward a new "anytime, anywhere, with
anybody" type of communication, through and over all-IP multi-access
networks together with multimode terminals and intelligent access
selection that will allow the user to be in some sense always best
connected, is the motivation behind the paper by Dr. Ivan Armuelles
and Prof. Tomas Robles, Technical University of Madrid, Spain. In
such a scenario, a mobile terminal is the subject of interactions
that presuppose new functionality requirements, including novel QoS
support mechanisms. To assist the meeting of these requirements, they
describe a proposed new "Enhanced Socket Interface for Next
Generation Systems," which is an advanced interface defined as a
component of an open and modular terminal architecture developed in
MIND, an EU IST project.
QoS provision for wireless mobile nodes operating freely and
reconfigurably in a multitude of heterogeneous access networks was
further addressed by Dr. Nikos Passas and Prof. Lazaros Merakos of
the University of Athens in their paper "Unified QoS Provision in
Wireless Access Networks." Transferring to wireless systems the QoS
advantages DiffServ and IntServ bring to the core and edge of wired
IP networks cannot be done smoothly. They describe some of the new
problems and potential remedies characteristic of wireless and mobile
access that these IETF IP QoS fixed network solutions must face, and
for which they suggest potential remedies.
Managing ABC
ABC decisions naturally will be made on the basis of criteria, many
of which will be set down in multidimensional user, terminal,
network, and service profiles. Such profiles will grow in complexity
with time, in parallel with the growth in the range of terminal,
network, and service access options, and consist of complex sets or
arrays of competing parameters. How this may be managed in a dynamic
adaptable way is no small challenge. Deriving a structured
implementable policy-based management framework as a way to begin
addressing this need is the subject of a paper by Hakima Chaouchi and
Prof. Guy Pujolle, Paris 6 University, France, "Policy-Based
Management Framework for Always Best Connected Users." Their
framework allows for awareness of specific application QoS
requirements, terminal capabilities, and the characteristics of the
available access networks. They include an outline specification for
a suitable network and ABC terminal architecture. In addition, they
propose policy-based handover management. This is a handover
triggered not only on the basis of a fading signal, but taking into
account other constraints such as cost, resource availability, and
load balancing. They discuss how such a policy-based handover will be
useful in a heterogeneous 4G wireless network environment.
Ad Hoc Networking
Within the vision of 4G as consisting of multiple heterogeneous
networks, there is little doubt that one genus of these is being
spoken of as a new multiform dynamic networking infrastructure with
important potential contributions, particularly in the areas of
always-on and ABC. Naturally then, a number of forward-looking papers
on ad hoc networking were selected for presentation at the workshop.
In the first of these Dr. Takayuki Yamamoto et al., Osaka
University, Japan, proposed in "Routing in Ad Hoc Networks for
Processing Many Short-Lived TCP Connections" a novel low-latency
hybrid routing protocol (LHR) that resolves a number of performance
problems of existing routing algorithms designed to support
persistent TCP connections. Their technique aims to bypass the large
overhead of exchanging routing tables, and the large latency for both
an initial route search and new route searches in the case of link
disconnections. Their LHR combines on-demand route search and
proactive route maintenance. With comparative performance analysis
they demonstrate its superior performance.
QoS in IP ad hoc networks is addressed in two papers. In one,
presented by Michael Barry, University of Limerick, Ireland, "QoS
Techniques in Ad Hoc Networks," a multidimensional network model is
described where different elements of ad hoc networks and their
influences on RRM and QoS can be investigated.
Anelise Munaretto, Guy Pujolle, et al., Paris 6 University,
France, proposed an optimized version of the classical link state
routing protocol (OLSR) as a basis for designing a QoS-based routing
scheme (QOLSR) suited to mobile ad hoc networks (MANETs), where route
choice is based less on hop metrics and more on delay metrics.
ANWIRE Panel Session
Dr. Jorge Pereira, EU Commission, chaired the workshop panel session
focused on a review of ongoing activities of the ANWIRE task forces
(TFs), in the context of the 4G and ABC debate. It included a
particularly interesting paper and presentation entitled "Towards an
ANWIRE B3G Wireless System Integration Architecture" by M. Siebert
(RWTH), H. Chaouchi (LIP6), A. S. Jahan (KCL), I. Demeure (ENST), I.
Armuelles (UPM), L. Palma, & L. M. Correia (IST-TUL), J Liu
(CTTC), M. O'Droma (UL), V. Friderikos (KCL), W. Xing (UCY), and N.
Alonistioti (UoA). With the aim of generating creative proposals for
an integrated system service architecture, this paper summarised a
detailed structured review of current system integration R&D
efforts happening within Europe and elsewhere in the world. The
ANWIRE TF is clearly making a good start toward evolving an initial
set of requirements for a targeted architecture by extracting,
crystallizing, and integrating ideas from many excellent but perhaps
independent and autonomous projects.
Another interesting paper was presented by Vasilis Friderikos of
KCL, UK, on behalf of his 12 co-authors from six EU research centres
C. F. Bader (CTTC), C. Pinart (CCTC), Z. Boufidis (UoA) , N.
Passas (UoA), I. Ganchev (UL), M. Dikaiakos (UCY), C. Christophi
(UCY), E. Tsiakkouri (UCY), Wei Xing (UCY), T. Robles (UPM), J.
Salvacha (UPM) and E. Tsontsis (KCL) "Analysis of User
Requirements for Future Heterogeneous Wireless Networks." In this
they indicate that the attempt to define an overall user-centric
approach to heterogeneous mobile/wireless network design was one of
the main driving forces in establishing 4G user requirements. The
ANWIRE TF has organized these requirements structurally within three
generic classes: connectivity, service provision, and adaptability
and reconfigurability. Also, objective criteria for QoS
characterization are being explored.
Other formal contributions were made by Dr. G Morbito, University of
Catania, on efficient and always-on connectivity, by Dr. N.
Alonistioti and Dr N. Houssos, UoA, Greece, on 4G reconfigurability
architectures, and by Dr. I. Demeure, ENST, France, entitled "Towards
Adaptable Services."
Dr Periera opened the discussion to the floor and, naturally,
participation was lively, especially with lateral thinking
contributions from such distinguished academics as Profs. Hamid
Agvami, Bernhard Walke, Barry Evans, Kimmo Raatikainen, Guy Pujolle,
and others. These may be listened to on the DVD[6].
Conclusion
ANWIRE's first international workshop was a success in both
technical content and level of the papers presented. As may be seen
from the papers presented at this workshop, fundamental research in
wireless networking is moving rapidly into a new phase. The papers
originate from a wide spectrum of authors covering 12 countries and
three continents. Overall the workshop and these proceedings
represent a fine contribution to the research efforts and debates in
this field, and provide the reader with new and stimulating insights.
Copies of the proceedings may be ordered at http://www.ece.ul.ie/anwire-workshop-glasgow/proceedings/index.html,
or obtained on loan from the British Lending Library.
References
[1] The Book of Visions 2000 Visions of a Wireless
World, v.1.0 3.5, IST Wireless Strategic Initiative, Nov 2000; http://www.wireless-world-research.org
[2] M. O'Droma et al., "Always Best Connected" Enabled 4G
Wireless World," Proc. 12th IST Summit on Mobile and Wireless
Commun., Aveiro, Portugal, June 2003, pp. 71016.
[3] E. Gustafsson and A. Jonsson, "Always Best Connected," IEEE
Wireless Commun., Feb. 2003, vol.10, no.1, pp. 4955.
[4] R. Becher et al., "Broadband Wireless Access and Future
Communications Networks," Proc. IEEE, vol. 89, no. 1, Jan
2001, pp. 5875.
[5] http://www.anwire.org
[6] http://www.ece.ul.ie/anwire-workshop-glasgow/proceedings/index.html
Sixth Baiona Workshop on Signal Processing in Communications
By Carlos Mosquera and Fernando Pérez-González,
Universidad de Vigo, Spain.
At a time when so many conferences compete for getting attendees and
good quality papers, we are proud to host a small workshop known as
the Baiona Workshop on Signal Processing in Communications. The use
of the same location for every edition, the lovely village of Baiona,
northwestern Spain, just 20 km from Portugal on the Atlantic coast,
makes this workshop different from most conferences in the field. The
workshop site, a medieval castle transformed into a nice hotel, helps
create an atmosphere conducive to technical discussions and social
life. On top of that, we cannot forget the nearby Cies Islands, a
great natural resort that hosts a regular social session of the
workshop.
The first reason for choosing Baiona back in 1991, when the first
Baiona Workshop took place, was its proximity to the campus of the
University of Vigo. Our university was involved at that time in an
European COST action that grouped quite a few European research
groups with specific interests in signal processing and its
applications. The workshop served as a meeting point for those
involved in the COST action, with some additional contributions from
outsiders that helped to enhance its scientific quality. With the
passing of time, the workshop gained an identity of its own, and has
acted as a bridge between American and European scientists, with the
incorporation of people from Asia as well. We should remember that
Baiona was the first European village to learn that Christopher
Columbus and its three vessels had reached the "Indias." The Pinta
arrived back in Baiona to announce the discovery of what would be
referred to as the "new world." Five centuries later communication
technologies have reached amazing frontiers, but many challenges lie
ahead awaiting a new Columbus to solve them.
The invitation of outstanding plenary speakers is a must at every
Baiona Workshop. They are expected to give encouraging tutorials on
different topics linked to signal processing in communications. These
tutorials are designed to help people understand recent advances and
challenges in different areas of signal proccesing in communications.
The list of names is quite long, and without trying to be exhaustive
we have enjoyed, among others, the dissertations of Thomas Kailath,
Sergio Benedetto, David Messerchsmitt, Peter Hoeher, Simon Haykin,
Andrea Goldsmith, and Giorgios Giannakis. Next we include the names
of the invited speakers and the topics of their lectures in this 2003
edition:
- Edward J. Delp: "Multimedia Security: So What's the Big Deal?
- David Falconer: "Frequency Domain Processing in Broadband
Wireless Systems"
- Giorgios B. Giannakis: "Full-Rate Full-Diversity
Complex-Field Space-Time Coding"
- Andrea Goldsmith: "Design Challenges in Future Wireless Systems"
- Marco Luise: "A Review of Soft-Information-Based
Synchronization for Iterative Data Demodulators"
- Fernando Pereira: "Developments and Trends on Multimedia
Content Adaptation"
In addition to the invited lectures, about 60 contributions are
selected to be presented in poster sessions, thus providing an
exciting environment for discussion. The authors of the best papers
are offered the opportunity to write extended versions to be reviewed
as possible contributions to the Signal Processing Journal of
Elsevier.
At this edition, precoding and cross-layer design were presented as
some of the most promising schemes to increase communication rates
and provide adequate quality of service for end user demands. "Dirty
paper coding" and "turbo synchronization" are fancy terms to describe
state-of-the-art research in coding with side information and joint
synchronization-turbo decoding, respectively. Signal processing also
has an important role to play in digital content protection and
multimedia content description.
All the documentation regarding this workshop, including the papers
in downloadable format, can be found at http://www.baionaworkshop.org.