Conference PaperPDF Available

Hybrid Seamless Mobility Supporting Pervasive Service Collaboration

  • Whitestein Technologies AG, Switzerland

Abstract and Figures

This paper discusses an approach to seamless hybrid connectivity bridging infrastructure-centric and ad-hoc networks to autonomically maximize the potential for sustained connectivity for terminal device users and their services. The architecture and functional attributes of a hybrid connectivity manager are described, and illustrated through deployment in an incident scenario where rescue workers require seamless hybrid connectivity to support sustained collaboration. The HCM extends an existing technology, the living systems connection agent for seamless vertical handover between infrastructure networks, with support for ad-hoc connections, and employs a policy-based logic engine a software-agent autonomic controller.
Content may be subject to copyright.
Hybrid Seamless Mobility Supporting Pervasive
Service Collaboration
Dominic Greenwood
Whitestein Technologies AG
Pestalozzistrasse 24
8032 Z¨
urich, Switzerland
Roberto Ghizzioli
Whitestein Technologies AG
Pestalozzistrasse 24
8032 Z¨
urich, Switzerland
Monique Calisti
Whitestein Technologies AG
Pestalozzistrasse 24
8032 Z¨
urich, Switzerland
Abstract This paper discusses an approach to seamless
hybrid connectivity bridging infrastructure-centric and ad-hoc
networks to autonomically maximize the potential for sustained
connectivity for terminal device users and their services. The
architecture and functional attributes of a Hybrid Connectivity
Manager are described, and illustrated through deployment in
an incident scenario where rescue workers require seamless
hybrid connectivity to support sustained collaboration. The HCM
extends an existing technology, the Living Systems Connection
Agent for seamless vertical handover between infrastructure
networks, with support for ad-hoc connections, and employs a
policy-based logic engine a software-agent autonomic controller.
Next Generation Networking (NGN) consists of IP packet-
based network deployments capable of providing telecom-
munications services across multiple broadband, QoS-enabled
transport technologies wherein service-related functions are
explicitly separate from underlying transport-related technolo-
gies. NGN also intrinsically supports mobile users with the
ability to communicate and access ubiquitous services irre-
spective of changes in location, terminal device or access
connection. By virtue of these characteristics NGN allows
myriad service providers to offer their services to users across
multiple connection types.
Due to this delineated design and support for multiple
service providers, within emerging NGN deployments such
as British Telecoms 21CN initiative [15] [14] the global
provisioning of pervasive telecommunications services is be-
coming rapidly suffused with user-driven content and third-
party services accessible across multiple, seamless access
technologies. Seamless access is of significant value to users
as it transparently converges multiple wireless communications
technologies through uninterrupted vertical-handover between
the various access connections available to a user device
according to connection availability, user preferences, operator
and service provider policies. Mobile-IP based technologies
such as the Living Systems Connection Agent [5] provide
seamless connectivity in accordance with the emerging IEEE
802.21 standard [16] [17] [14]. Session continuity is assured
as connections transition across access technologies allowing
client devices to maintain the same IP address for an entire
session and applications to run uninterrupted.
To date, state-of-the-art seamless vertical-handover, de-
fined by the IEEE 802.21 standard as Media independent
handover (MIH), most typically supports handover between
infrastructure-centric access connections such as cellular, wire-
less LAN and WiMAX networks. Our approach extends this
model to support Social Area Networks (SANs) through
geographically localised ad-hoc connections over technologies
such as Bluetooth, ad-hoc wireless LAN, and infrared, etc.
This hybrid network [11] [4] is a substrate allowing end-to-
end connections between terminal devices that may consist of
multiple connection segments seamlessly transitioned across
infrastructure and ad-hoc network technologies.
Over this hybrid network substrate resides the service pro-
visioning domain wherein conventional operator-provisioned
services are accompanied by passive content and active ser-
vices contributed by so-called prosumers [18] and third-party
provider services. These services, which may also draw on pas-
sive content resources, can collaborate across the potentially
changing underlying network landscape by virtue of seamless
handover technology to form short or long lived composite
service assemblies. While such collaborations may draw on
widely dispersed services and content, of particular interest
is the ad-hoc seamless mobility extension discussed in the
following sections which allows content and services located,
and perhaps generated by, diverse prosumer terminal devices
to collaborate within geographically constrained Social Area
Networks that span devices present within the range of ad-hoc
technologies such as Bluetooth and ad-hoc wireless LAN.
Such composite assemblies of services result from localised
collaboration as appropriate to scenarios such as the emer-
gency worker coordination case outlined in Section V, or more
generally, cases where prosumers choose to collaborate with
one another by generating and sharing services and content in
a common location. In all cases while users may choose to take
advantage of access to other locally located terminal devices
to collaborate via short-range ad-hoc technologies, they will
nevertheless remain connected via seamless connectivity to
available infrastructure networks. This implies that service
composite assemblies may, and will, be constructed from
services drawn from both local and remote sources dependent
on application context; in fact such assemblies may change
both their constituents and the means of communicating with
those constituents dynamically and in real-time.
Creating content, combining it into short-lived, ad-hoc
composites with other diverse content types, and seamlessly
provisioning it over multiple access technologies is a central
component of the Telco 2.0 vision, substantially informed
by operator network convergence strategies and the user-
driven content revolution of Web 2.0. This paper proposes
a technology substrate for offering seamless connectivity for
client device users across a hybrid of standard infrastructure
networks (cellular and wireless) and ad-hoc peer-to-peer net-
works (bluetooth, wireless ad-hoc, etc.).
The following section discusses some relevant published
work. Section III presents the architecture of a proposed Hy-
brid Connection Manager (HCM). Section IV then describes
the connection and usage aware connectivity management
features of the HCM and section V illustrates a real-life
scenario within which the HCM prototype has been evaluated.
Among published works dealing with hybrid networks
consisting of a combined infrastructure and mobile ad-hoc
connectivity, only a few then consider either the requirements
and influence of the user applications running over the auto-
nomic communications subsystem, or the roles of end-users
in deployment scenarios.
An example of hybrid network management is provided in
Hauge et al. [11] who discuss an approach to the combined use
of 3G cellular and ad-hoc networks. The authors conclude that
hybrid networks provide the opportunity to transmit service
data to a higher percentage of interested mobile terminals
than when using only an infrastructure network. Nevertheless,
the role of their user-level service (multicast) within hybrid
networks is not considered. The same can be said of the
work of the Delay Tolerant Networking Research Group [1]
chartered as part of the Internet Research Task Force (IRTF).
This group focuses on network aspects providing end-to-end
connectivity in disruptive environments without considering
how application contexts influence the achievement of con-
nection/delivery goals. We address this issue as a component
of the HCM usage-aware approach (see section IV).
Another important work in this area is that of Kappler et
al. [12] who use a policy-engine to address hybrid network
composition. Although this aspect is in line with our approach,
the authors do not consider the discovery of relevant network
nodes, and policies are only used for the composition of
network devices, whereas our approach also considers the
possibility of user-level service composition.
The notion of Unified Messaging (UM), as reported in van
der Meer et al. [13] for example, is also closely related to
our work from the perspective of supporting both fixed and
mobile users with universal access to communication services.
The central concept of UM is the capability of the messaging
system to select the most appropriate terminal or application
for an incoming message according to availability, status,
and other parameters. A UM system is designed to adapt
terminals to different kinds of services via content adaptation
processes guided by user rule policies. However, the particular
approach documented in [13] identifies CORBA as a suitable
means of engineering the middleware software for handling
UM; an approach, in our opinion, not entirely suitable for
application in pervasive environments where seamless mobility
is a paramount issue and small footprint devices are the norm,
as is particularly the case in disruptive environments.
Additionally, published work relating to the use of ambient
and pervasive technologies in disruptive environments or dis-
aster management tends to focus more on service composition
or other application-oriented aspects without considering the
underlying networks issues. Such an example is the reported
work of Kristensen et al. [6] which focuses on IT support in
major incidents, such as the use of bio-monitors, patient iden-
tification and collaboration tools for response units, without
considering how these applications could, or should, behave in
the presence of network disruptions. The HCM system reduces
this gap between specific autonomic aspects purely based on
the network management and the autonomic aspects based on
the user-level services deployed in disruptive environments.
The Hybrid Connection Manager (HCM) is terminal device
logic extending an existing seamless mobility technology
known as the Living Systems Connection Agent (LS/CA) [5]
which enables vertical-handover between Wireless LAN and
Cellular technologies (i.e., GPRS, UMTS, HSPA). The LS/CA
is based on Mobile IP technology which offers session conti-
nuity by allowing a client to maintain the same IP address for
an entire session and applications to thus run uninterrupted.
The HCM extends the LS/CA with support for seamless han-
dover between ad-hoc networks including Bluetooth, ad-hoc
Wireless LAN and IrDA. The result is seamless connectivity
across any available combination of hybrid infrastructure and
ad-hoc networks. Both the original LS/CA and the HCM ex-
tension automatically detect available networks, with support
for new technologies added via a plug-in architecture.
The HCM mirrors the LS/CA in being designed as an
autonomic system that shifts selected configuration tasks from
users to a specialised self-regulating subsystem. A local
policy engine is used by each HCM deployment for auto-
configuration purposes, allowing autonomic adjustment of
behaviour in accordance with environmental changes. The
HCM is also capable of auto-optimization by monitoring
network resources and adapting its behaviour to meet end-
user and application service connectivity requirements, i.e.,
automatically handing-over sustained sessions between differ-
ent access connections when a connection failures occur, or to
simply maximise the goal of remaining connected to the best1
available connection.
The HCM software architecture is depicted in Figure 1 and
consists of an autonomic manager and several managed ele-
ments: an user application interface service, a network bearer
1A ’best’ connection may be considered to be that with the optimal QoS
interface service and policy service. The autonomic manager
uses a policy engine decision logic to control seamless han-
dover by responding to network events, application events in
accordance with dynamically configurable policies. Policies
are used to guide the autonomic decisions that the HCM
autonomic controller can take in response to events sensed
from the environment. A simple example is the detected failure
of an infrastructure connection (say WLAN); a predefined
obligation policy mandates that the HCM re-route2messages
via one or more alternative network technologies to either their
final destination or another node with an active infrastructure
connection. An example of such a policy is described in
Section V.
Sensors Effectors
Bearer Service
Sensors Effectors
Effectors Sensors
Network Components
User Applications
Managed Elements
Autonomic Manager
Control Agent
Fig. 1. HCM software architecture.
The HCM software is designed to interact with different
layers of the user-device communication stack. In the default
case it interacts with lower-layer communications components
dealing with bearer technologies, e.g., UDP, Bluetooth, etc.,
and with upper-layer components dealing with, for example,
routing protocols, discovery, and services/applications. Other
components may also interface to the HCM through provided
software interfaces.
A prototype of the HCM has been created using the
JADE [2] (Java Agent DEvelopment framework) multi-agent
system development and runtime environment. As JADE3
supports J2ME and MIDP it was an ideal candidate for
initial fast-prototyping work. The HCM autonomic manager
is designed for deployment as a JADE agent and the managed
element interfaces are deployed as JADE kernel services, both
of which are executable within the JADE runtime.
A. The Managed Elements
The HCM control agent interacts with the external world
via JADE kernel services. Each service is controlled through
2Routing of traffic, especially over ad-hoc connections, is considered
beyond the scope of this paper
3JADE documentation and software is available from
sensors and effectors. Effectors produce actions relating to
instructions received from the HCM agent; they implement the
Command design pattern [7]. Sensors collect information from
the external world and provide it to the agent for processing;
they implement a simplified version of the Half-Sync/Half-
Async design pattern [3]. The HCM has three JADE kernel
The Network/Application Service is used by the HCM
agent to interact with upper-layer network services, such as
routing and discovery, and application services facing the user.
Messages passing through this interface may be inspected by
the HCM to determine whether any action is necessary by the
autonomic manager in accordance with specified policies. A
selection of interfaces are available allowing communication
with a broad range of network services and applications.
The Bearer Service is used by the HCM agent to interact
with the lower-layer bearer interfaces to both infrastructure
and ad-hoc network endpoints. Messages passing through this
interface may be inspected by HCM to determine whether any
action is necessary by the autonomic manager in accordance
with specified policies. A selection of interfaces are available
for a broad range of network technologies.
The Policy Service is used by the HCM agent to interact
with the local policy engine (see Section III-C). In brief, this
engine possesses the operational rules that must be applied to
control (or not control) the way in which messages are treated
by the HCM. The HCM agent uses these policy rules to effect
this control.
B. The Autonomic Manager
This is the software agent that controls the HCM system
such that it exhibits the following autonomic behaviours:
Sensing: By installing sensors in the managed elements the
agent is able to monitor new application messages to be sent,
new messages received from the network or new network sta-
tus events. Data is collected both asynchronously (the managed
elements notify of a status changing) or synchronously (the
agent explicitly request for information).
Compiling Knowledge: Intercepted messages and received
events are used to create an internalized model of the external
world. This compiled knowledge base contains information
such as statistical flow data, historical fault logs, active and
treated faults, discovered devices and the services those de-
vices provide.
Decision Control: Whenever the internal knowledge base
is updated, the HCM agent triggers a call to the local policy
engine which dictates the policy constraints that must guide
decision-making by the agent. This aspect is discussed in more
detail in Section III-C.
Proactivity: Actions can be executed by the HCM agent
immediately, or postponed until some point in the future. In
order to schedule such future actions the HCM implements a
model of time allowing proactive planning.
C. Decision Making
A core aspect of the HCM system is its reasoning system.
The HCM agent receives messages from user-level applica-
tions and probes the environment using the previously dis-
cussed managed element kernel services. Decisions on how to
treat the received messages are made locally using policies;
a set of constraint rules governing system behaviour. One of
the goals of the HCM is to provide the end user with an
easy means of authoring policies that will control the various
autonomic features. In order to modify the HCM behaviour
at runtime, these policies are dynamically loaded when the
device is running.
Policies are not coded directly within the agent behaviours.
To be more flexible, the agent uses the Policy Service managed
element, shown in Figure 1, to issue events to a policy
engine and wait for a set of recommended actions to per-
form. The particular policy server employed by the HCM is
Ponder24[8] citerussello, used as a local library, which uses
an XML-based policy description language to define events
and policies to be processed by the Ponder2 policy engine.
The result of a policy is an action the HCM agent has to
perform. Currently, the HCM deals with obligation policies.
An obligation policy is an Event Condition Action (ECA)
rule in the deontic sense [10]. Given E, and Cis true, it
is obligatory that the agent performs A.
Two of the primary features of the HCM are support for
connection-aware and usage-aware communication.
Connection-Aware Communication is the ability of the
HCM to remain aware of all available (active and inactive)
infrastructure and ad-hoc network connections, their parame-
terization and performance characteristics (e.g., Quality of Ser-
vice (QoS) characteristics). This set of autonomic behaviours
are triggered by changes in network resources, with some of
the most significant operations being:
Network handover : The capability of dynamically
switching from one network type to another when com-
municating with other devices. This decision can be
taken based on the reachability of a device over different
infrastructure or ad-hoc networks and on network avail-
abilities. For example, we can consider handovers from an
infrastructure technology (e.g., supporting UDP connec-
tions) to an ad-hoc technology (e.g., suuporting Bluetooth
connections) when communicating with a device in the
local neighbourhood.
Routing optimization : The capability of enhancing
multi-hop routing among network nodes. Parameters
which affect these decisions can be based on several QoS
parameters such as response delay, nominal and available
bandwidth between network nodes, transmission errors,
etc. For example, a self-optimization feature of the HCM
fitting into this group is the proactive evaluation of the
transmission delay between two interoperating devices
and consequently the use of an alternative path to reach
the same target node.
Other behaviours within this category include those acting in
response to failovers or high network load, etc.
Usage-Aware Communication consists of a set of autonomic
behaviours related to the usage of deployed user-level services.
For example, a group of rescue workers are on the site of
a major accident with human casualties and must constantly
maintain communication with both with one another and with
a response unit concerning their findings and the positions of
injured people. Due to the potentially disruptive nature of the
environment network availability may be intermittent, but the
goal to reliably deliver communication must persist. In this
scenario the typical autonomic decisions to be taken include:
Transmission contingency: The capability of providing
alternatives to the default means of transmitting a mes-
sage. This specifically includes making the best possible
use of multi-technology transmission paths including
cellular networks, IP infrastructure networks, satellite
systems, MANETs, etc. An important parameter in these
decisions is the importance of the message to be sent. An
example is simultaneously sending high priority messages
via two or more different technologies, and therefore
routes, to improve the chances that they are successfully
delivered to target nodes. The use of extra resources is
justified by the importance of the content to deliver.
Content adaptation: Adapting the content of a message
(or stream). These decisions can be based on several
parameters such as the number and the importance of the
messages/streams to be sent. Examples include applying
a codec to reduce the bandwidth consumed by a video
stream, or simply stripping out the audio component and
sending this in lieu of the video.
Deferred service provisioning: Waiting until a con-
nection is available before making routing decisions to
mitigate uncertainty relating to the choice of optimal
technologies or paths. Also in this case the parameters
able to trigger this type of decisions are the volume
and the significance of the information to be sent. This
includes the need to buffer messages while awaiting a
Role management: Specifying user defined conditions
which must be met before taking a particular action. A
parameter which affects these decisions is the role of the
end-user in a particular scenario. For example within a
major incident response workers are divided into response
units, structured in a hierarchical manner. In this scenario
we can envisage policies which ensure a message (e.g.,
notification of an event) is sent to the right recipients in
the role hierarchy (e.g., escalating).
All of these capabilities are controlled by user-definable
policies that may be simply and rapidly specified/modified on-
the-fly by either the user or local/remote automated routines.
One domain in which the HCM has been demonstrated
through experiment to be of considerable value is in supporting
collaboration between rescue workers at incident sites [6].
The various respondees to an incident may include police,
firefighters, paramedics, coastguards, the military and other
specialist teams. Moreover these people often only represent
the vanguard of a much larger team of personnel working
off-site including logistics support, consultants, administrators,
etc. It is critical that communication between all parties
involved in an incident remain available at all times, and that
different channels of communication be available according
to task and location constraints. Personnel working in close
proximity at the incident site may share information in a
Social Area Network relating to services such as positioning
information, map topologies, biometrical data, etc., via ad-
hoc connections, perhaps as a preference or in the absence
of a reliable infrastructure connection. The critical aspect is
the assured preservation of communication even in such poten-
tially disruptive environments. Sessions must be as continuous
as possible, providing a consistent experience by seamlessly
adapting configuration according to available access technolo-
gies, device types and locations.
An early HCM prototype has recently been a key component
of experiments involving rescue personnel operating at a major
incident site. In one particular experiment a building fire is
considered. This particular scenario demands fast and decisive
action, often in life-threatening situations. It also requires
collaboration between numerous people located in different,
often changing areas: personnel at the incident site including
fire-fighters, police, and paramedics, at the command center,
in vehicles, and others.
Each of the people involved, and many of the vehicles and
other equipment, are associated with one or more electronic
devices such as radios, biosensors, GPS, health recorders,
handhelds, tablet PC, etc. In the fire scenario different de-
vices run different crisis-relevant applications including VOIP
clients, instant messengers, map services and other collabora-
tive tools.
This scenario is illustrated by Figure 2. Two Fire-Fighters
(FF) are moving in relative proximity to one another attempt-
ing to evacuate people from a building on fire. They are
using small, wearable HCM activated PDAs and are able to
communicate with local CCTV cameras that are broadcasting
video streams of the location on several access channels. The
duty of the fire-fighters is to explore the incident site and notify
the Command Center (CC) and the other local team members
of findings related to visited building(s), i.e., the positions of
injured people. To do this, they make special marks on the
map displayed on their PDA. They may also take pictures
using the PDAs in-built camera to assist the command center
with gaining a visual overview of the incident location and
In terms of connectivity, FF1 and FF2 are connected via
both ad-hoc (AH) and infrastructure (IN) networks, and the
two CCTV cameras are capable of only broadcasting images
over a local ad-hoc connection. In this situation indicated
where certain infrastructure connections fail due to some
disruption at the incident site, fire-fighters remain able to
communicate using their ad-hoc connections while in relative
Infrastructure Radio Link
Ad-Hoc Radio Link
Fig. 2. Rescue worker scenario.
proximity to one another. If an infrastructure connection were
to become re-established the HCM may decide, according
to prescribed policies, to switch switch FF2 back to this
connection allowing him to communicate directly with the CC
rather than via FF1.
In this particular scenario, the firefighter’s devices are
equipped with the following policies:
1) IF (infrastructure connected) THEN send map data to
CC and FF via INFR.
2) IF (!infrastructure connected) THEN send map data to
CC and FF via HOC.
3) IF (infrastructure connected) THEN send high resolu-
tion pictures to CC via INFR.
4) IF (!infrastructure connected) THEN send low resolu-
tion pictures to CC via HOC.
Figure 3 shows the definition of the fourth policy statement
in this list. This policy sends low resolution pictures to the
command center when the device is not infrastructure con-
nected. It receives media status events which denote whether
a network interface is available or not. This event has two
associated parameters: the interface protocol and its current
status. The only condition that triggers this policy is that the in-
frastructure connection is unavailable (see the XML condition
element). The action sendLowResolutionPicture triggered by
this policy notifies the HCM agent to decrease the resolution
of the pictures for the command center and to send them
using the ad-hoc network (see the protocol attribute of the
sendLowResolutionPicture XML element).
In this scenario, FF1 has policies (1) and (3) activated. When
FF2 moves into an area where the infrastructure connection
fails, this is detected and policies (2) and (4) automatically
become active. The HCM running on FF2’s PDA is thus
notified by the policy engine and hands over all communica-
tion with FF2 to the available ad-hoc connection with FF1.
14 D. Greenwood and R. Ghizzioli
command center (CC) and the other local team members of findings related to
visited building(s), i.e., the positions of injured people. To do this, they make
special marks on the map displayed on their PDA. They may also take pictures to
assist the command center with gaining a visual overview of the incident location
and status.
(a) The conducted mocked-up situation. (b) Example of policy to send low resolution
pictures to the command center towards the
ad-hoc network when the infrastructure con-
nection is not working anymore.
Figure 6. Real World Scenario.
In terms of connectivity, FF1 and FF2 are connected via both ad-hoc (HOC)
and infrastructure (INFR) networks and FF3 only via an ad-hoc connection. In this
situation, through the multi-hop capabilities of RASCAL all four actors (the three
firefighters and the command center) are able to communicate one another. For
example FF3 communicates with the command center via the ad-hoc connection
with FF2.
In this particular scenario, the RASCALized devices used by the firefighters
are equipped with the following policies:
1. IF (infrastructure connected) THEN send map data to CC and FF via
2. IF (!infrastructure connected) THEN send map data to CC and FF via
3. IF (infrastructure connected) THEN send high resolution pictures
to CC via INFR.
4. IF (!infrastructure connected) THEN send low resolution pictures
to CC via HOC.
Figure 6(b) shows the definition of the final policy in the list presented above.
This policy sends low resolution pictures to the command center when the device
Fig. 3. Policy statement to send low resolution pictures to the command
center via the ad-hoc network when the infrastructure connection is no longer
Given the importance of sending images to the command
center and giving the low nominal bandwidth of the ad-hoc
channel, pictures are first automatically reduced in quality (i.e.,
resolution) before transmission.
Later, when FF2 returns to an area with infrastructure
network coverage, communications with the command center
are automatically returned to the infrastructure connection with
images once again sent in normal, high resolution.
Real world experimentation with emergency services is
ongoing with results to be published in a forthcoming paper; a
particular focus will be analysing the effectiveness of service
collaboration across the seamless hybrid network substrate. We
are also assessing a related scenario where clusters of rescue
personnel at different sites use the HCM to actively collaborate
when coordinating search and rescue activities.
This paper has provided an introduction to a Hybrid Con-
nection Manager technology based on autonomic communi-
cations software components designed to manage seamless
connectivity across infrastructure and ad-hoc networks. The
HCM can be interfaced with device communication stacks and
offers features including automated network handover, routing
optimization, transmission contingency, content adaptation,
deferred service provisioning and role management. The key
components of each local HCM deployment are a software
agent autonomic control logic and a policy engine with which
the user is able to define rules guiding and constraining the
policy-driven agent control logic.
The HCM is an early prototype technology extending an
existing product, the Living Systems Connection Agent -
winner of the 2005 GSMA award for best mobile enterprise
application through it’s incarnation as Swisscom’s Unlimited
Data Manager5. Particular ongoing work includes areas that
enhance the autonomic capabilities of the system including
the incorporation of improvements to the policy expression
language, and to the autonomic controller logic. The im-
provements to the policy language include a new means of
expression using a form of process-algebra over actions. This
allows the expression of a set of actions (i.e., workflow)
rather than the current limitation to atomic actions. Further
planned improvements relate to the inspectability of the HCM
system and its use in composing dynamic assemblies of
computational devices and services with flexible, self-adaptive
communicative connections.
[1] S. Farrell and V. Cahill Delay and Disruption Tolerant Networking. Artech
House Publishers, 2006.
[2] F. L. Bellifemine, G. Caire and D. Greenwood Developing Multi-Agent
Systems with JADE. John Wiley & Sons, 2007.
[3] J. M. Vlissides, J. O. Coplien and N. L. Kerth Pattern languages of
program design 2. Addison-Wesley Longman Publishing Co., 1996.
[4] F. Hu and S. Kumar The integration of ad hoc sensor and cellular
networks for multi-class data transmission. Elsevier Ad-Hoc Networks
Journal, 4:2, 2006, 254–282.
[5] D. Greenwood and M. Calisti The Living Systems Connection Agent:
Seamless Mobility at Work Communication in Distributed Systems (KiVS
07), 2007.
[6] M. Kristensen, M. Kyng, E. T. Nielsen IT support for healthcare
professionals acting in major incidents. 3rd Scandinavian conference on
Health Informatics, 2005, 37–41.
[7] Gamma, Erich and Helm, Richard and Johnson, Ralph and Vlissides, John
Design patterns: elements of reusable object-oriented software. Addison-
Wesley Professional, 1995.
[8] N. Damianou, N. Dulay, E. Lupu and M. Sloman The Ponder Policy
Specification Language. POLICY ’01: Proceedings of the International
Workshop on Policies for Distributed Systems and Networks, Springer-
Verlag, 2001, 18–38.
[9] G. Russello, C. Dong, and N. Dulay Authorisation and Conflict Resolution
for Hierarchical Domains. IEEE Workshop on Policies for Distributed
Systems and Networks, Bologna, Italy, 2007.
[10] R. J. Wieringa and J.-J. Ch. Meyer Applications of deontic logic in
computer science: a concise overview. Deontic logic in computer science:
normative system specification, John Wiley and Sons Ltd., 1993, 17–40.
[11] M. Hauge and O. Kure Multicast Service Availability in a Hybrid 3G-
cellular and Ad Hoc Network. International Workshop on Wireless Ad-
Hoc Networks, 2004.
[12] C. Kappler, P. Mendes, C. Prehofer, P. Poyhonen and D. Zhou A
Framework for Self-organized Network Composition. Proc. of the 1st IFIP
International Workshop on Autonomic Communication, 2004.
[13] S. van der Meer, S. Arbanowski and T. Magedanz An Approach for
a 4th Generation Messaging System. Proc. of 4th IEEE Symposium on
Computers and Communications (ISCC’99), 1999, 156–163.
[14] K. N. Choong, V. S. Kesavan, S. L. Ng, F. Carvalho, A. L. Low and
C. Maciocco SIP-based IEEE802.21 media independent handover: a BT
Intel collaboration BT Technology Journal, 2007, 25:2, 219–230.
[15] M. H. Reeve, C. Bilton, P. E. Holmes and M. Bross, Networks and
systems for BT in the 21st century, BT Technology Journal, 2005, 23:1,
[16] F. Cacace and L. Vollero, Managing mobility and adaptation in up-
coming 802.21 enabled devices, Proceedings of the 4th international
workshop on Wireless mobile applications and services on WLAN
hotspots (WMASH ’06:), 2006, 1–10.
[17] A. de la Oliva, T. Melia, A. Vidal, C.J. Bernardos, I. Soto and
Albert Banchs, IEEE 802.21 enabled mobile terminals for optimized
WLAN/3G handovers: a case study, SIGMOBILE Mobile Computing
Communications Review, 2007, 11:2, 29–40.
[18] S.B. Lee, H.J. Choi and Sang Gyoo Sim, A Multiple Authorship Model
in User Generated Content, Proceedings of the 5th IEEE Consumer
Communications and Networking Conference (CCNC), 2008, 1070-1074.
[19] J-L. Yoon, Telco 2.0: a new role and business model, IEEE Communi-
cations Magazine, 2007, 45:1, 10–12.
Full-text available
Due to the large success of wireless networks and portable devices, the pervasive computing paradigm is becoming a reality. One of the most challenging objectives to be achieved in pervasive computing environments is to allow a user to perform a task by composing on the _y the environment's service and resource components. This involves automatic matching and selection of services across various devices in the pervasive environment. Existing approaches mostly consider only functional aspects for service and component matching and do not consider various non-functional aspects such as user preferences, device capabilities in terms of software and hardware, and network heterogeneity of devices. We present an approach for dynamic selection of components and devices in a pervasive environments considering all the aforementioned aspects simultaneously. First, we provide a modeling of abstract and concrete application, device capabilities and resources, user preferences as well as modeling of the underlying connected platform. Device capabilities are represented by our extended CC/PP model and user preferences using our extended CP-net model. We model both the user task and the underlying network services, along with service requirements, user preferences and device capabilities, as graphs. The heterogeneity of communication protocols is also considered in the graph. The algorithmic aspects have been treated by providing algorithms for service and component matching, application mapping on network platform and user preference evaluation. For description of application composition extended SCA model is used. Departing from an abstract composition, we arrive on achieving a concrete application composition which may be distributed across more than one device. If during the application execution a new, better device appears, the application is recomposed to replace the existing components by the newer ones. This also implies the continuity of session from one device to another. A prototype implementation and its evaluation are also provided.
Full-text available
En raison du grand succès des réseaux sans _l et des appareils portatifs, le paradigme de l'informatique pervasive est devenu une réalité. L'un des plus di_ciles objectifs à atteindre dans de tels environnements est de permettre à l'utilisateur d'exécuter une tâche en composant à la volée, les services et les ressources de l'environnement. Cela implique la correspondance et la sélection automatique de services à travers divers dispositifs de l'environnement pervasif. Les approches existantes considèrent souvent seulement les aspects fonctionnels des services et ne prennent pas en compte diff érents aspects non-fonctionnels tels que les préférences utilisateur, les capacités des dispositifs en termes matériels et logiciels, et l'hétérogénéité du réseau de ces dispositifs. Nous présentons une approche pour la sélection dynamique des composants et des dispositifs dans un environnement pervasif en considérant simultanément tous les aspects précédemment mentionnés. Premièrement, nous proposons une modélisation abstraite et concrète de l'application, des capacités des terminaux et des ressources, des préférences des utilisateurs, ainsi que la modélisation de la plate-forme réseau sous-jacente. Les capacit és des dispositifs sont représentées par notre extension du modèle CC/PP et les préférences des utilisateurs en utilisant notre extension du modèle CP-Net. Nous mod- élisons sous forme d'un graphe la tâche de l'utilisateur et des services réseau sous-jacent, ainsi que les exigences des services, des préférences utilisateur et les capacités des dispositifs. L'hétérogénéité des protocoles de communication est également considérée dans les graphes. Les aspects algorithmiques ont été traités en fournissant des algorithmes pour la correspondance entre les services et les composants, pour la projection des applications sur la plate-forme de composants existants et pour l'évaluation des préférences utilisateurs. Pour la description de la composition de l'application nous proposons un modèle SCA étendu. Partant d'une composition abstraite de services, nous arrivons à réaliser une composition concrète de l'application distribuée à travers les dispositifs existants. Si pendant l'exécution un nouveau meilleur dispositif apparaît, l'application est recomposée en tenant compte des nouveaux composants. Cela permet de réaliser la continuité de la session d'un dispositif vers un autre. Une mise en oeuvre d'un prototype et son évaluation sont également fournis.
Full-text available
One emerging characteristic of electronic devices is the increasing number of connectivity interfaces (aka NICs1) towards the outside world. That obviously translates in a set of technical issues related to their management in order to provide seamless connectivity when the connections move from one interface to another. The IEEE 802.21 is a recent effort of IEEE that aims at providing a general interface for the management of NICs. In this paper we discuss how the upcoming standard may be effectively exploited in a mobile context in order to hide network heterogeneity to end users. To accomplish this task, we propose a centralized element called Mobility Manager interfacing with the 802.21 sublayer and responsible for the application of connectivity policies. Based on a real testbed, we showed that the new standard and the MM can be used to improve network performance experienced by the end user. Moreover we showed how the MM can interact with adaptive applications in order to improve further the range of usability of real-time applications.
Full-text available
This paper focuses on development of it support for health-care professionals acting in major incidents. We introduce the participatory design approach as adequate for analysis, de-sign and development of technologies for use in complex envi-ronments and situations, and describe the actual participatory design efforts. Then we present our analysis of the need for it-support in relation to major incidents. The analysis concen-trates on four main problem areas. These are: a) Use of bio-monitors, b) Person identification and registration of data, c) Communication and d) Overview. We then describe the Blue-Bio biomonitor prototype, a wireless multifunction biomoni-tor. BlueBio data can be accessed by the healthcare profes-sionals independent of where they are located and displayed on different types of devices tailored to the needs of the indi-vidual professional. Finally we discuss some challenges re-garding the prototype development process.
Learn how to employ JADE to build multi-agent systems! JADE (Java Agent DEvelopment framework) is a middleware for the development of applications, both in the mobile and fixed environment, based on the Peer-to-Peer intelligent autonomous agent approach. JADE enables developers to implement and deploy multi-agent systems, including agents running on wireless networks and limited-resource devices. Developing Multi-Agent Systems with JADE is a practical guide to using JADE. The text will give an introduction to agent technologies and the JADE Platform, before proceeding to give a comprehensive guide to programming with JADE. Basic features such as creating agents, agent tasks, agent communication, agent discovery and GUIs are covered, as well as more advanced features including ontologies and content languages, complex behaviours, interaction protocols, agent mobility, and the in-process interface. Issues such as JADE internals, running JADE agents on mobile devices, deploying a fault tolerant JADE platform, and main add-ons are also covered in depth. Developing Multi-Agent Systems with JADE: • Comprehensive guide to using JADE to build multi-agent systems and agent orientated programming. • Describes and explains ontologies and content language, interaction protocols and complex behaviour. • Includes material on persistence, security and a semantics framework. • Contains numerous examples, problems, and illustrations to enhance learning. • Presents a case study demonstrating the use of JADE in practice. • Offers an accompanying website with additional learning resources such as sample code, exercises and PPT-slides. This invaluable resource will provide multi-agent systems practitioners, programmers working in the software industry with an interest on multi-agent systems as well as final year undergraduate and postgraduate students in CS and advanced networking and telecoms courses with a comprehensive guide to using JADE to employ multi agent systems. With contributions from experts in JADE and multi agent technology.
In this paper, we look at various mobility management protocols and handover frameworks in use in BT. We also report on the results from a collaborative proof of concept mixed network prototype for a seamless handover system using Intel’s early implementation of the IEEE802.21 media-independent handover standard. The paper reports the design, implementation and deployment issues/options of a handover mechanism using the Intel framework which includes a connection manager and mixed network IEEE802.21 adaptation layer together with BT’s SIP-based audio/video application in a heterogeneous Wi-Fi, WiMAX and Ethernet network environment.
This paper provides an overview of BT’s 21st century network architecture. Analysis shows that this will deliver an improved customer experience, faster service provision and a significant overall cost reduction on today’s base. These aims can only be achieved by a ground-breaking approach which moves from multiple networks each providing one service to a true multiservice network where the services are delivered by combining reusable capabilities. The architecture is therefore based on network, intelligence and operational support systems built from reusable components and an application layer that can expose these capabilities through open interfaces.
Having expertise is one thing; conveying it to others is quite another. Patterns do just that." Patterns are a literary form with roots in literate programming, in a design movement of the same name in contemporary architecture, and in the practices common to the ageless literature of any culture. This volume, with contributions from the biggest names in the patterns community, is the second in a series documenting patterns for professional software developers. These patterns capture solutions to a plethora of recurring problems in software design and development, including language-specific patterns and idioms; general- and special-purpose patterns; architectural patterns; process and organizational patterns; expositional patterns; and patterns for concurrent programming, distributed systems, and reactive systems. This new collection not only reveals secrets of great software professionals but also makes those secrets easy to apply to your own work.
Proceedings of the first conference on Pattern Languages of Programs