ID-Services: An RFID Middleware Architecture for
Mobile Applications
Joachim Schwieren and Gottfried Vossen
European Research Center for Information Systems (ERCIS)
University of Münster, Leonardo-Campus 3, 48149 Münster, Germany
Abstract. The use of RFID middleware to support application development for
and integration of RFID hardware into information systems has become quite
common in RFID applications where reader devices remain stationary, which
currently represents the largest part of all RFID applications in use. Another
field for applying RFID technology offering a huge set of novel possibilities
and applications are mobile applications, where readers are no longer fixed. In
order to address the specific issues of mobile RFID-enabled applications and to
support developers in rapid application development, we present the architec-
ture of an RFID middleware for mobile applications. The approach has been
used to implement MoVIS (Mobile Visitor Information System), a mobile ap-
plication which allows museum visitors to request individually adapted multi-
media information about exhibits in an intuitive way.
1 Introduction
Many information systems are being used to manage objects or processes from the
“real” physical world (e.g., products, shipping goods or other physical resources).
The gap that exists between the physical world and its domain-specific model in the
information system which is also referred to as “physical digital divide” [13], needs to
be bridged by manual user interaction. Apart from that, auto-ID technologies like
barcodes, smart-cards and RFID (Radio Frequency Identification) can help to inte-
grate information systems very closely with the physical world and to align them with
physical processes. This reduces manual interaction and helps to increase reliability,
to speed up processes and to reduce costs. Especially RFID has emerged tremendous-
ly in the last years, mainly due to lower prices and a higher level of maturity of the
available technology [8]. However, the actual success of RFID is based on its supe-
rior characteristics: it allows a contact-less and robust identification without requiring
a line of sight. This works from a few centimeters up to several meters, depending on
the specific RFID technology used. These characteristics are offered by no other auto-
ID technology. Apart from very specialized, highly integrated applications (e.g., car
immobilizers, contactless micro-payment systems, access control systems) where
RFID is already widely established, most effort is currently put on implementing
RFID in business contexts (e.g., supply chain automation, tracking and tracing of
shipping goods, etc.) [2]. But also industry can benefit from RFID technology in
Schwieren J. and Vossen G. (2008).
ID-Services: An RFID Middleware Architecture for Mobile Applications.
In Proceedings of the 2nd International Workshop on RFID Technology - Concepts, Applications, Challenges, pages 19-31
DOI: 10.5220/0001728500190031
Copyright
c
SciTePress
order to extend the level of automation in production and manufacturing processes
[10]. Even though these RFID applications originate from completely different fields,
most applications have in common that the reader device is stationary whereas the
transponders are mobile. In this paper we present a framework for RFID application
development for the opposite case of mobile readers that opens a new branch of RFID
uses which we refer to as “mobile RFID”.
Indeed, a rather novel field for applying RFID technology is mobile applications
where the RFID reader device itself is a mobile unit (usually combined with a mobile
computing device) [19, 20]. Since today’s mobile computing devices such as PDAs
(Personal Digital Assistant), UMPCs (Ultra Mobile PC) or mobile phones have a
relatively strong computing power and offer many functionalities, the combination
with RFID offers a huge set of new possibilities and applications. Because mobile
applications usually have a higher affinity to the physical world in which they are
used, the need for a close integration into the physical world is very high. RFID can
be used to directly interact with physical objects which are equipped with transpond-
ers. This can be used to reduce the complexity of user interfaces and can help to
speed up and align processes. Examples for mobile RFID applications are for exam-
ple maintenance applications [18] where objects like e.g. fire extinguishers need to be
maintained from time to time. However, also field force automation applications that
help field staff to quickly access information about product samples they carry with
them or keeping track of field service activities are examples of mobile RFID applica-
tions. Also in the healthcare and medical context, mobile RFID applications can be
used to identify patients or medicine to prevent errors in treatment [4] and to speed up
documentation processes [27]. Another application is tracking and documenting sen-
try patrols at fixed physical locations. Finally, a more consumer-oriented use case is
mobile guide systems like MoVIS [24], which allows museum visitors to request
information about exhibits in a very intuitive way by simply touching RFID trans-
ponders that are attached to exhibits. And yet another possible application of mobile
RFID is using PDAs for in-field e-learning applications that for example teach engi-
neers about the different components of a complex machine while standing in front of
it.
Though there are many possible use cases for mobile RFID applications, these ap-
plications are still not very widely spread. A major reason for this is the fact that
developing such applications is not a trivial task: The available hardware is very hete-
rogeneous and many different standards and technologies exist. Since the integration
of the specific RFID hardware is in the main focus, the development is often very
hardware-oriented. This means that developers have to take care of both, integrating
the hardware, modeling the physical context and implementing the application’s
business logic. While developing MoVIS, we have encountered several challenges
and general requirements that can also be applied to other mobile RFID applications
like identifier resolution, information retrieval on physical objects or tracking physi-
cal interactions. Therefore we have developed an architecture of an RFID middleware
for mobile applications called ID-Services. ID-Services has been used to implement a
production version of MoVIS after a first prototype version has been developed to
evaluate the general concept of this application.
The remainder of this paper is organized as follows: In Section 2 we cover related
work and derive the motivation for our approach from that. We then present the ID-
20
Services architecture in Section 3. Section 4 discusses a particular use case for ID-
Services, and Section 5 concludes the paper.
2 Related Work and Motivation
As stated before, there are many possible use cases for mobile RFID applications.
However, it is difficult to develop such applications since the integration of the RFID
hardware often requires a very low-level interaction. Also, the developer has to model
the physical context that bridges the physical world with an information system. In
the context of stationary RFID applications a special component in the system’s ar-
chitecture is used which is often referred to as RFID middleware [11]. In contrast to
traditional middleware (like for example CORBA
1
), this RFID middleware is seen as
a mediator between the RFID reader hardware and the underlying information sys-
tems. There is no exact definition of what an RFID middleware is and where it is
located in an application’s architecture [5]; some RFID reader devices even include a
middleware in the reader’s embedded controller unit.
The idea of using a special component within a system’s architecture to mediate
between RFID hardware and other components is very common. The term RFID
middleware has been used to describe such a component. All major DBMS and ERP
vendors provide products that they refer to as RFID Middleware [17]. Even though
this term is frequently used in the literature and also in commercial products, there is
no exact definition of what an RFID middleware is and what exact functionality it is
expected to have [5]. Furthermore the term RFID middleware is almost exclusively
used in the context of RFID applications where the RFID reader units are stationary.
Most attention is put on typical problems of these applications such as handling large
amounts of data or detecting and handling read errors. This especially applies to ap-
plications in the context of the EPCglobal network
2
. An RFID middleware compo-
nent named “Savant” has also been part of the initial EPCglobal architecture specifi-
cation [9]. Finally EPCglobal has decided to leave the implementation of the middle-
ware up to software vendors and to only specify the interface (ALE – Application
Level Events) between the RFID middleware and business application. Several im-
plementations of this specification already exist [6]. While the attention on RFID
middleware is almost completely focused on stationary applications, it has only been
regarded very little in the mobile context [28]. However, especially for mobile appli-
cations a mobile RFID middleware could help to make the development process of
mobile RFID-enabled application easier and more efficient. When dealing with mid-
dleware approaches in the mobile context, the focus is mostly put on different aspects
of ubiquitous computing applications such as context modeling, context detection, or
data and network connectivity management but not on integrating RFID in order to
create a direct bridge to the physical world [7, 25, 23].
In the field of mobile RFID applications, NFC (Near Field Communication) has
gained some attention over the last years [1]. NFC is a standard released by the NFC
1
CORBA (Common Object Request Broker Architecture)
2
EPCglobal Inc.: http://www.epcglobalinc.org
21
Forum
3
for hardware, protocols, and partly also software components to primary
enable mobile phones with RFID capabilities. Most NFC standards have either been
ratified as ISO standards or cover existing ISO standards
4
. Since NFC offers different
sub features like peer-to-peer communication, smart-card emulation and basic RFID
reader capabilities for accessing certain 13.5 MHz high-frequency transponders, there
are many possible use cases for NFC. Most use cases focus on applications that sup-
port transactions like micro-payment or electronic ticketing. But also applications like
“smart posters” where the URL to a web page is encoded into a passive RFID trans-
ponder which is embedded into public posters [21]. NFC seems to be a promising
solution for mobile RFID applications but currently lacks support in both services
provided and hardware. Also only very few mobile phone handsets are available so
far that support NFC. For application development only a basic SDK from Nokia is
available [22]. The NFC SDK is mainly focused on low-level support of the develop-
er. High-level concepts are only supported partly and need to be explicitly addressed.
NFC is a standard that focuses on mobile RFID applications but cannot be regarded
as middleware that supports a developer with ready-to-use high-level functionality.
The idea of providing a framework that allows a rapid application development
process (RAD) for RFID applications has already been proposed so far [14]. Since
mobile applications are getting more important, a RAD approach is also reasonable
here. A middleware that addresses both data management and services and offers out-
of-the-box functionality can help to facilitate this approach. A similar idea is also
presented by Wu et al. [28]. They propose a service-oriented architecture that gives
developers the possibility of constructing a customized mobile OSGi
5
-compliant
RFID middleware to support different combinations of requirements. Apart from that
a fully fledged RFID middleware for mobile applications is not yet available.
In order to offer a better support for developing mobile RFID applications, we pro-
pose an architecture for a mobile RFID middleware in this paper. While RFID mid-
dleware in the stationary context often has to handle large volumes of data and is
mainly used for supporting data collection processes, in a mobile scenario the re-
quirements are different: In most cases information about physical objects that is
stored in a database either within the mobile device itself or on a remote database
server needs to be retrieved upon selecting a physical object by its RFID transponder.
Also in some cases the transponder has the ability to store data on the chip. Apart
from that not only data aspects need to be addressed; in many cases it is desirable to
also map functionality which physical objects advertise. These can be referred to as
“physical services”. Finally, certain events like reading a transponder need to be cap-
tured and processed further in order to facilitate process integration. Other general
aspects of mobile information systems are connectivity and data management. A
mobile RFID middleware will also have to address these. Some mobile applications
serve as thin clients and are connected to a remote server via networks like WLAN,
GSM, etc. Other applications work completely offline while there are also many hy-
brid variants that cache or replicate data in order to work also in temporary offline
situations.
3
The NFC Forum: http://www.nfc-forum.org
4
The NFC Forum Specifications: http://www.nfc-forum.org/specs/
5
OSGi (Open Services Gateway initiative)
22
These requirements have been taken into account when designing our ID-Services
middleware. Apart from hardware abstraction a main concept of ID-Services are
proxy objects that represent a placeholder for a physical object in the virtual world of
the mobile information system that can be accessed in the same way by the developer
as objects known from OOP (object oriented programming). The developer does not
(necessarily) have to care about RFID specific characteristics like transponder IDs or
how to access the reader hardware – he can fully focus on the development of the
actual business logic of the application. The idea is to provide a flexible RFID mid-
dleware for mobile applications that allows a rapid application development (RAD)
by mapping data, services and process related events to proxy objects that can be used
by developers in a convenient and easy way.
3 ID-Services Architecture
ID-Services is an architecture of an RFID middleware for mobile applications. Basi-
cally it is not bound to a specific hardware or an operating system platform. We have
chosen the Microsoft .NET-Compact-Framework
6
for a reference implementation
(ID-Services.NET) because applications written for this platform can be executed on
multiple devices that use the Windows Mobile or Windows CE
7
operating system, no
matter what processor is actually used [29]. Furthermore developing applications for
this platform is well supported and a lot of components and functionality like data-
base systems, replication, access to web-services, etc. are provided by the .NET-CF.
Furthermore many mobile applications in the industry and business context today are
Windows CE-based.
The main idea of the ID-Services middleware is to support the developer as much
as possible when implementing an RFID-enabled application but also to still offer a
maximum of flexibility. Therefore we have chosen a three-layer-approach which
offers at each layer access to RFID-specific functionality based on different levels of
abstraction (see Figure 1). Each layer has a core concept on which the main focus is
put on. While the first layer is dedicated to hardware abstraction, the second layer
offers a seamless integration of data functions and processes by providing “proxy
objects”. Finally the third layer provides access to high level concepts such as “physi-
cal hyperlinks” [24], “physical services” and process integration. Developers can
choose with which layer they want to interact and to what extend, depending on the
type of application being developed.
The (lowest) first layer (ID Device Layer) represents a hardware abstraction layer
that allows the developer to access the RFID reader device over a standardized inter-
face. This layer offers all features and functions that can be expected from a “concep-
tional” mobile RFID device in a standardized form. These include basic functions like
reading or writing transponder IDs or accessing the transponder’s build-in memory
on a raw level (if available). Because not all reader devices offer the same functio-
6
Microsoft .NET Compact Framework:
http://msdn2.microsoft.com/en-us/netframework/aa497273.aspx
7
Microsoft Windows CE: http://www.microsoft.com/windows/embedded/default.mspx
23
nality and not all transponders have features such a writeable memory, the ID Device
Layer also offers an interface to query the reader’s and the transponder’s capabilities.
Usually the device capabilities are known in advance to the developer but this ap-
proach allows writing a single piece of software that supports multiple hardware
devices with different capabilities and features. Also operating the reader device is
done by the ID Device Layer: A reader can either be used in a user-triggered mode
which means that the user of an application activates the reader to select a transpond-
er for example by pressing button or it can be used in a “polling mode” where the
reader continuously scans for transponders within range and reacts as soon a trans-
ponder has been detected. This type of operation mode is especially useful when the
intention of the application is to provide an intuitive physical user-interface that is
touching a transponder with the mobile device [26]. In order to support a specific
RFID device, the driver or API that is supplied by the device manufacturer, is ac-
cessed by a hardware-specific ID-Services-Connector components which maps the
functionality of the specific hardware to the ID Device Layer. This abstraction allows
using different RFID hardware devices without any changes to the application or
other parts of the middleware. Only the ID-Services-Connector needs to be modified
or re-implemented. Another advantage of this approach is that an RFID device does
not necessarily have to be a “real” hardware device. A reader device can also be com-
pletely implemented in software (for example the “DummyReader” that is part of the
ID-Services.NET reference implementation). Such virtual RFID devices can support
(automated) testing scenarios or may be helpful during the development phase when
the actual hardware is not available to all developers, because mostly mobile applica-
tions are being developed using an emulator instead of the actual hardware.
If developers want to implement the required, RFID-specific functionality them-
selves, this can be done by accessing the various features of the ID Device Layer. So
far ID-Services offers a hardware abstraction that allows access to RFID hardware
and to perform rather low-level RFID specific tasks. If the developer needs more
support when interacting with physical objects, the second layer which is called Phys-
ical Object Layer offers further-reaching functionality.
The core concept of the Physical Object Layer is “proxy objects”. Physical objects
from the real world have characteristics that need to be mapped with the data model
and functions of the information system. The idea of proxy objects is to keep all
RFID-specific aspects such as transponder IDs away from the developer and to pro-
vide a placeholder for a physical object in the information system that can be directly
accessed and modified in the same way as “ordinary” objects known from object
oriented programming. The idea of mapping database entities into objects is not new
and is a well known and proved concept used in many applications. In most cases an
OR-Mapper (object-relational mapper) is used to fulfill this task [3]. But the idea of
proxy objects goes beyond this approach. While typical meta-data like the name of
the physical object is stored in the proxy objects properties, it also offers support for
methods and events. Methods are used to advertise services that are provided by
physical objects (referred to as “physical services”) and events are used to allow the
application developer to perform certain tasks during the interaction with the physical
objects. While the properties and the methods of a proxy object depend on the physi-
cal object which the proxy object represents, the events that can be fired by a proxy
object are always the same: OnCreated(), OnModified() and OnCommit()
24
The OnCreated() event is fired when a proxy object is created. The only way to
“create” a proxy object at runtime is to scan the RFID transponder which is associated
with the physical object. As a result this event can be used for logging or tracking
purposes. If a property of the proxy object is modified the OnModified() event is
fired. Finally the OnCommit() event gets fired when a proxy object has been mod-
ified and its properties are now being made persistent.
Fig. 1. Three-level-architecture of the ID-Services RFID middleware.
Proxy objects can either represent real, physical objects or “virtual objects” that
can be used to describe fixed locations or transponders without any object affiliation.
The basic process of interacting with proxy objects is described in the following:
Before a physical object can be used in the application, an RFID transponder needs to
be attached to it. This transponder needs to have a unique identifier which is referred
to as Tag-ID. The level of uniqueness of this identifier depends on the specific appli-
cation. Applications like those in the EPCglobal network need a world-wide unique
identifier
8
that also fulfills certain criterions like a scalable resolution process [15]
while in a closed-loop scenario, an application-wide uniqueness of the identifier is
sufficient. Since the idea of the Physical Object Layer is to hide all low-level RFID-
specific aspects from the developer, there is no need to enter the Tag-ID into the
system. ID-Services also provides a functionality to register new transponders into the
system. The process of registration takes the Tag-ID and maps it to a proxy object. A
proxy object is created by the Proxy Object Factory which is closely integrated with
8
EPC (Electronic Product Code)
25
the OR- and service-mapper. The source of the data that is mapped with the proxy
object can either be a local database or any other kind of (remote) data source (for
example a data-centered web service). It is also possible that data which is mapped to
a property is originated from the transponder’s internal memory which makes ad-
dressing the transponders memory very easy. To be able to fulfill the mapping task,
the Proxy Object Factory uses meta-data that has been provided by the developer in
advance. This meta-data contains the Tag-ID, the data source that is used for the
mapping and the value of a primary key filed that refers the object’s data. Also fur-
ther mapping information like the names and data types of the mapped properties that
are exposed by the proxy objects are provided. Additionally to the mapping of the
properties, also methods can be mapped to the proxy object. Therefore either locally
implemented method calls or remote web services are referenced in the proxy object
meta-data. The parameters of these methods represent the context of a method invoca-
tion and are filled with values upon invocation. Typical context elements are the user
(-name) that uses the application and the point of time when the service is invoked.
Parameter values that have been queried from the user over the applications GUI can
also be used.
In order to speed up the process of modeling proxy objects a concept similar to the
“prototypes” from the experimental object oriented programming language SELF [12]
is used. Prototypes can be regarded as class definitions. But in contrast to traditional
OOP programming models, no inheritance is available. Instead a copy of an existing
prototype is created and modified according to the developers needs (for example by
adding or removing properties). We use the term “template” instead of “prototype” in
this context. Templates can also be merged to quickly add new properties or methods
to a new object definition. This very flexible concept allows a quick creation of proxy
objects and helps to speed up the OR- and service mapping process. In order to ana-
lyze the structure of a proxy object at runtime, a reflection mechanism can be used, if
required.
Before the Proxy Object Factory can create a proxy object out of a template and
the object’s data from the database, a resolution service maps the Tag-ID with the
appropriate proxy object meta-data. This resolution service can either be a simple
table look-up on a locally available database (which is usually replicated in order to
propagate changes back to the backend-system) or a remote resolution service (for
example a service similar to an ONS (Object Naming Service) like proposed in to the
EPCglobal architecture). The resolution service is a pluggable component in the
architecture that can also be implemented by the developer to support different kinds
of resolution types.
After the proxy object has been created, it can be used by the application in the
same way as “traditional” objects know from OOP. All data management and the
invocation of methods are controlled by the OR- and service-mapper. The whole
process which consists of the Tag-ID resolution, the proxy object creation and finally
the use of the proxy object is shown in Figure 2.
Finally the idea of the third level (High Level Concepts Layer) which has not yet
been implemented so far in the reference implementation is to provide access to typi-
cal high-level concepts of mobile, RFID-enabled applications such as “physical
hyperlinks”, “physical services” and a seamless integration of RFID-functionality into
mobile (business) processes. While interaction with proxy objects still requires the
26
Fig. 2. Resolution of Tag IDs into proxy objects that represent physical objects.
developer to implement the desired features himself, the High Level Concepts Layer
is supposed to provide ready-to-use functionality for certain applications. In case of
physical hyperlinks a local or remote resource (for example a web page) is refe-
renced, which needs to be retrieved in order to be presented to the user. Physical
services might need additional parameters from the user in order to be invoked.
Another aspect regards the fact that an RFID transponder might advertise more than
just one physical hyperlink or service. This leads to the conclusion that the user will
have to use some kind of selection mechanism, for example a menu on the mobile
computing device. Currently we investigate what characteristic functionality of mo-
bile RFID applications can be generalized in an appropriate way to be used for other
application types. A major requirement is that this functionality has to be “compact”
enough to be provided in a ready-to-use way by the ID-Servcies Middleware.
4 Use Case and Implementation
ID-Services can be used for many types of mobile, RFID-enabled applications. Those
applications that make use of high-level concepts such as physical hyperlinks, physi-
cal services or that require a close integration of physical objects into business
processes derive most value from the use of this middleware architecture. An applica-
tion that we have developed in cooperation with Elatec GmbH, a German RFID- and
electronics company, is MoVIS (Mobile Visitor Information System) [24], a PDA-
based multimedia guide system for museum visitors. By touching RFID tags that are
27
attached to or near the exhibits, a visitor can access individually adapted multimedia
information (like text, speech, images, video and even interactive content such as
quizzes) with a PDA that has a 125KHz LF RFID reader integrated. The basic idea of
MoVIS is to use RFID transponders as an intuitive extension of the user interface in
order to allow an easy access to information about museum exhibits. With MoVIS a
visitor can explore the museum on his own initiative or do a guided tour. The trans-
ponders are either used to identify exhibits in the explorative mode or to verify exhi-
bits in the tour mode.
Fig. 3. Architecture of the Mobile Visitor Information System.
The application and the content are stored locally on the mobile device. This “of-
fline approach” has been chosen to make the system more stable and to lower the
systems requirements (for example a WLAN network is not necessary). The content
is replicated on the device from a central database server using the RDA (Remote
Data Access) replication mechanism of SQL Server CE (see Figure 3). A standalone
smart-client application that connects to the server via .NET-Remoting is used by the
museum staff to add and maintain content and to analyze the reports that are being
generated by the system from the anonymously collected tracking data of the visitors.
The tracking data is generated by the code following the OnCreated() events of
the exhibit proxy objects.
The use of ID-Services highly simplifies the use of RFID in this case because the
application developer does not have to deal with low-level concepts such as Tag-IDs
etc. – he can fully concentrate on working with objects such as “exhibits”, “rooms”
and the actual content. But also the museum staff can profit from this, because com-
mon tasks such as resetting a device after it has been returned by a visitor or synchro-
nizing data with the backend server have been made available through physical ser-
vices. By touching special transponders, which are not available to visitors, the de-
sired maintenance service is invoked which makes them very easy to execute without
28
any further interaction on the PDA’s screen. These “service tags” are implemented as
virtual proxy objects. But also adding new exhibits (transponders) to the system is
simple because it is supported by ID-Services. By using another “service tag”, the
device can be put into a mode that allows registering new transponders. The new
transponder is simply attached to the exhibit and then scanned with the PDA’s reader
device. The museum operator can give the transponder a speaking name and after
synchronizing with the backend server, the newly registered transponder can be asso-
ciated with an exhibit or content. Both the museum operator and the developer of the
application will never see the ID which is stored on the ready-only transponders that
are used for this application. The whole ID resolution is done internally by ID-
Services. MoVIS is currently being used and evaluated at the “Mühlenhof” open-air
museum in Münster, Germany to inform visitors about the exhibits and their historical
background.
5 Summary and Conclusions
The use of RFID middleware as a central component in RFID-enabled applications
has become common for most applications where stationary readers are involved. An
RFID middleware is usually used to abstract from the actual hardware and to provide
RFID-specific functionality such as cleaning or aggregating RFID data, as well as
handling massive data-streams and to offer a close integration into information sys-
tems. In the case of mobile RFID applications, such a kind of middleware has not
been used so far. Therefore developers have to directly access the hardware over the
drivers and APIs that are supplied by the device manufacturers. This makes applica-
tion development a very hardware-centered task. Even though mobile RFID applica-
tions do have different requirements in comparison to “stationary” RFID applications,
the concept of a middleware component that mediates between the hardware and the
actual application can also be applied here, as well. While a basic hardware abstrac-
tion fosters compatibility and portability and also allows an easier testing process,
typical high-level concepts and functionality such as mapping physical objects with
entities from databases or services can also be supported.
Our approach allows an easy, yet flexible access to RFID-specific functionality in
the context of mobile applications, depending on the individual level of support that
is needed by the developer for a specific application. A developer can choose to use
very basic RFID functionality such as reading a transponder’s ID or a transponder’s
memory storage with only a few lines of code over a standardized API or he can
make use of ready-to-use high-level functionality. This includes concepts like physi-
cal hyperlinks or physical services as well as common ID management and resolution
tasks. The flexibility is achieved by separating the functionality into three bottom-up-
stacked layers that can be accessed according to the developers needs.
The conceptual architecture of ID-Services is basically platform-neutral and can be
used for any kind of mobile computing device. The reference implementation ID-
Services.NET which makes use of several built-in features of the .NET Compact
Framework and also the SQL Server CE database engine has been used to implement
MoVIS, an RFID-based, mobile visitor information system for museums. Many other
29
applications from completely different domains use similar high-level concepts which
are natively supported by ID-Services. An issue that has not been addressed so far is
security and privacy. Some applications will require ways to protect data that is stored
on the transponders or prevent that data which is stored on the mobile device or the
network is being accessed by non-authorized users [16]. The further development of
ID-Services will also focus on this issue, as well as the ability to allow a seamless
integration into existing mobile (business) processes which will be a part of the High
Level Concepts Layer.
References
1. ABI Research: “Near Field Communication (NFC): Leveraging Contactless for Mobile
Payments, Content and Access”, 2007
2. ABI research: RFID Annual Market Overview, 2007, http://
www.abiresearch.com/products/ market_research/RFID_Annual_Market_Overview
3. Ambler, W.S.; Mapping Objects to Relational Databases: O/R Mapping In Detail, 2006,
http://www.agiledata.org/essays/mappingObjects.html
4. Anshel, M. & Levitan, S. Reducing medical errors using secure RFID technology SIGCSE
Bull., ACM, 2007, 39, 157-159
5. Burnell, J. What Is RFID Middleware and Where Is It Needed? RFIDupdate.com, 2006
6. Chen, N.; Chang, T.; Chen, J.; Wu, C. & Tzeng, H. Reliable ALE Middleware for RFID
Network Applications CSREA EEE, 2007, 183-189
7. Dabkowski, A. XML-Based Middleware for Mobile Systems Berliner XML Tage, 2003,
432-438
8. Dortch, M., Aberdeen Group, Inc.: "Winning RFID Strategies for 2008", 2008,
http://www.aberdeen.com/summary/report/benchmark/4205-RA-winning-rfid-strats.asp
9. EPCglobal, "The Application Level Events (ALE) Specification, Version 1.0," Specifica-
tion, February 8, 2005
10. Finkenzeller, K. RFID Handbook: Fundamentals and Applications in Contactless Smart
Cards and Identification John Wiley & Sons, Inc., 2003
11. Floerkemeier, C. & Lampe, M. RFID middleware design - addressing both application
needs and RFID constraints GI Jahrestagung (1), 2005, 277-281
12. Hölzle, U. & Ungar, D. Reconciling responsiveness with performance in pure object-
oriented languages ACM Trans. Program. Lang. Syst., ACM, 1996, 18, 355-400
13. Jeffery, S.R.; Alonso, G.; Franklin, M.J.; Hong, W.; Widom, J. Virtual devices: an extensi-
ble architecture for bridging the physical-digital divide. Tech. Rep. UCB-CS-05-1375, UC
Berkeley CS Division, 2005
14. Kim, Y.; Moon, M. & Yeom, K. A Framework for Rapid Development of RFID Applica-
tions Computational Science and Its Applications - ICCSA 2006, 2006, 226-235
15. Kindberg, T. Implementing physical hyperlinks using ubiquitous identifier resolution
WWW '02: Proceedings of the 11th international conference on World Wide Web, ACM,
2002, 191-199
16. Konidala, D. & Kim, K. Mobile RFID Applications and Security Challenges Information
Security and Cryptology ICISC 2006, 2006, 194-205
17. Leaver, S. C.; Mendelsohn, T.; Spivey Overby, C. & Yuen, E. H. Evaluating RFID Mid-
dleware Forrester Research, Inc., 2004
18. Legner, C. & Thiesse, F. RFID-Based Facility Maintenance at Frankfurt Airport IEEE
Pervasive Computing, IEEE Computer Society, 2006, 5, 34-39
30
19. Liu, S.; Wang, F. & Liu, P. Integrated RFID data modeling: an approach for querying
physical objects in pervasive computing CIKM '06: Proceedings of the 15th ACM interna-
tional conference on Information and knowledge management, ACM, 2006, 822-823
20. Nath, B.; Reynolds, F. & Want, R. RFID Technology and Applications IEEE Pervasive
Computing, IEEE Educational Activities Department, 2006, 5, 22
21. NFC Forum: NFC Smart Poster Record Type Definition, SPR 1.1, NFCForum-
SmartPoster_RTD_1.0, 2006
22. Nokia 6131 NFC SDK 1.1, http://www.forum.nokia.com/info/sw.nokia.com/id/ef4e1bc9-
d220-400c-a41d-b3d56349e984/Nokia_6131_NFC_SDK.html
23. Park, S.; Kim, D. & Kang, B. Context-aware Middleware Architecture for Intelligent Ser-
vice in Mobile Environment CIT '06: Proceedings of the Sixth IEEE International Confe-
rence on Computer and Information Technology (CIT'06), IEEE Computer Society, 2006,
240
24. Schwieren, J. & Vossen, G. Implementing Physical Hyperlinks for Mobile Applications
using RFID Tags. IDEAS 2007, 154-162
25. Son, M.; Kim, J.; Shin, D. & Shin, D. Research on Smart Multi-agent Middleware for
RFID-Based Ubiquitous Computing Environment Agent Computing and Multi-Agent Sys-
tems, 2006, 787-792
26. Välkkynen, P.; Korhonen, I.; Plomp, J.; Tuomisto, T.; Cluitmans, L.; Ailisto, H. & Seppä,
H. A user interaction paradigm for physical browsing and near-object control based on tags
2003
27. Wang, S.; Chen, W.; Ong, C.; Liu, L. & Chuang, Y. RFID Application in Hospitals: A
Case Study on a Demonstration RFID Project in a Taiwan Hospital hicss, IEEE Computer
Society, 2006, 8, 184a
28. Wu, J.; Wang, D. & Sheng, H. Design an OSGi Extension Service for Mobile RFID Appli-
cations icebe, IEEE Computer Society, 2007, 0, 323-326
29. Yao, P.: Microsoft .NET Compact Framework for Windows CE .NET, July 2002. http://
msdn2.microsoft.com/en-us/library/ms836805.aspx
31