
2.4 Realization of other information
systems
In the frame of an elaborated programming lecture at
our University the topic of J2ME was introduced
(Weghorn, 2003). The students could select as as-
sessment work one implementation project out of a
defined list of information systems, which are con-
structed like the one, which is described in detail
above. In total twenty teams of two students were
funded, and they were and they still are doing a de-
velopment of information systems on car traffic
channels, public transportation, railway connections,
and skiing arenas in the Alps.
The student teams are implementing both parts
of the information system – the terminal software,
and the central service agent. Of course, since this
work is part of a learning lesson, the results are not
all prefect. Only a few teams came close to the
above described optimal system concept. But in the
end, in sum a series of information systems are im-
plemented, which can be useful for various exem-
plary situations.
3 SIMULATION AND REAL
WORLD DEVICES
Due to budget limitations at our University, wireless
tools for digital phone networks cannot broadly be
developed and tested in the target environment. This
is also not really required, because simulation envi-
ronments are widely available for developing and
testing in particular J2ME applications. One simula-
tor is supplied within the wireless toolkit from the
company SUN Microsystems (accessible from
http://java.sun.org/j2me). If the developer aims for a
specific target device, all the big phone manufactur-
ers (Motorola, Nokia, Siemens, …) operate WEB
sites for developers, where simulators of various
JAVA-enabled phone models can be obtained. In
our practical experiments, simulation was used to
develop and officially assess the many different pro-
jects.
Some of the student teams ambitiously wanted to
run their tools on real physical devices and teleph-
ony networks. Hence, a few of the projects were
demonstrated in the real world. During this, it turned
out that the implementation of the J2ME idea is not
yet sufficiently elaborated on many devices. Even
when using the high-level UI only, the developer has
to take care of behaviours of mobile phone devices,
which were not in accordance to the original J2ME
specifications. Workarounds can be used in these
cases, but the problem that is shown by this experi-
ence is that at the moment it is not possible to rely
on a software, which is constructed in full accor-
dance to the J2ME specifications.
Without pointing to any specific vendor – be-
cause most phone vendors have similar styled prob-
lems – it has to be remarked that the level of the
average phone J2ME has not yet reached a profes-
sional state. Furthermore, security restrictions pre-
vent that the UI can be truly minimized, because in
many phones the user has to manually select the
applied network data link, when the JAVA MIDlet
starts an HTTP query. By this kind of mechanism,
for which there exist certainly good security argu-
ments, the terminal software cannot be optimised to
the full degree.
Another issue is the network access speed on real
world devices. The query for the traffic channel in-
formation takes on a real wireless phone network
approximately 15 seconds, while in the simulation
environment the query is processed in around a sec-
ond. Although this appears on a first glance as a
back draw, the use of the described service will be
much more efficient than the possible alternatives.
Considering the practical example that one has to
decide after breakfast which way to drive to the
working office, the information collection through
the proposed system would require at maximum half
a minute, and around three keypad presses ( user
input actions). The alternative would be to use a
voice announcement service, which would take at
least the same time, and the user has to select his
particular information out of this generalised service,
or to use WEB browsing on a desktop computer.
Especially the latter method would consume several
minutes for obtaining the desired decision input
(booting of the computer, connecting to the Internet,
browsing action, shutting computer down).
4 FUTURE TECHNOLOGIES
Starting the discussion of wireless data accesses with
the relatively simple SMS communication system
(section 1) probably does not appear as appropriate.
Nevertheless, due to recent developments in J2ME
technology, this ancient communication method will
come into question for the described information
systems. One important feature of the most recent
specification of wireless JAVA (it is named MIDP
version 2.0) is the capability of PUSH mechanisms
(Fig. 5).
With the PUSH system, J2ME tools can be acti-
vated automatically (Ortiz, 2003). For this, the
MIDlets have to register for the required service. In
the particular application here, the terminal display
EFFICIENT INFORMATION RETRIEVAL FROM HANDHELD TERMINALS WITH WIRELESS DIGITAL PHONE
INTERFACE - Personalized information access on mobile phones and PDAs
59