Image Processing Reference
In-Depth Information
security features. As they are frequently located at the premises requiring access permit, eavesdrop-
ping or message tampering would require a physical access to the medium. A potential solution
to provide a certain level of security is the access point (PLC, for instance) control. The emerging
Ethernet-based fieldbuses are more vulnerable to attacks on account of using the Ethernet and the
TCP/IP protocols and services. Here, the general communication security tools for TCP/IP apply.
At the device and embedded level, the limited computing, memory, and communication band-
width resources of controllers embedded in the field devices pose considerable challenge for the
implementation of effective security policies which, in general, are resource demanding. his limits
the applicability of the mainstream cryptographic protocols, even vendor-tailored versions. he oper-
ating systems running on small footprint controllers tend to implement essential services only, and
do not provide authentication or access control to protect mission and safety-critical field devices. In
applications restricted to the Hypertext Transfer Protocol (HTTP), such as embedded Web servers,
Digest Access Authentication [], a security extension to HTTP, may offer an alternative and viable
solution. In case of the denial of service (DoS) attack, the processor is preoccupied with handling
communication interrupts potentially compromising the (hard) real-time requirements and opera-
tional safety as a result—clever interrupt priority allocation and/or selection are needed. Interrupts
handling outside the normal operational conditions can cause with time battery draining on bat-
tery powered devices, making embedded controllers to become unavailable—a serious problem
in wireless sensor networks deployed on the factory floor where the function of the unavailable
node cannot be taken over by other nodes. In general, an embedded controller is expected to with-
stand autonomously many security attacks; the “buffer overflow” attack, for instance, which has the
potential to crash the system—proper error and exception handling is required here.
1.4 Wireless Sensor Networks
Another trend in networking of field devices has emerged recently, namely, wireless sensor networks,
which is another example of networked embedded systems. Here, the “embedded” factor is not so
evident as in other applications; particularly true for the ad-hoc and self-organizing networks where
the nodes may be embedded in the ecosystem or a battlefield, and to mention some.
Althoughreportsonactualapplicationsarescarce,andpotentialapplicationsintheprojectedareas
are still under consideration, the wireless sensor/actuator networks are in the deployment stage by
the manufacturing industry. he use of wireless links with field devices, such as sensors and actuators,
allows for flexible installation and maintenance, mobile operation required in case of mobile robots,
and alleviates problems with cabling. A wireless communication system to operate effectively in the
industrial/factory floor environment has to guarantee high reliability, low and predictable delay of
data transfer (typically, less than ms for real-time applications), support for high number of sensor/
actuators (over in a cell of a few meters radius), and low power consumption, and to mention
some. In the industrial environments, the characteristic for the wireless channel degradation arti-
facts can be compounded by the presence of electric motors or a variety of equipments causing the
electric discharge, which contribute to even greater levels of bit error and packet losses. One way to
partially alleviate the problem is either by designing robust and loss-tolerant applications and con-
trol algorithms, or by trying to improve the channel quality; all subject of extensive research and
development.
There is quite a difference in the requirements between the self-organizing and ad-hoc wireless
sensor networks and those imposed by the industrial applications. Some of the major characteristics
of the self-organizing and ad-hoc wireless sensor networks involve self-containment, lack of prear-
ranged network topology (organized by nodes on ah-hoc basis), and the ability to self-heal (network
