HOW TO ASSESS RELIABILITY OF INDUSTRIAL WIRELESS

SOLUTIONS

Lutz Rauchhaupt and Marko Krätzig

Institut für Automation und Kommunikation e.V. Magdeburg, Steinfeldstr. 3, Barleben, Germany

Keywords: Wireless Communication, Industrial Communication, Performance Investigation.

Abstract: Wireless communication is an emerging technology for industrial automation applications. Many solutions

are available which more or less consider industrial related requirements. One of the main concerns of

industrial automation system users is the reliability of wireless communication. The subject of this paper is a

method to assess reliability of wireless communication from the point of view of industrial automation

applications. Characteristic parameters are introduced which can be used in analytical studies, in network

simulations or measurements to assess reliability with respect to intended industrial control processes. In

particular the different use cases for the characteristic parameters are stated as well as the stochastic nature

of these parameters. Finally the influences are mentioned which have to be taken into account while

assessing wireless industrial communication systems.

1 INTRODUCTION

Wireless communication technologies are widely

spread in daily life. The price of wireless products is

thereby the main design aspect with respect to the

consumer market. Reliability is one of the minor

design goals. Therefore, almost everyone has had

negative experiences with such technologies and has

developed concerns regarding the usage of wireless

in industrial communication.

Indeed a number of measures are applied to

make industrial fit products from cheap solutions of

the consumer market (Dzung, 2005), (Weczerek,

2005), (Siemens, 2007). However, how can users be

convinced that wireless solutions meet the requested

reliability of industrial automation applications?

This paper starts with a definition of reliability

with respect to the application area - industrial

automation. Thereafter a model is introduced from

which relevant characteristic parameters are derived

which are used to assess reliability. Some examples

follow which show how to assess wireless solutions

with the described approach.

2 MODEL

2.1 Requirements

First we would like to clarify what reliability means

in context of wireless industrial communication. A

user of an industrial communication system expects

a certain process value, e.g. position or temperature,

at a certain interface within a defined time frame

without any errors under defined conditions. This is

an informal definition. In order to be able to assess

the degree of fulfilment of this requirement by

means of simulation or measurement, a formal

model is required.

First of all this model has to take into account the

application field - the industrial automation. The

parameters to be investigated have to be in line with

the design criteria of industrial automation systems.

Parameters such as Data Throughput or Bit Error

Rate are normally not useful to design a particular

automation application which e.g. shall manufacture

a product in a certain time frame or with a certain

cycle.

Furthermore, the model has to consider that there

is no general wireless standard available for

industrial automation which fits to all

communication tasks. Several different technologies

are used for industrial automation. A unified

interface between communication and application is

122

Rauchhaupt L. and Krätzig M. (2008).

HOW TO ASSESS RELIABILITY OF INDUSTRIAL WIRELESS SOLUTIONS.

In Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - RA, pages 122-130

DOI: 10.5220/0001503101220130

 SciTePress

not available. Therefore the model has to be

independent of a certain wireless technology and

even more it has to be open for future developments

in wireless communication such as Ultra-Wideband.

Last but not least the model should represent the

conditions of reality as accurate and complete as

possible and necessary. The following section

introduces an approach which fulfils the mentioned

requirements.

2.2 Approach

The abstraction of a distributed automation

application using wireless communication is shown

Figure 1. Wireless communication modules are

seen as an internal or external part of automation

devices. The automation devices have to fulfil

certain functions in a distributed automation system

and for that they have to communicate in our case

using a wireless communication media. From the

point of view of the automation system, the

communication characteristics at the interface

provided by the wireless solution are important.

These communication characteristics have to fit to

the time and error categories used in the industrial

automation area as introduced later in this document.

It must be clearly defined as to what the

communication interface is, upon which the

characteristics are related to. This interface consists

of a hardware part such as Ethernet or Dual Ported

RAM and a software part such as a communication

protocol or a driver. Besides a clear statement

concerning the communication interface and the

communication characteristics, the conditions have

to be described under which the characteristic values

are valid. The conditions can be described by a

number of influencing values which have different

origins. It is obvious that the communication system

itself affects the characteristics concerning e.g.

topology or data rate. It is also evident that the

communication media has influence because of

other users of the spectrum or because of the effects

of multi path fading. Furthermore, the characteristics

depend on the options chosen in the devices, which

means on its configuration. It is sometimes forgotten

that also the application affects the characteristic

values in the sense of the size of a packet or the

cycle of requests on the communication system.

Radio Communication Media

Automation Application

Conditions

Communication Characteristics

Communication Interface

Network Elements

Distributed

Automation

Module

Radio

Component

Distributed

Automation

Module

Radio

Component

Figure 1: Model approach for the assessment of wireless

industrial communication systems.

2.3 Characteristic and Influencing

Parameters

The analysis of literature concerning the usage of

characteristic parameters to describe and assess

communication behaviour has shown that there are

remarkable differences. Moreover, the definitions

come mostly from the application field of Ethernet,

Internet or telecommunication which does not fit to

the application field of industrial automation (e.g. in

(DIN EN 61491, 1999), (DIN EN 61209, 200)). That

is why it was necessary to find appropriate

definitions. The following characteristic parameters

are proposed to assess wireless communication

systems with respect to industrial automation

applications:

 Transmission delay

 Response time

 Update time

 Data throughput

 Packet loss rate

 Residual error rate

 Activation time after energy saving mode

 Energy requirements

It has to be mentioned that it is not required nor

recommended to use all parameters at the same time

to characterise a communication solution for a

certain application. The following sections show

exemplary which parameters fit to which kind of use

cases. The definitions of the listed characteristic

parameters can be found in (VDI/VDE 2185, 2007).

It is obvious that the values of the characteristics

are influenced by several parameters. That is why it

is important to know these parameters and their

values. Some of the parameters can be set with

certain values but it is also possible that the

parameters can not be influenced. In this case it is

important to determine the value of the parameter to

be able to assess the determined characteristic value.

HOW TO ASSESS RELIABILITY OF INDUSTRIAL WIRELESS SOLUTIONS

123

The first set of influencing values is related to

the application. This includes

 A background communication load, which

exists in addition to the communication under

investigation

 A user data length (packet size)

 A distance between the radio components

 An application period

 A relative moving speed between the radio

components

 A relative moving direction between the

radio components

The second set of influencing values is related to

the radio technology and the radio devices. It

includes

 A topology

 A frequency band

 A security functionality

 A safety functionality

 A type, direction and gain of antenna

 A transmission power

 A data rate via the physical media

 A media access control method

 A retry limit in case of errors

 A data rate at the communication interface

 A communication cycle

The third set of influencing values is related to

the environment in which the communication will

take place. It includes

 An application area

 Electromagnetic disturber

 Other frequency users

 Environmental conditions

Taking into account the listed influencing

parameters while determining target-oriented

relevant characteristic parameters, these can be used

to assess the time and error behaviour of a wireless

communication solution with respect to automation

applications.

2.4 Use on Reliability Definition

Now we can define the term reliability more specific

and we can describe how to assess reliability. In line

with the definition of chapter 2.1, reliability can be

seen as the degree in which you can expect that a

wireless communication solution meets the limits of

relevant characteristic parameters. With this

definition it is obvious that the assessment of

reliability needs stochastic measures. The

characteristic parameters are random variables.

Their behaviour follow probability density

functions. The reliability is the probability that a

value of a characteristic parameter is less or equal to

the limit defined by the automation application.

3 ASSESSMENT OF

RELIABILITY

3.1 Event Driven Data Transmission

Event driven data transmission is relevant for

process variables which indicate that a certain state

is assumed. For example when a work piece reaches

a specified position that it can be machined or when

a fluid reaches a defined level in a tank. In these

cases it is of interest as to how long it takes to

transfer the information from sensor to the control

unit e.g. programmable control logic (PLC). The

appropriate characteristic parameter to assess the

behaviour of a communication system is the

transmission delay.

Test Consumer

comm.ind

Test Producer

comm.req

Wireless System

under Test

Communication

Module

Communication

Module

transmission

delay

Communication

Interface

Communication

Interface

Figure 2: Definition of the transmission delay.

The definition of the transmission delay is based

on a producer consumer model (see Figure 2). It is

the time duration from the beginning of the handing

over of the first user data byte of a packet at the

communication interface in the test producer, up to

the handing over of the last user data byte of the

same packet at the communication interface at the

test consumer. It may be necessary to transmit

several telegrams between the communication

modules e.g. for acknowledgment. Furthermore,

network elements such as base stations may be

involved in the communication producing additional

delays. All these delays are covered by the

transmission delay.

As already mentioned, the characteristic

parameters and therefore the transmission delay are

random variables. Next it is shown which parts of

the transmission are randomly distributed and which

are constant. Furthermore, it is shown which parts

are specific for wireless transmission and what

makes these special considerations necessary.

To get a deeper understanding of enlarged

transmission delays, the most important segments of

a transmission delay value are listed in Table 1.

ICINCO 2008 - International Conference on Informatics in Control, Automation and Robotics

124

Table 1: Time segments which influence the transmission

delay value.

Time Segments Remark

Latency of

application

interface

The data transfer between application

module and communication module

may influence the transmission delay

value remarkably.

Latency of

implemen-

tation

The implementation of the

communication module influences the

transmission delay value remarkably.

User data

length

The user data length is related to the

data which is generated or consumed

by the automation application.

Data rate Bd

This rate is the radio transmission rate

of the user data. Sometimes a symbol

rate is given. In this case a symbol may

consist of more than one bit. The

header of a packet containing the user

data may be transmitted with another

data rate.

Technology

constant

The technology constant contains all

technology relevant protocol overheads

which are the same for each

transmission such as fixed idle times or

the time to transmit headers or tails.

Technology

variable

The technology variable contains all

technology relevant protocol overheads

which may vary for different

transmissions such as the time to get a

clear channel or the back-off time.

Depending on the technology,

acknowledgments are required to

complete a transmission.

umber of

retries

If a transmission is disturbed, the

packet is usually retransmitted. This

may be possible at different layers.

Transmission

deadline

DL In some cases the transmission is

terminated when a deadline is

exceeded.

Time

allocation of

additional

connections

Tac If there is more than one connection

established, the time allocated to the

other connections within the same

system has to be taken into account.

Global time

slot

GTS

In systems with TDMA the maximum

transmission delay can be calculated

considering the global time slot.

The random nature of the transmission delay is

being caused by the latency of the application

interface and implementation, by the technology

variable, the number of retries and the time

allocation for additional connections. In contrast to

wired communication, the wireless transmission is

affected much more by environmental influences.

Therefore, the random behaviour of the technology

variable together with the number of retries and the

time allocation for additional connections may

influence the transmission delay remarkably

Taking into account the time segments listed in

Table 1, the dependency of the transmission delay

can be described in different ways. The first way is

the given formula (1) and is illustrated in Figure

()

td ai i ai i ud ud tc tv r ac

T =f T (p),T (p),T (c),T (c),L ,Bd ,T ,T ,N ,T (1)

Tai(p) Tai(c)Ttc1 Lud/BdudTi(p) Ti(c)Ttv2 …Tac

* Nr

Ttd

Ttv1 Ttc2

Figure 3: Time segments of transmission delay depending

on re-transmissions, data rate and data length.

The second way to describe the transmission is

given in formula (2) and Figure 4. The

maximum transmission delay is fundamentally

influenced by the maximum allowed deadline which

covers the random behaviour of the media related

time segments.

tdmax ai i ai i

T = f(T (p), T (p), T (c), T (c), DL) (2)

Tai(p) Tai(c)Ti(p) Ti(c)DL

Ttdmax

Figure 4: Time segments of transmission delay depending

on a transmission deadline.

The third way to describe the transmission delay

is given in formula (3) and shown in Figure 5.

The maximum transmission delay is fundamental

depending on the number of retries and the global

time slot.

tdmax ai i ai i r GTS

T = f(T (p), T (p), T (c), T (c), N , T ) (3)

Tai(p) Tai(c)Ti(p) Ti(c) N

* T

GTS

Ttdmax

Figure 5: Time segments of transmission delay depending

on a global time slot.

As an example typical results of transmission

delay measurements are depicted in Figure 6. The

lower part of the figure shows the number of

packets, relative to the sample size, with certain

transmission delay values. The above described

random nature of the transmission delay can be

observed. The reasons for the different values are

HOW TO ASSESS RELIABILITY OF INDUSTRIAL WIRELESS SOLUTIONS

125

mainly transmission retries because of disturbances

and delays due to an occupied media. The curve

follows a Beta probability density function.

The probability distribution function of the

measurement is depicted in the upper part of Figure

6. It shows how many packets are transmitted by a

certain point of time.

Figure 6: Probability Density Function and Probability

Distribution Function (Cumulative Density Function) of

packets' transmission delay.

To assess the time behaviour of a wireless

solution, the well known statistical parameters for

the centre (e.g. mean value) and for the variation

(e.g. standard deviation) can be used. Our

experience is that the 95th percentile value P95 is

the best indicator for relevant changes in the time

behaviour e.g. because of disturbances. It is a trade

off between a feasible sample size (e.g. one million

packets) and an adequate significance.

The maximum value is not qualified for

assessment since it is a single value of a series of

measurements and it is not sure that the real

maximum value is captured. An infinite

measurement of the transmission delay would be

necessary or an inference to a larger population

using methods of interferential statistics. However,

the maximum value is considered so far as it

influences the value of the 95th percentile P95.

The assessment of the reliability of an event

driven data transmission means a comparison of a

limit for a statistical parameter given by the

application with the statistical parameters of a

measurement.

3.2 Cyclic Data Transmission

Most of the control processes in industrial

automation are cyclic. A process image is taken

cyclically via input devices or process interfaces. It

is processed by a controller and the result is output

via output devices or interfaces. One example is the

control of an overhead monorail system. The

position is acquired cyclically and as a result the

control information is transferred to the drive.

Also for these cases it is of interest as to how

long it takes to transfer data e.g. from the position

sensor to the controller. However, using the

transmission delay to assess the time behaviour

could be misleading. The problem is that in these

cases different cyclic processes are involved which

are not synchronised. Taking as an example a rotary

encoder sensor in which the communication buffer is

cyclically updated with the position information

which is cyclically transferred to a controller. The

effect of the asynchronism is shown in Figure 7.

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29

Number of Packets

Transmission Delay [ms]

TTD 20 TTD 21

Figure 7: Effect of asynchronous cyclic processes on

transmission delay.

Depending on the initial situation and the period

of the cyclic processes, the transmission delay may

have very different statistical parameters. In the case

TDD20, the communication cycle is an integral

multiple of the update time and therefore the

transmission delay value is constant. In the case

TDD21, first the transmission has been started just

before the buffer update (maximum value) and next

the transmission starts just after the buffer update

(minimum value). Thus, the mean value may

decrease even when the communication cycle

increases as shown in Table 2. The variation on the

other hand becomes much higher.

Since this behaviour is random and in reality

more than two cyclic processes are involved, it is

possible that the influence on the wireless

transmission is overlaid by the effect shown in

Figure 7 and can therefore possibly not be assessed.

This behaviour must at least be considered.

In most cases the update time is the appropriate

characteristic parameter to assess the time behaviour

of communication systems with cyclic data transfer.

The update time is ascertained according to the

ICINCO 2008 - International Conference on Informatics in Control, Automation and Robotics

126

producer-consumer-model (see Figure 8). This

means the period of time is from the delivery of a

packets` last user data byte, from the communication

interface of a consumer to the application, until the

delivery of the last user data byte of the following

packet of the same producer. Therefore the update

time is at least as long as the transmission delay

between producer and consumer, prolonged by the

time of the application update within the producer.

Table 2: Transmission delay values for different

communication cycles.

TTD20 TTD21

Buffer update 40 ms 40 ms

Communication cycle 20 ms 21 ms

*(T

/Bd

) 13 ms 13 ms

Transmission Delay (Mean Value) 33 ms 23,8 ms

Consumer

comm.ind

Producer

comm.req

Wireless System

under Test

Communication

Module

Communication

Module

Communication

Interface

Communication

Interface

update time

Figure 8: Definition of cycle time at producer-consumer

model.

Figure 9: Probability Density Function and Probability

Distribution Function (Cumulative Density Function) of

packets' update time.

As an example typical graphs are depicted in

Figure 9 for the probability function and the

distribution function of packets with certain update

time values. The update time is a Gaussian

distributed random value. The mean value indicates

in the first place the usability of the wireless

communication system for a certain cyclic control

process. Another important parameter is the span in

the automation area known as jitter. However, it has

to be mentioned that the measured minimum and

maximum values which are used to calculate the

span are most likely not the absolute extreme values.

Again an infinite measurement of the update time

would be necessary or an inference to a larger

population using methods of interferential statistics.

Therefore the span can only be assumed with a

certain probability.

Furthermore, measurements have shown that the

standard deviation value is well suited in order to

indicate influences on the wireless communication

system. Therefore, this parameter together with the

span can be used to assess the reliability of cyclic

data transmissions. The mean value follows the

application cycle if the system is correctly

configured. This means it is equal to the application

cycle which is the period e.g. a sensor value is

updated in the communication buffer.

3.3 Assessment of Error Behaviour

Up to now it was assumed that none of the

transferred data got lost. That means the configured

number of re-transmissions were sufficient to

transfer the user data successfully. This chapter

discusses how the reliability is assessed using the

Packet Loss Rate (PLR).

The packet loss rate (PLR) is ascertained

according to the producer-consumer-model. It

reveals how many of the packets, transferred from

the application to the communication interface

within the producer, are transmitted from the

communication interface to the application within

the consumer. The packet loss rate is determined as

follows:

tx rx

PLR

−

(4)

Where N

means number of transmitted packets

and N

means number of received packets.

In principle industrial wireless communication

solutions are designed to cope with the special

environmental conditions. They are considered

robust against interferer and industrial propagation

conditions. Therefore, in principle no packets

disappear. However, tacking into account the

maximum limits of the transmission delay the

situation changes. A remarkable packet loss rate can

be noticed in the case where a packet is considered

HOW TO ASSESS RELIABILITY OF INDUSTRIAL WIRELESS SOLUTIONS

127

to be lost when a certain value of transmission delay

is exceeded (or a deadline is missed). The

calculation of the packet loss rate for a certain use

case can be done as shown in formula (5). The

number of packets is acquired, which have

transmission delay values less or equal to the limit

defined by the use case. The difference to the sample

size is assumed to be lost packets. Thus, the packet

loss rate is calculated with respect to use case

specific limit of transmission delay. Especially

interferers cause higher PLR values. Considering the

packet losses and the transmission delay, the

reliability can be assessed by comparing the PLR

with the required packet loss rate of the use case.

Since also the PLR

is a random value, measures

should be foreseen by the application for the case

that data is not received within the expected time

frame.

tx rx TD TDmax

NN(t|t T )

PLR

−≤

= (5)

In the following chapter some examples are

given on how the reliability can be assessed.

4 EXAMPLES

4.1 Overview

The examples presented in this chapter shall

illustrate how the characteristic parameters are used

for certain purposes. The measurement scenarios,

the results and their assessment are not the topic of

this paper.

The results presented in this chapter come from

measurements with an IEEE 802.11g based

industrial communication solution. Different

influences have been investigated. A test system

generated packets with a length of 64 octets and

transferred them with an application cycle of 15 ms

to the interface of the test producer. At the test

consumer, the packets are transferred to the test

system. The test system measured the values of the

characteristic parameters. In none of the presented

cases could a packet loss concerning (4) be

investigated.

The systems under investigation are not specified

in detail in the current document since the project in

which the measurements are made has not yet been

completed. Moreover, in this paper the method of

assessment is in the focus and not the absolute

results of the tests.

4.2 Assessment of Event Driven Data

Transmission

The following figures show the number of packets

which are received after a certain time in line with

the above given definition of transmission delay.

Figure

10 shows the result of a measurement with an

industrial wireless solution made in an absorber hall

that means without any environmental influences.

Figure 10: Histogram of transmission delay in absorber

hall.

In Figure 11 the same system was placed in a

factory hall. Influences due to the environment can

be ascertained.

Figure 11: Histogram of transmission delay in factory hall.

Figure 12: Histogram of transmission delay in factory hall

with interferer.

ICINCO 2008 - International Conference on Informatics in Control, Automation and Robotics

128

In Figure 12 an interferer was finally activated

and a considerable influence can be ascertained.

However, comparing the values in Table 3 with a

transmission delay limit e.g. of a high speed I/O

system which is 10 ms, it is obvious that the

reliability of the communication is comparable for

all investigated conditions. In particular there was no

packet loss concerning the definition of formula

(5). That means even when the wireless

communication is noticeably influenced, this does

not mean that the requirements of a certain

application can not be fulfilled.

Table 3: Transmission Delay Values.

Transmission Delay [ms] Min. Max. P95

Absorber Hall 0,6 2,8 0,8

Factory Hall 0,6 4,5 0,8

Interferer 0,7 13,1 2,0

4.3 Assessment of Cyclic Data

Transmission

Figure 13 to Figure 15 show the update time for the

same scenarios described in the previous section.

As shown in Table 4 the mean values of the

update times are equal to the application cycle of

15 ms for all scenarios. By contrast the span differs.

Taking a limit for the span from ±1,5 ms in the first

case the requirement is fulfilled. In the second case

614 packets and in the third case 12.563 packets are

out of range. This results in a reliability of 99,9%

and 97,4% concerning the definition of this paper.

Figure 13: Histogram of update time in absorber hall.

Table 4: Update Time Values.

Update Time [ms] Min. Max. Mean

Absorber Hall 13,2 16,8 15,0

Factory Hall 11,2 18,8 15,0

Interferer 7,0 23,0 15,0

Figure 14: Histogram of update time in factory hall.

Figure 15: Histogram of update time in factory hall with

interferer.

5 CONCLUSIONS

In the paper we presented a proposal on how to

assess the reliability of industrial wireless solutions.

A fundamental requirement for such a method is the

focus on industrial automation applications. That is

why characteristic values such as transmission delay,

update time and packet loss are used in the way

defined in this paper. It was pointed out that these

parameters are random variables which mean the

statistical parameters have to be considered.

The described method has been used to assess

the coexistence of different industrial wireless

communication solutions. Furthermore, these

solutions are currently being used to assess the

possibility of using wireless communication in

automation applications with safety requirements. A

test system is available which supports the

measurements of the described characteristic

parameters (Rauchhaupt, 2006).

The approach can be used for analytical studies,

simulations and tests. The method is required for

wireless communication since the dimension of

influences can be remarkably greater than in wired

systems.

HOW TO ASSESS RELIABILITY OF INDUSTRIAL WIRELESS SOLUTIONS

129

The work described in this paper is accompanied

by important manufacturers of automation and radio

solutions, and users of such systems which work

together in the German Society of Measurement and

Automation. As a result the characteristic parameters

presented in the paper are introduced in the

VDI/VDE-Guideline 2185 "Radio based

communication in industrial automation".

ACKNOWLEDGEMENTS

This paper presents results from the project "Funk-

Transfer-Tester für industrielle Funklösungen"

which deals with the development of a method and a

tool used to investigate the time and error behaviour

of radio solutions for industrial automation

applications. The project is funded by the German

Ministry of Commerce and Labour within the

research program "Förderung von Forschung und

Entwicklung bei Wachstumsträgern in

benachteiligten Regionen" - INNOVATIVE-

WACHSTUMS¬TRÄGER (INNO-WATT).

REFERENCES

DIN EN 61491, 1999. "Elektrische Ausrüstung von

Industriemaschinen - Serielle Datenverbindung für

Echtzeit-Kommunikation zwischen Steuerungen und

Antrieben (IEC 61491:1995, modifiziert); Deutsche

Fassung EN 61491:1998"

DIN EN 61209, 2000.

"Navigations- und

Funkkommunikationsgeräte und -systeme für die

Seeschifffahrt – Integrierte Brückensysteme (IBS) -

Betriebs- und Leistungsanforderungen, Prüfverfahren

und geforderte Prüfergebnisse (IEC 61209:1999);

Deutsche Fassung EN 61209:1999"

VDI/VDE 2185, 2007.

"Funkgestützte Kommunikation in

der Automatisierungstechnik (Radio based

communication in industrial automation)"

Rauchhaupt, L., Gnad, A., Hintze, E., 2006.

"Wireless

Automation Under Test"

, embedded world 2006

conference, Nurnberg, Germany, proceedings part I,

ISBN 3 7723 0143 6, pp. 529

Dzung, D., Apneseth, C., Endresen, J., Frey, J.-E., 2005.

“Design and Implementation of a Real-Time Wireless

Sensor/Actuator Communication System"

10.1109/ETFA.2005.1612710, Volume: 2, pp. 10 – 19,

ISBN: 0-7803-9401-1

Weczerek, J., 2005.

"Erfahrungen beim industriellen

Einsatz von Bluetooth als Wireless-IO-Lösung."

VDI-

KfiT-Fachtagung Funkgestützte Kommunikation in

der industriellen Automatisierungstechnik,

Düsseldorf, Germany, ISBN 3-18.091889-6, pp. 83-93

Siemens, 2007.

"Industrial mobile communication."

Siemens AG, Order No. 6ZB5530-1AM02-0BA2

ICINCO 2008 - International Conference on Informatics in Control, Automation and Robotics

130