© 2010 IEEE. Reprinted, with permission, from 2010 IEEE International Symposium on Electromagnetic Compatibility Proceedings.
2010 IEEE EMC Symposium Best Technical Paper Nominee
Wireless digital communication systems using the 2.4 GHz ISM (Industrial, Scientiﬁc and Medical) band are earning more and more popularity from day to day. Short range communication systems like Wireless LAN, Bluetooth and ZigBee are the most widely known and used standards nowadays. These communication systems are especially used for cable reduction and replacement in home, ofﬁce and industry. As these systems are – under special restrictions – nearly worldwide license free, they are very easy and cheap to integrate into electronic systems. With the use of wireless communication systems costs can be reduced and electronic communication systems can be easily extended. In either case, the Electromagnetic Compatibility (EMC) of the wireless communication systems has to be ensured so that the functionality and reliability of the system itself and other electronic systems in their vicinity is not reduced.
The effects of man-made electromagnetic interferences on electronic systems is an ongoing and widely researched topic. The effects of high-power and transient sources is a matter of particular interest and is summarized in the technical term High-Power Electromagnetics (HPEM ). The interferences caused by HPEM sources are refered in the topics of Intentional Electromagnetic Interferences (IEMI).
The effects of some HPEM sources on electronic equipment have been investigated in  and . The classiﬁcation of these effects on system level is given in . Building on the previous investigations, this article deals with the effects and estimations of susceptibility on wireless communication systems inside the 2.4 GHz ISM band by using UWB and radar pulses. In this article the main focus of the coupling path is laid on the antenna and the effects on the receiver system of a wireless communication system by changing the power, the pulse length and the pulse repetition frequency.
SOURCES OF HIGH-POWER AND TRANSIENT DISTURBANCES
The research and description of the effects of high-power transient disturbances is centralized in the technical term HPEM. Natural sources (e.g. lightning (LEMP) or electrostatic discharge (ESD)) and military sources (e.g. Nuclear Electromagnetic Pulse (NEMP), Ultra Wideband Pulse (UWB) or High-Power Microwave (HPM)) are the area of research in HPEM. In addition the technical term of High Intensity Radiated Fields (HIRF) is especially used in aviation  for the exposition of electronic devices and life-forms in vicinity of high electric ﬁeld strengths caused by, for example, the radar. The general parameters of the following pulses are the pulse duration and the pulse repetition frequency.
Radar-Radio Detection and Ranging
The range of application of radar is subdivided into military and civilian purposes. Especially for civilian usage, the radar is being used for surveillance and navigation of ship trafﬁc inshore and in port entrances, navigation aid and Search-and-Rescue (SAR). In civilian shipping, navigation is being performed by using s-band (3 GHz) and x-band (9.4 GHz) pulse radars. Table 1 from  outlines the typical parameters of a civilian s-band radar as it is used on every modern vessel. From the general radar equation the received radar signal power Pr of a radar target is calculated by
Table 1: Typical parameters of a civilian s-band radar for navigation
Typical HPEM sources are the LEMP, NEMP, UWB or the HPM. These HPEM sources have very high amplitudes and power. Due to this these sources can be seen as a serious thread for every system in their vicinity. The electromagnetic pulses of HPEM sources are characterized by the rise time tr and the pulse duration td, which is expressed as full width half maximum (FWHM). Different rise times and pulse durations are summarized in Table 2.
Table 2: Time parameters for different electromagnetic pulse forms
The UWB pulse is characterized by the fastest rise time of less than 200 ps and the shortest duration of less than 5 ns. compared to the other electromagnetic pulses. Therefore the frequency spectrum of the UWB pulse covers frequencies up several gigahertz. The threat of UWB pulses compared to the other pulses is therefore much higher and is being used inside for the following measurements.
The unipolar ultra wideband pulse uUWB(t) can be approximated in time domain with the use of a double exponential function 
where Vp denotes the maximum voltage of the pulse, α and β are time constants and k is a normalization factor, which is calculated by
MEASUREMENT OF THE INFLUENCE ON A GENERAL RECEIVER OF A WIRELESS COMMUNICATION SYSTEM
Communication theory subdivides a digital wireless communication system into three main sections: transmitter, channel and receiver. Figure 1 depicts the general structure of a digital wireless communication system. The transmitter itself consists of a source and the related quadrature modulator which mixes the digital modulated signals (inphase and quadrature component) into the passband. After the channel the passband signal is being mixed in the receiver with the help of the quadrature demodulator into the baseband and is being transferred to the sink.
Figure 1: Block diagram of a digital wireless communication system
Coupling into the Receiving Antenna
The interference of digital wireless communication systems against external electromagnetic disturbances is separated by frontdoor and backdoor coupling . While frontdoor coupling describes the coupling of electromagnetic energy via the antenna in the receiver of a wireless communication system, backdoor coupling describes the coupling of electromagnetic energy via the geometry of the system (e.g. pcb structures) and the connected cables respectively.
In Figure 2 the coupling of a typical s-band radar (measurement is performed at a radar station) and a UWB pulse (measurement is performed by using a UWB pulse generator PBG3 from Kentech Inc.1) into a standard bar antenna for the ISM band is exempliﬁed. Both plots have been normalized to a peak electrical ﬁeld strength of 1 kV/m of the incident pulses at the point of installation. The used radar pulses can be varied in time (1 μs, 0.3 μs and 0.08 μs). Figure 2(a) shows the coupling of an s-band radar pulse with a pulse length of 1 μs and Figure 2(b) the coupling of a radiated UWB pulse in an antenna placed the a Gigahertz Transversal Electromagnetic (GTEM) cell.
[ 1Parameter Kentech PBG3: peak voltage at 50 Ω: 12.5 kV rise time: below 100 ps, pulse length (FWHM): 3 ns]
While the maximum amplitude of the coupled s-band radar pulse reaches approximately 4 V, the maximum amplitude of the coupled UWB pulse reaches approximately 15 V. The time response of the coupled UWB pulse lasts approximately 20 ns.
Effect on the Baseband Signal
The electromagnetic coupling via the receiving antenna of a digital wireless receiver can result in corruption or disturbance of the receiving data. As the packaging of a wireless communication system receiver is very small and compact, the effect of the different coupled pulses is measured with the help on a typical quadrature demodulator.
A quadrature modulator from Analog Devices (AD8347, ) build on an evaluation board has been used (Figure 3). The external local oscillator can be varied from 0.7 GHz to 2.7 GHz and is being selected during the measurements to a carrier frequency inside the 2.4 GHz ISM band.
The effects of the different coupled pulses with varying ﬁeld strengths are presented in Figure 4. Only the inphase component of the baseband is being presented as the effect at the quadrature component shows the same behavior. Figure 4(a) and 4(b) show the effect of a radar system with electrical ﬁeld strengths varying from 0.8 kV/m to 1.8 kV/m (Figure 4(b) to 2.6 kV/m). With increasing ﬁeld strength the voltage deviation is rising up to 3 V. The inﬂuence length of the pulses matches at lower electric ﬁeld strengths nearly the radar pulse length, but is extended at electric ﬁeld strengths over 1.8 kV/m about approximately 300 ns.
The effect of a coupled UWB pulse for electric ﬁeld strengths from 70 V/m to 2.2 kV/m is plotted in Figure 4(c). With increasing electric ﬁeld strength the voltage deviation is increasing too. Peak voltages of up to 2.5 V are reached. The inﬂuence length at the highest ﬁeld strength of 2.2 kV/m is nearly 200 ns and approximately ten times higher than the measured length of the coupled UWB pulse of Figure 2(b).
Effect of Pulse Repetition Frequency and Pulse Duration on a “Real” Wireless Communication System
The previous measurements have shown that the inﬂuence of the used pulses causes disturbances in the baseband of a typical receiving unit. A worst-case scenario for the effects of disturbances caused by pulses is the assumption that a communication link is disrupted for the time of the disturbance inﬂuence. This worst-case scenario can be measured with the use of a high-frequency switching unit (HF switch) which is integrated between channel and receiving unit of the wireless communication system. To simulate a possible disturbance the switch blocks the communication between transmitter and receiver for a speciﬁc time. The HF switch is controlled by using a standard waveform generator to vary the disturbing length (pulse duration) and the disturbance repetition frequency (pulse repetition frequency).
The general measurement setup is depicted in Figure 5. The transmitter and receiver are connected by using 50 Ω SMA cables so that the communication is not performed over air. The channel is represented by an attenuator. The effect of different pulse lengths and pulse repetition frequencies on a ”real” wireless communication network can be investigated with this measurement setup. The following paragraphs will show the effects on a Wireless LAN and a Bluetooth network. The results will show the performance of the different wireless networks by measuring the maximum throughput (TCP mode) and the packet error rate (UDP mode) of the wireless communication link.
Results for Wireless LAN
This paragraph illustrates the result for a Wireless LAN 802.11g adhoc network at a transmission rate of 54 MBit/s. Two typical WLAN PCI cards in two PCs have been connected with an attenuator of 40 dB as depicted in Figure 5. The maximum reached data rate without any disturbances has been measured to approximately 28.8 MBit/s. The measurement has been performed by varying the pulse length from 0 ns (connection always on) to 1000 ns and the pulse repetition frequency from 0 kHz (connection always on) to 10 kHz. The results of this measurement for TCP and UDP transfer mode are summarized in Figure 6. In Figure 6(a) the data rate in per cent of the maximum achieved data rate is being plotted, Figure 6(b) plots the packet error rate in per cent respectively.
Comparing the two plots the transfer modes correlates in their behavior at different pulse lengths and pulse repetition frequencies. If pulse disturbances are not longer than 100 ns and the disturbance repetition frequencies are less than 1 kHz, reductions of at most 20% of the maximum data rate are observed, the packet error rate reaches a maximum of 20% respectively. A complete break down of the WLAN communication link is observed at disturbance lengths longer than approximately 300 ns and at disturbance repetition frequencies more than 3.5 kHz.
Results for Bluetooth
This paragraph summarizes the results for a Bluetooth network by using two USB Bluetooth 2.0 EDR (Enhanced Data Rate) devices connected to two PCs. The attenuator of this setup has been chosen to 30 dB. Again the pulse length and the pulse repetition frequency have been changed as described in the previous paragraph. The results of this measurement are depicted in Figure 7 for the TCP transfer mode whereas the maximum data rate without any disturbances has been measured to 1.6 MBit/s.
Looking again at the border of a reduction of the maximum data rate to 80%, the disturbance durations are approximately 350 ns and the disturbance repetition frequencies are less than 0.5 kHz. A complete break down of the Bluetooth communication link is observed at disturbance durations of more than 600 ns and disturbance repetition frequencies of nearly above 2 kHz.
Comparing the Bluetooth results to the Wireless LAN 802.11g results from the previous paragraph, the inﬂuence of the disturbance duration on the Blutooth communication is less than the disturbance duration on the WLAN communication link. But comparing the inﬂuence of higher disturbance repetition rates at longer disturbance duration the communication link is more inﬂuenced at higher disturbance repetition rates.
The measurements have shown how high-power and transient pulses couple into a receiving antenna of a wireless communication system. A standard antenna for the use inside the 2.4 GHz ISM band has been exposed by an UWB and radar pulses with different pulse durations. At a peak electric ﬁeld strength of 1 kV/m the maximum coupled voltage reaches up to 15 V for the UWB pulse and 4 V for the radar pulses.
Based on these coupling measurements the effect of these pulses on the baseband of a generic quadrature modulator has been investigated. The results show that the disturbance duration is increasing with increasing peak ﬁeld strength of the incident electromagnetic pulse. The caused effects in the baseband are achieving voltages up to 3 V and disturb the baseband signal.
In the end the inﬂuence of different pulse lengths and pulse repetition frequencies on a ”real” wireless communication link has been investigated. A worst-case analysis by switching the communication link on and off with a fast HF switch showed that a Bluetooth communication system is less interfered at short pulse durations (up to 300 ns) than a WLAN 802.11g communication link. Otherwise the WLAN communication was able to transmit data up to pulse repetition frequencies up to 3.5 kHz while the Bluetooth communication was only able to transmit data up to 2 kHz pulse repetition frequency. However both communication links have break down at higher pulse repetition rates and pulse lengths.
A complete destruction of the system has not been observed and after switching the disturbance off, the communication systems have worked without any disturbances as before.
- C. E. Baum, “From the Electromagnetic Pulse to High-Power Electromagnetics,” Proc. IEEE, vol. 80, June 1992.
- D. Nitsch, M. Camp, F. Sabath, J. L. ter Haseborg, andH. Garbe, “Susceptibility of Some Electronic Equipment to HPEM Threats,” IEEE Trans. Electromagn. Compat., vol. 46, August 2004.
- D. Nitsch and H. Garbe, “Susceptibility of Personal Computer Systems to Fast Transient Electronic Pulses,” IEEE Trans. Electromagn. Compat., vol. 48, August 2006.
- F. Sabath, “Classiﬁcation of Electromagnetic Effects at System Level,” Proc. of EMC Europe 2008, Hamburg, Germany, September 2008.
- Environmental Conditions and Test Procedures for Airborne Equipment, RTCA Std. DO-160, 2005.
- RADAR 1100, Technical Manual, SAM Electronics GmbH, Hamburg, Germany, 2005.
- C. Klünder, A. Schoof, and J. L. ter Haseborg, “On the Use of Wireless Communication Systems on Ships,” Proc. of EMC Europe Workshop 2007, Paris, France, June 2007.
- B. Standler, Protection of Electronic Circuits From Overvoltages, New York: Wiley, 1989.
- M. W. Wik, R. L. Gardner, and W. A. Radasky, “Electromagnetic Terrorism and Adverse of High-power Electromagnetic Environments,” 13th International Zurich Symposium and Technical Exhibition on EMC, Zurich, Switzerland, February 1999, pp. 181–185.
- “Datasheet Quadrature Demodulator AD8347,” Analog Devices, 2005.
Christian Klünder, email@example.com
Jan Luiken ter Haseborg, terHaseborg@ieee.org
Institute of Measurement Technology and Electromagnetic Compatibility Hamburg University of Technology