Arab J Sci EngDOI 10.1007/s13369-014-0953-6

RESEARCH ARTICLE - ELECTRICAL ENGINEERING

Fuzzy-Based Assessment of Health Hazards of a ReferenceAntenna

Selcuk Comlekci · Ozlem Coskun · Mesud Kahriman

Received: 13 June 2012 / Accepted: 31 December 2012© King Fahd University of Petroleum and Minerals 2014

Abstract This paper discusses briefly a fuzzy-based assess-ment of health hazard due to electromagnetic radiation. TheRF electromagnetic fields, out of the measurement points,were calculated by the developed software based on fuzzylogic. The electric and magnetic field components of RF radi-ation value at any point can be compared with national/inter-national standards and limits easily using this software. Thereis currently a general consensus in the scientific and standardscommunity that the most significant parameter, in terms ofbiologically relevant effects of human exposure to radiofre-quency electromagnetic fields, specific absorption rate is thespecific energy absorption rate in tissue, a quantity properlyaveraged in time and space and expressed in watts per kilo-gram. The Institute of Electrical and Electronics Engineersrecognizes that there is public concern regarding the safety ofexposure to the radio frequency and microwave fields fromhand-held, portable, and mobile cellular telephones. Interna-tional organizations have established guidelines for humanexposure to radio frequency energy. While these guidelinesdiffer in some respects, their limits in the frequency rangeused by wireless communications devices are broadly simi-lar. The consensus of the scientific community, as reflectedin these exposure guidelines, is that exposure to RF energywithin the recommended limits stated in these guidelines issafe. However, there is a scientific discontinuity in view ofhealth hazards. In this study, a fuzzification/defuzzificationmethod of the discontinuity problem makes the “soft” bound-aries between hazardous regions and non-hazardous regions.In future studies, more sophisticated fuzzy methods shouldbe tested for more realistic solutions.

S. Comlekci · O. Coskun (B) · M. KahrimanDepartment of Electronics and Communication Engineering, Facultyof Engineering, Suleyman Demirel University, 32260 Isparta, Turkeye-mail: ozlemcoskun@sdu.edu.tr

Keywords Fuzzy logic · Safety standard · Health hazard

1 Introduction

The utilization of electromagnetic (EM) energy has increasedrapidly since the late 1990s. A number of organizations haveestablished limits for human exposure to EM fields. The stan-dards vary somewhat in their exposure limits and in other

123

Arab J Sci Eng

particulars. However, the frequencies used for wireless com-munications systems are broadly similar for (in) these dif-ferent guidelines. All of these guidelines include provisionsfor different exposure situations. These include limits forwhole-body exposure or partial body exposure that are morerelevant to the users of wireless communications. The stan-dards also require that the exposure be averaged over timeperiods ranging 6–30 min [1].

The International Commission on Non-Ionizing Radia-tion Protection (ICNIRP) guidelines [2], and the Instituteof Electrical and Electronics Engineers (IEEE) standard [1],specify occupational and general public (ICNIRP)/controlledenvironment and uncontrolled environment (IEEE) thresh-old levels for whole body and local rates of electromagneticenergy absorption, expressed in terms of the specific absorp-tion rate (SAR), measured in watts per kilogram. For a humanbody radiated by a base station antenna, the electromagneticenergy absorption depends in part on the antenna body dis-tance, so that, for a given exposure threshold level, thereis a corresponding minimum distance required between theantenna and the body.

International organizations have established guidelines forhuman exposure to radio frequency energy. These includethe IEEE C95.1 standard and The ICNIRP guidelines [2].Despite a considerable amount of speculation in the scientificliterature, no mechanism has established a standard such thatelectromagnetic fields at levels below recommended limitscan produce biological damage of clinical consequence [3–8].

Mousa [9] studied the electromagnetic radiation emittingfrom some cellular base stations around the city of Nablus.The study was performed at several sites. The readings obtai-ned were compared to some international standards and guide-lines. It has been noticed that the maximum measured valuewas only 0.007 % of the ICNIRP and 0.005 % of the FCCinternational limits. Furthermore, the values measured rep-resented not only radiation emitted from the mobile base sta-tions, but also that emitted from all other sources of radiationin the range of 200 kHz to 3 GHz. The signals here can haveeither destructive or instructive interference at some specificpoints, so it is recommended that the radiation emitting fromthe base stations should be investigated together with othersources such as local TV, FM and WLAN transmitters. Thiscan be achieved using a suitable spectrum analyzer. Anotherimportant issue is that the radiation exposure to the mobilestation itself should be measured since it may have a muchlarger value being very close to the users [9].

In Kaluski and Stasierski’s [10] work, a rough numericaltechnique for the calculation of the near EM field distributionin the vicinity of FM and TV antenna systems was presented.Faraone et al. [11] investigated the character of the averagepower density in the close proximity of base-station antennas.In 2003, a new measurement method for radiation emanating

from AM, FM, and TV antennas and mobile phone base sta-tions was proposed by Shay et al. [12]. Cicchetti and Faraone[13] proposed a prediction formula for estimating the peakequivalent power density in the near-field of cellular base-station array antennas.

Recently, Larcheveque et al. [14] studied the impact ofsmall-scale fading on the estimation of local average powerdensity for radio frequency exposure assessment. Joseph et al.[15] studied a low-cost measurement method for the extrac-tion of the relative phases of the fields of the base station andbroadcast antennas. In the end, Colak and Kocsalay workedRF electromagnetic field distribution around a TV broadcastantenna. They developed artificial neural network based soft-ware to estimate RF EMF in a small area around TV broadcastantennas [16].

2 Fuzzy Model and Study Design

The fuzzy logic method can be used to control processes thatare complex and nonlinear in the traditional control structure.In fuzzy systems, effective results can be obtained based onuncertain linguistic knowledge. Therefore, the fuzzy logicmethod is convenient for cases where the system is complex,and the result cannot be found using the traditional meth-ods or cases where the information is infinite or uncertain.Fuzzy logic is fit for soft computing in engineering prob-lems. In particular, uncertainties on the boundary conditionscan be solved using the soft computing approach of fuzzylogic [17,18].

A novel model has been used to find a realistic relation-ship between health hazard (or SAR) and electromagneticradiation (measured and calculated). The main objective isto overcome the problem of uncertainty regarding the eval-uation and classification of hazardous regions in the vicinityof antenna.

SAR is a unit of measurement used in the standard and itmeasures the amount of radio frequency energy

SAR = σ

∣∣E2

∣∣

ρm(1)

where is effective incident electric field value, mass densityof tissue, conductivity of tissue. The commercial field probesoperating in the wireless communication bands are sensitiveto and the reading is usually expressed in. These instrumentsare referred as isotropic E-field probes [19].

Exposure limits for radio frequency radiation have beenestablished by the Institute of Electrical and Electronics Engi-neers (IEEE) and the International Commission on Non Ion-ising Radiation Protection (ICNIRP). Safety distance (fromantenna to measuring point) can be found as

d =√

30 × P × 10G/10

E(2)

123

Arab J Sci Eng

where is output power of antenna, is antenna gain (25 dBi forcommercial antenna), is maximum permissible electric fieldintensity [1].

The main purpose of this paper is to study the use of fuzzymodeling for the analysis of health hazards due to electro-magnetic radiation.

3 Measurements and Modeling

In this study, there were a total of 50 measuring points in thecenter of City of Isparta, Turkey. The two service providershad total 5 pylons of 900 MHz reference antennas in this res-idential area. For system validation, our model was tested inview of obtained and calculated electric fields. These mea-surements have been conducted using the spectrum analyzer/satellite receiver meter unit which was used to investigate thereflections and background noises in the measuring media.Repetition time, frequency, and amplitude of spectrum of RFenergy (900 MHz) were also investigated, observed, and ver-ified by the satellite level meter that is PROMAX, MC-877C(Barcelona/Spain). All the reflection and exposure measure-ments were carried out by utilizing the Portable RF SurveySystem, HOLADAY, HI-4417 (MN/USA) with its standardprobe as well. The probe is able to select and obtain the vectorsum on the X, Y and Z axis. In order to see if they matched,the measured and calculated results were compared with eachother.

If one measures the field density value using the measuredor calculated electric field intensity in the above equation, thesafety distance can be calculated. The most common valuesof parameters can be obtained as shown in Figs. 1 and 2.

The normalized electric field can be expressed as.

%E =⎛

⎝

∣∣∣E

′ ∣∣∣

|E | − 1

⎞

⎠ × 100 (3)

|E ′|: Obtained electric field value analytically,

Fig. 1 Electric field versus distance from antenna with limit valuein 900 MHz communication system. This limit is recommended byICNIRP as the safety distance [20]

Safety distance vs power

4,05,06,07,08,09,0

10,011,012,013,014,0

0,0 5,0 10,0 15,0 20,0 25,0 30,0 35,0

Output power (W)

Saf

ety

dis

tan

ce (

m)

Fig. 2 Safety distance (m) from antenna can vary by means power

Table 1 Measured and calculated electric field values versus distance

Distance fromantenna (m)

Calculatedelectric field(V/m)

Measuredelectric field(V/m)

SAR relatedhazard grade(%) × 1,000

1 435.4 400 1,600

5 87 90 8.1

10 43.5 45 2

10.6 41 40 1.6

50 8.7 10 0.1

100 4.4 5 0

500 0.9 1.4 0

Measured electric field determines health hazard dealing with SAR

Pt(Watt)

Distance(m)

18 RulesMamdani

Percentage(%)

Fig. 3 Proposed fuzzy model to the system

|E |: Permissible electric field value to provide safety (42V/m). So, the normalized percentage can be used as a “hazardgrade”.

The electric field results obtained from these measure-ments were used to establish a fuzzy model. This modelrequires some results obtained from open area measurements.The model was used for the prediction of E field values. Soone needs only a validated model without any measurementprocess. The SAR defines the local E field and the energyabsorbed into tissue. Our model predicts the E field value intissue, or SAR. The predicted values from the model weretested and validated. According to the basic electromagnet-ics, we had to use some rules. These electric field measure-ments are tabulated in Table 1.

The Fuzzy Logic Toolbox of MATLABv6 was utilizedto establish our model at the Suleyman Demirel University

123

Arab J Sci Eng

Table 2 FAM table and rulebase of the system Pt (W) Distance (m)

Very very near Very near Near Mid Far Very far

Low Very harmful Mid Harmless Harmless Very harmless Very harmless

Mid Very very harmful Low harmful Mid Harmless Harmless Harmless

High Very very harmful Harmful Low harmful Mid Harmless Harmless

Fig. 4 Comparison of fuzzifiedharm zones between directionaland omni directional antennas.Fuzzified zones. Bold zonesrepresent more hazards

Fig. 5 Evaluation of harmful can be obtained analytically. SAR isknown directly proportional to electric field. Health hazard is definedas dealing with SAR

Engineering Faculty Lab. The fuzzy model consists of twoinputs (transmitter power of base station antenna, distancebetween base station antenna and measuring point) and oneoutput as a percentage for the expected health hazard. Thisfuzzy inference system (FIS) model is shown in Fig. 3 andthe fuzzy associative memory (FAM) table is provided inTable 2.

Fig. 6 Outputs of the model. Fuzzified zones can be classified as “softtransition” among the areas

4 Conclusion

After the defuzzification of the system, the crisp values areutilized to compare the analytical results for the calculated

123

Arab J Sci Eng

safety distance results. Graphical representations are pro-vided in the form of graphs in Figs. 4 and 5. Figure 6 rep-resents the fuzzified solutions to the regional health hazardsusing the fuzzy model presented. There have been a suffi-cient number of matches, between our results from the modeland our measurements that were mentioned in the Sect. 5.Figure 5 shows a close agreement between the measured andcalculated fields, especially in the near field.

5 Discussion

In this study, a new approach to obtain a risk assessment forthe energy radiated by a reference antenna is presented. Forinstance, the output power of the antenna varied between 5and 20 W in 900 MHz. According to the basic electromag-netic, electric field intensity decreases by distance in steps sothat the most effective criterion is a field at a certain point. Inthis respect, it is not easy to establish certain limits or bound-aries among “harmful” or “harmless” regions. Using the pro-posed method, one can classify (in view of hazards) somepoints in the vicinity of an antenna. It can be seen in Figs.4, 5; Table 1 that the relative risk calculated from the fuzzymethod and from the analytical solution matches each other.MATLAB-FIS gives acceptable linguistic outputs. Due to thevariable traffic condition, adaptive or proper models shouldbe created. Moreover, 3D solutions are always an essentialrequirement for real-time geographic conditions. In the futurestudies, more agreeable fuzzy models will be developed formore reliable risk assessment mapping of directional anten-nas.

References

1. IEEE C95.1-2005 IEEE standard for safety levels with respect tohuman exposure to radio frequency electromagnetic fields, 3 kHzto 300 GHz (2006). doi:10.1109/IEEESTD.2006.99501

2. Cooper, J.; Marx, B.; Buhl, J.; Hombach, V.: Determination ofsafety distance limits for a human near a cellular base stationantenna, adopting the IEEE standard or ICNIRP guidelines. Bio-electromagnetics 23, 429–443 (2002)

3. Bernardi, P.; Cavagnaro, M.; Pisa, S.; Piuzzi, E.: Human exposureto radio base-station antennas in urban environment. IEEE Trans.Microw. Theory Tech. 48, 1996–2002 (2000)

4. Gosselin, M.C.; Christ, A.; Kuhn, S.; Kuster, N.: Dependence of theoccupational exposure to mobile phone base stations on the proper-ties of the antenna and the human body. IEEE Trans. Electromagn.Compat. 51, 227–235 (2009)

5. Meyer, F.J.C.; Davidson, D.B.; Jakobus, U.; Stuchly, M.A.: Humanexposure assessment in the near field of GSM base-station antennasusing a hybrid finite element/method of moments technique. IEEETrans. Biomed. Eng 50, 224–233 (2003)

6. Hardell, L.; Sage, C.: Biological effects from electromagnetic fieldexposure and public exposure standards1. Biomed. Pharmacother.62, 104–109 (2008)

7. Markov, M.: Nonthermal mechanism of interactions between elec-tromagnetic fields and biological systems: a calmodulin example.Environmentalist 31, 114–120 (2011)

8. Yalçin, S.; Erdem, G.: Biological effects of electromagnetic fields.Afr. J. Biotechnol. 11, 3933–3941 (2012)

9. Mousa, A.: Electromagnetic radiation measurements and safetyissues of some cellular base stations in Nablus. JESTR 4(1), 35–42(2011)

10. Kaluski, M.; Stasierski, L.: Electromagnetic field estimation in thevicinity of panel antenna system for FM and TV broadcasting.IEEE Trans. Broadcast. 41, 136–142 (1995)

11. Faraone, A.; Tay, R.Y.S.; Joyner, K.H.; Balzano, Q.: Estimation ofthe average power density in the vicinity of cellular base-stationcollinear array antennas. IEEE Trans. Veh. Technol. 49, 984–996(2000)

12. Shay, W.T.; Lao, R.R.; Liang, W.: Practical measurement procedurefor EM radiation from base stations in Taiwan. Asia-Pac. Conf.Environ. Electromagn. CEEM-2003 605–608 (2003)

13. Cicchetti, R.; Faraone, A.: Estimation of the peak power density inthe vicinity of cellular and radio base station antennas. IEEE Trans.Electromagn. Compat. 46, 275–290 (2004)

14. Larchêveque, E.; Dale, C.; Wong, M.F.; Wiart, J.: Analysis of elec-tric field averaging for in situ radiofrequency exposure assessment.IEEE Trans. Veh. Technol. 54, 1245–1250 (2005)

15. Joseph, W.; Verloock, L.; Martens, L.: Reconstruction of the polar-ization ellipse of the EM field of telecommunication and broadcastantennas by a fast and low-cost measurement method. IEEE Trans.Electromagn. Compat. 48, 385–396 (2006)

16. Colak, I.; Koçsalay, I.: Neural network estimation of RF electro-magnetic pollution near TV broadcast transmitters. Microw. Opt.Technol. Lett. 50, 1004–1010 (2008)

17. Zadeh, L.A.: Fuzzy logic = computing with words. IEEE Trans.Fuzzy Syst. 4, 103–111 (1996)

18. Zadeh, L.A.: The roles of soft computing and fuzzy logic in theconception, design and deployment of intelligent system. IEEEAsia Pac. Conf. Circuits Syst 3–4 (1996)

19. Blanch, S.; Romeu, J.; Cardama, A.: Near field in the vicinity ofwireless base-station antennas: an exposure compliance approach.IEEE Trans. Antennas Propag. 50, 685–692 (2002)

20. Ahlbom, A.; Bergqvist, U.; Bernhardt, J.; Cesarini, J.; Grandolfo,M.; Hietanen, M.; Mckinlay, A.; Repacholi, M.; Sliney, D.; Stol-wijk, J.: Guidelines for limiting exposure to time-varying electric,magnetic, and electromagnetic fields (up to 300 GHz). InternationalCommission on Non-Ionizing Radiation Protection. Health Phys.74, 494–522 (1998)

123