The Influence of Load Impedance, Line Length, and ... - IEEE Xplore

180

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 23, NO. 1, JANUARY 2008

The Influence of Load Impedance, Line Length, and Branches on Underground Cable Power-Line Communications (PLC) Systems Justinian Anatory, Student Member, IEEE, Nelson Theethayi, Member, IEEE, Rajeev Thottappillil, Senior Member, IEEE, M. M Kissaka, and N. H. Mvungi

Abstract—An underground cable power transmission system is widely used in urban low-voltage power distribution systems. In order to assess the performance of such distribution systems as a low-voltage broadband power-line communication (BPLC) channel, this paper investigates the effects of load impedance, line length, and branches on such systems, with special emphasis on power-line networks found in Tanzania. From the frequency response of the transfer function (ratio of the received and transmitted signals), it is seen that the position of notches and peaks in the magnitude are largely affected (observed in time-domain responses too) by the aforementioned network configuration and parameters. Additionally, channel capacity for such PLC channels for various conditions is investigated. The observations presented in this paper could be helpful as a suitable design of the PLC systems for better data transfer and system performance. Index Terms—Branched network, broadband power line, channel capacity, interconnections, load impedance, low-voltage channel, multipath, transfer function, underground cable.

I. INTRODUCTION

R

ECENTLY, the power line have been proposed as an infrastructure for broadband services, digital entertainment systems, and other services in most countries. However, the network is affected in terms of channel attenuation, variable position of notches, signal distortions, channel capacity, etc., depending on network topology and its operation. Based on [1]–[7], the stochastic signal attenuations, notches, etc. are due to the switching ON/OFF of equipment (sudden changes in loads), the addition of new lines or customers, etc. For this reason, in the case of a buried power transmission system working as a BPLC system, we shall study the influence of the number of branches (N ), line lengths (d ) from the transmitter to the receiver and branched line length (X), terManuscript received December 27, 2006; revised March 5, 2007. This work is a collaboration research between the Faculty of Electrical and Computer Systems Engineering, University of Dar es Salaam, Tanzania through SIDA/SAREC and the Division for Electricity and Lightning Research, Uppsala University, Sweden. Paper no. TPWRD-00831-2006. J. Anatory, M. M. Kissaka, and N. H. Mvungi are with the Faculty of Electrical and Computer Systems, University of Dar es Salaam, Dar es Salaam, Tanzania (e-mail: [email protected]; [email protected]; [email protected]. tz). N. Theethayi and R. Thottappillil are with the Division for Electricity and Lightning Research, Uppsala University, Uppsala, Sweden (e-mail: [email protected]; [email protected]). Digital Object Identifier 10.1109/TPWRD.2007.911020

) on minal loads (infinite and low), impedances (Z and Z the signal response based on transmission-line analysis. The information presented in this paper could be helpful for BPLC system design that involves buried cables. II. SYSTEM UNDER STUDY In general, the transmission-line equations that are to be solved for voltage and currents on the line are given by (1). Assume that the line is oriented in the direction

(1) In (1), V and I are the voltage and current on the line and Z and Y are the total series impedance and shunt admittance of the line. The configuration of an underground cable power line beginning from a distribution transformer is shown in Fig. 1(a) and the corresponding transmission-line representation for an adjacent conductor return is shown in Fig. 1(b). Fig. 1(c) shows the cross section of a typical power cable that has four conductors. Thus, for the case with adjacent conductor return, the series impedance and shunt admittance of the line needed for solving (1), including the possible skin effect phenomena [8] of the conductor, are shown and

(2a) (2b) (2c) (3) Equation (2) is an approximation based on [8] and in (3), is the dielectric loss angle. For the configuration corresponding to Fig. 1(a), is the depth of the cable below the ground, with radius and is the line length. The per-unit length transmission-line parameters are given by (4) for self inductance and capacitance of the line [6]

0885-8977/$25.00 © 2007 IEEE

(4)

ANATORY et al.: INFLUENCE OF LOAD IMPEDANCE, LINE LENGTH, AND BRANCHES

181

(a)

Fig. 2. Power-line network with distributed branches.

characteristic impedance of any terminal with source while and are the source voltage and load impedance, respectively, based on Fig. 2. M is the total number of distributed nodes (1, 2 d M ) (b)

(5a) (5b) source otherwise

(5c)

In (6), is the impulse response obtained by an inverse Fourier transform of (5a) (c)

(6)

Fig. 1. (a) Underground cable with an adjacent conductor as a return conductor. (b) Per-unit length equivalent underground circuit with an adjacent conductor as a return. (c) Cross section of a typical power-line cable (four conductors) [6].

Based on [6] for frequencies of interest and hence and are neglected in the present analyses. III. POWER-LINE CHANNEL MODEL For a transmission line with distributed branches (e.g., Fig. 2), the generalized transfer function can be represented by (5a), is the total number of branches connected [9]–[11]. In (5a), at a given node (e.g., see “node 1” in Fig. 2) and terminated in , and represent any arbitrary load. Let n, m, M, any branch number, any referenced (terminated) load, number of reflections (with a total of L number of reflections), transfer function between a given source point (transmitter) to a referenced load termination m, and a transmission factor with reference to any load termination m, respectively. With these, the at referred node “d” is given by signal contribution factor (5b), where is the reflection factor at the referred node d is the propagabetween line to the referenced load m, tion constant of line at referred node “d” that has line length . All terminal reflection factors , in general, are given by (5c), except at the source, where is the source reis the source impedance, is the flection factor [10]. Also,

IV. INFLUENCE OF LINE LENGTH A. Length From Transmitter to the Receiver We consider the power-line cables used in low-voltage systems, where the line ABC (see Fig. 3) is NAYY150SE with radius (r) per conductor that is equal to 6.9099 mm, the insumm), relative permittivity 4, using lation thickness ( H/m, (4) the per-unit length parameters are (L C pF/m). The branched cable BD is NAYY35RE with radius (r) per conductor equal to 5.9161 mm, the insulamm), relative permittivity 4, using tion thickness ( H/m, (4) the per-unit length parameters are (L C pF/m) [5]–[7]. The characteristic impedance while the of the line was calculated as propagation constant was , where the paand is as described previously. The charrameters acteristic impedance of line AB is Z while the characteristic impedance of BD is Z . In Fig. 3, . For sensitivity, the analysis of the line length AC was varied as 1.2 km, 600 m, 300 m, and 150 m, with point B always at the midpoint of AC. The branched line length (BD) was kept constant at 15 m and terminated in 10 k . The transfer function was calculated by taking the voltage ratio of point C to that of A [i.e., using

182


Fig. 3. Power-line network with branch.

Fig. 5. Simulation results for a received signal in the cable link with one branch for different values of ac per Fig. 3: (a) 1.2 km, (b) 600 m, (c) 300 m, and (d) 150 m.

Fig. 4. Transfer functions for a cable link with one branch for different values of ac per Fig. 3: (a) 1.2 km, (b) 600 m, (c) 300 m, and (d) 150 m.

(5)]. Fig. 4(a)–(d) shows the transfer function relating the voltages at the load and sending end for various lengths of AC with the same boundary conditions at line terminations. From Fig. 4, the peaks and notches in frequency response do not vary with either frequency or line length. The position of peaks and notches is independent of the line length from transmitter to receiver. However, the attenuation increases as line length and frequency increases. Fig. 5(a)–(d) is the received signal at the load Z for a signal from the source at the sending end having a shape of rectangular pulse with an amplitude 2 V (pulse width of 1 s and shifted by 0.5 s) for different cable lengths. From the time-domain signatures of the received pulse at the load Z (calculated from the frequency response of the voltage at Z and using ifft), it is observed that the signal attenuation increases with cable lengths and the distortion increases in such a way that the signal tends to lose its original shape. This confirms that in the underground cables, the signals encounter both attenuation and distortions which depend on transmission-line length. The effects of branch length are studied next. B. Branched Length The configuration is the same as in Fig. 3; but now the length AC is kept constant at 1.2 km, while the length of BD was varied as 10, 20, 30, and 40 m. Point D was terminated in 10 k . Similar investigations as conducted previously were carried out using (5). The transfer function for all cases relating the voltage at the load (Z ) and launched voltage at point A is shown in Fig. 6(a)–(d). It is observed that the position of notches and peaks is case dependant (depending on branched line length). As the branched line length increases, the number of notches increases. The attenuation in each case increases with frequency.

Fig. 6. Transfer function for the cable link of 1.2 km with one branch of length for different values of BD per Fig. 3: (a) 10, (b) 20, (c) 30, and (d) 40 m.

The generalized expression for frequency position (f in megahertz) of an th notch in terms of branched line length (X in m) is approximately given by (7)

(7) Fig. 7(a)–(d) shows the received time-domain signal for the same rectangular pulse as mentioned in the previous section injected at the sending end. From the results, it is seen that the distortions increase with branched lengths. Next, let us study the effect of the number of interconnected branches. V. INFLUENCE OF THE NUMBER OF BRANCHES A. Multiple Branches at Single Node Consider the configuration in Fig. 8. The line length of AC is 1.2 km with all branches that are 15 m long concentrated at node B and terminated in 10 k with cable parameters that are the same as before. We vary the number of branches as 2, 4, 5, and 6 and Fig. 9(a)–(d) shows the transfer functions for a different number of branches. It is seen that the sharpness of notches decreases with an increase in the number of branches


Fig. 7. Simulation results for the received signal in the cable link of 1.2 km with one branch of length for different values of BD as per Fig. 3: (a) 10, (b) 20, (c) 30, and (d) 40 m.

183

Fig. 10. Simulation results for the received signal in the cable link with multiple branches at the single node as per Fig. 8: (a) two branches, (b) four branches, (c) five branches, and (d) six branches.

Fig. 11. Power-line low-voltage network with distributed branches. Fig. 8. Power-line network with five multiple branches at a single node.

B. Distributed Branches

Fig. 9. Transfer function for the cable link with multiple branches at the single node as per Fig. 8: (a) two branches, (b) four branches, (c) five branches, and (d) six branches.

at the same node, but with a further increase in the number of branches, there is a tilt in the notch shape leading to a reduction in attenuation. Fig. 10(a)–(d) is the received time-domain signal at the load Z for a rectangular signal injected at the sending end as in the previous example for various branch numbers. It is seen in the low-voltage channel that by increasing the number of branches at the same node causes the received signals to be more attenuated and distorted.

Now consider the underground cable with distributed branches as in Fig. 11. The number of branches was increased in the link between point A and J. The length AJ was 1.2 km while all branches were 15 m long. The number of branches was increased as 2, 5, 10, and 15 such that in each case, they were equally distributed between A and J. The cable parameters are the same as before. The terminations were each branch in any of the cases that was 10 k . Fig. 12(a)–(d) is the transfer function for a different number of branches. It is seen that the positions of deep notches do not change for any case. As the number of branches increases, the attenuations tends to increase. Fig. 13(a)–(d) shows the received time-domain signal at the load Z for a rectangular signal injected at the sending end as in previous example for various branch numbers. It is observed that the peaks of the signal increase with the number of branches due to successive reflection from the distributed branches. VI. INFLUENCE OF LOAD IMPEDANCE A. Low Resistive Load We now consider the effects of load impedances. The load impedances at point D in Fig. 3 were varied as 5, 10, 20 , and characteristic impedance of branch BD. The length of AC is 1.2 km and BD is 15 m. Fig. 14(a)–(d) is a transfer function for

184

Fig. 12. Transfer function for the cable link with distributed branches as per Fig. 11: (a) two branches, (b) five branches, (c) ten branches, and (d) 15 branches.

Fig. 13. Simulation results for the received signal in the cable link with distributed branches as per Fig. 11: (a) two branches, (b) five branches, (c) ten branches, and (d) 15 branches.


Fig. 15. Simulation results for a received signal in the cable link of 1.2 km with a branch of 15 m for different loads at D as per Fig. 3: (a) 5 , (b) 10 , (c) 20

, and (d) Zo.

Fig. 16. Transfer function response for the cable link of 1.2 km with a branch of 15 m for different loads at D as per Fig. 3: (a) 50 , (b) 100 , (c) 1 k , and (d) 10 k .

the previous example for various loads. It is seen that as the load impedance increases, both signal attenuation and distortions tend to reduce. B. High Resistive Load

Fig. 14. Transfer functions for the cable link of 1.2 km with a branch of 15 m for different loads at D as per Fig. 3: (a) 5 , (b) 10 , (c) 20 , and (d) Zo.

all cases. It is seen that as the load impedance increases toward characteristic impedance, the notches tend to improve. Fig. 15(a)–(d) is the received time-domain signal at the load Z for the rectangular signal injected at the sending end as in

We consider the same case as in the previous example but now with higher impedances for the load varied as 50 , 100 , 1 k , and 10 k . Fig. 16(a)–(d) is a transfer function. It is seen that as the load impedance tends to be higher than channel characteristic impedance, the notches tend to become more prominent. Fig. 17(a)–(d) is the received time-domain signal at the load Z for the rectangular signal injected at the sending end as in previous example for various loads. It is seen that as the load impedance increases from lower to higher values, both signal attenuation and distortions tends to reduce as observed previously. The general expression for the position of notches with open-circuit load impedance is given by (8)

(8)


185

Fig. 18. Noise in the power-line network based on [13].

Fig. 17. Simulation results for a received signal in the cable link of 1.2 km with a branch of 15 m for different loads at D per Fig. 3: (a) 50 , (b) 100 , (c) 1 k , and (d) 10 k .

Next, let us see the influence of load impedance, line length, and branches on the channel capacity of the underground cable. VII. CHANNEL CAPACITY It is said that without electromagnetic shielding, power-line cables are sensitive to external noise from various radio-frequency devices and electromechanical equipment leading to electromagnetic-interference problems. There has been considerable effort by various researchers to model the noise in the PLC network [12]–[15]. The noise in PLC systems can be classified into three types, namely: 1) colored background noise, 2) narrowband noise, and 3) impulse noise. Colored noise is the sum total of various noise sources with low power. Narrowband (ingress) noise has amplitude-modulated signals caused by induction from radio-station signals in medium and shortwave bands. The power spectral density (PSD) for background noise is usually around 145 dBm/Hz and this is about 30 dB above the thermal noise floor [13]. The impulsive noise has a maximum amplitude of 40 dBm/Hz higher than background and/or narrowband noise. These conclusions are made based on measurements carried out by Liu et al. [12] for indoor cases and the equivalent noise model is documented by Chen [13]. We used an assumption that this typical noise can also be the same for the low-voltage power-line channel under study. A typical noise frequency response in the power-line network can be represented as shown in Fig. 18. Next, let us use the channel models and noise in the power line to determine the channel capacities in different power-line channels for various cases of branched and terminal load conditions. In the investigations to be presented, we first consider the channel with distributed branches as shown in Fig. 11. The number of branches was varied in the link between points A and J. The distance between points A and J was 1.2 km while all corresponding branch lengths were 15 m long. The distribution of the number of branches was varied as 2, 5, 10, and 15 for the sake of demonstration. Note that for each case, the distances between the branches were equal and equally distributed between the link A and J. Further, for each branch case corresponding to

the branch length and number of branches, the load impedance was varied as 5 , characteristic impedance, and 100 k . For any case treated, the channel transfer function is determined by using (5). The channel capacity was determined using (9), with a frequency variation between 1–30 MHz. S(f) is the received signal power and N(f) is the noise power, which is dependent on the transmitted signal power and channel transfer function as given in (10). The noise power level N(f) for different frequencies is considered based on Fig. 18, [13]. Due to the limitations on the transmitted power, the field strength is limited to 30 dB V/m. Thus, the allowed PSD can be estimated according to (11) and is found to be between 72 dBm/Hz and 52 dBm/Hz [13], corresponding to a coupling factor in the range of 65 dB and 45 dB for a distance of about 30 m [13]. For this study, we chose the range of PSD to be between 90 dBm/Hz to 30 dBm/Hz (9) (10) (11) Fig. 19 shows the variation of channel capacity in megabits per second against PSD in decibel meters per hertz for various branch numbers and different load terminations. It is seen that channel capacity decreases with an increase in the number of branches between the sending and receiving ends. Note that all of the branch lengths were kept constant at 15 m. For the link with fewer branches [see Fig. 19(a)], the influence of load impedance is negligible. However, it is seen that as the number of branches increases more than five, the influence of branches on channel capacity is predominant. Note that the channel capacity is minimum when the load impedances are terminated in characteristic impedance. The differences in the channel capacity are negligible whether the load is 5 or 100 k . It seems that the terminal load variations do not affect the channel capacity except with characteristic impedance terminations. But the channel capacity is less with larger branches. Fig. 20 shows the variation of channel capacity against PSD for different branch numbers and with different lengths from the transmitter to the receiver. Here, we keep the terminal loads at 10 k and the branched line length at 15 m. The direct lengths

186


Fig. 19. Channel capacity for a cable link of 1.2 km with a different number of branches and terminations per Fig. 11: (a) two branches, (b) five branches, (c) ten branches, and (d) 15 branches.

Fig. 20. Channel capacity for a cable link of 1.2 km with a different number of fixed branch lengths and load per Fig. 11: (a) two branches, (b) five branches, (c) ten branches, and (d) 15 branches.

(from transmitter to the receiver) were varied as 250 m, 500 m, and 1 km. It is observed that for a link with two branches, the influence of length is observable; however, as the number of branches increases, both the 250- and 500-m link have similar channel capacity. Fig. 21 shows the variation of channel capacity against PSD for different branch numbers and with different branch lengths. Here, we keep the terminal loads at 10 k and the length of the link at 1.2 km. The branch lengths are varied as 10, 20, and 40 m. It is seen that as the branch length increases, the channel capacity decreases.

Fig. 21. Channel capacity for a cable link of 1.2 km with different branch lengths and for a different number of branches and fixed branch loads per Fig. 11: (a) two branches, (b) five branches, (c) ten branches, and (d) 15 branches.

the transmitting and receiving ends. However, the signal attenuation increases as link length and frequency increases. Additionally, the distortion increases in such a way that the signal tends to lose its original shape. The position of notches is case dependent (i.e., dependent on the number of branches, branched line length, and terminal impedances on those branches). As the branched line length increases, the number of notches increases. The signal distortions also increase. The sharpness of notches decreases with an increase in the number of branches at the same node, but with a further increase in the number of branches, there is a tilt in the notch shape leading to a reduction in attenuation. The increasing number of branches at the same node causes the received signals to be more attenuated and distorted. As the number of distributed branches in the link between the sending and receiving ends increase, the attenuations tends to increase in such a manner that there could be a reduction in the available bandwidth. The reason could be due to the successive reflection from the distributed branches. As the terminal impedances on the branches increase to line characteristic impedance, the notches tend to improve and both signal attenuation and distortions tend to reduce. As the load impedance is increased beyond the corresponding line characteristic impedance, the notches tend to be more prominent and signal attenuation and distortion are further improved. The channel capacity decreases with an increase in the number of branches between the sending and receiving ends. The channel capacity is minimum when the load impedances are terminated in characteristic impedance. The sensitivity analysis presented here has important implications for the possible design considerations of PLC equipment for broadband applications using buried cable power systems.

VIII. CONCLUSION For an underground low-voltage cable channel, the position of notches in frequency response does not vary with either transmitted source frequency or direct line or link length connecting

REFERENCES [1] F. J. Canete, L. Diez, J. A. Cortes, and J. T. Entrambasaguas, “Broadband modeling of indoor power-line channels,” IEEE Trans. Consum. Electron., vol. 48, no. 1, pp. 175–183, Feb. 2002.


[2] A. B. Gutierrez, A. Darmand, V. Watt, and L. Ngalamou, “Design of an analog electronic interface for a power line based telephony system,” in Proc. IEEE ISPLC, Orlando, FL, Mar. 2006, pp. 232–238. [3] C. Papaleonidopoulos, C. G. Karagiannopoulos, D. P. Agoris, P. D. Bourkas, and N. J. Theodorou, “HF signal transmission over power lines and transfer function measurement,” in Proc. 6th IASTED Int. Conf., Rhodes, Greece, Jul. 3–6, 2001, pp. 502–505. [4] N. Pavlidou, A. J. H. Vinck, J. Yazdani, and B. Honary, “Power line communications: State of the art and future trends,” IEEE Commun. Mag., vol. 41, no. 4, pp. 34–40, Apr. 2003. [5] M. Zimmermann and K. Dostert, “A multipath model for the power line channel,” IEEE Trans. Commun., vol. 50, no. 4, pp. 553–559, Apr. 2002. [6] M. Zimmermann and K. Dostert, “Multi-path signal propagation model for the power line channel in the high frequency range,” in Proc. Int. Symp. Power-line Communications and its Applications, 1999, pp. 45–51. [7] M. Götz, M. Rapp, and K. Dostert, “Power line channel characteristics and their effect on communication system design,” IEEE Commun. Mag., vol. 42, no. 4, pp. 78–86, Apr. 2004. [8] S. N. Nahman and R. D. Holt, “Transient analysis of coaxial cables using the skin effect approximation,” IEEE Trans. Circuit Theory, vol. CT-19, no. 5, pp. 443–451, Sep. 1972. [9] J. Anatory, M. M. Kissaka, and N. H. Mvungi, “Channel model for broadband power-line communication,” IEEE Trans. Power Del., vol. 22, no. 1, pp. 131–145, Jan. 2007. [10] J. Anatory, M. M. Kissaka, and N. H. Mvungi, “The effects of load impedance, line length and branches in the BPLC-transmission lines analysis for medium voltage channel,” IEEE Trans. Power Del., vol. 22, no. 4, pp. 2156–2162, Oct. 2007. [11] J. Anatory, N. Theethayi, R. Thottappillil, M. M. Kissaka, and N. H. Mvungi, “The effects of load impedance, line length and branches in the BPLC–transmission lines analysis for indoor voltage channel,” IEEE Trans. Power Del., vol. 22, no. 4, pp. 2150–2155, Oct. 2007. [12] D. Liu, E. Flint, B. Gaucher, and Y. Kwark, “Wideband AC powerline characterization,” IEEE Trans. Consum. Electron., vol. 45, no. 4, pp. 1087–1097, Nov. 1999. [13] W. Y. Chen, Home Networking Basis: Transmission Environments and Wired/Wireless protocols. Upper Saddle River, NJ: Prentice-Hall, 2004, pp. 103–143, Pro. Tech. Ref.. [14] M. Zimmermann and K. Dostert, “The low voltage power distribution network as last mile access network- signal propagation and noise scenario in the HF-range,” Int.. J. Electron. Commun., vol. 54, no. 1, pp. 13–22, 2000. [15] M. Zimmermann and K. Dostert, “An analysis of the broadband noise scenario in powerline networks,” in Proc. Int. Symp. Powerline Communications Applications, Limerick, U.K., Apr. 2000, pp. 131–138.

Justinian Anatory (S’06) received the B.Sc. and M.Sc. degrees in electrical engineering and the Ph.D. degree in telecommunications engineering from the University of Dar es Salaam, Dar es Salaam, Tanzania, in 1998, 2003, and 2007, respectively. He was a Software and IT Engineer with Beta Communication Consulting Co. (T) Ltd., Dar es Salaam, before joining the University of Dar es Salaam in 2001. Currently, he is a Lecturer with the Faculty of Electrical and Computer Systems Engineering, University of Dar es Salaam. He was a Visiting Researcher in the School of Electrical and Information Engineering, University of Witwatersrand, Johannesburg, South Africa, in 2002 and Visiting Researcher at the EMC Group of the Division for Electricity and Lightning Research of Uppsala University, Uppsala, Sweden, in 2005 and 2006. His research interests include power-line communication, wireless communication, communication networks, and teletraffic engineering.

187

Nelson Theethayi (S’04–M’06) was born in India in 1975. He received the B.E. degree in electrical and electronics (Hons.) from the University of Mysore, Mysore, India, in 1996, the M.Sc.Eng. degree in high-voltage engineering from the Indian Institute of Science, Bangalore, India, in 2001, and the Ph.D. degree in electricity with a specialization in electrical transients and discharges from Uppsala University, Uppsala, Sweden, in 2005. Currently, he is a Researcher with the EMC Group of the Division for Electricity and Lightning Research of Uppsala University. His research areas are electromagnetic compatibility, high-voltage engineering, electrical power systems, modeling and experimental investigation of lightning phenomena and lightning interaction, analysis and design of lightning protection systems for power, and railway and communication systems. Dr. Theethayi is a member of the Technical Committee of Lightning (TC-5) of IEEE-EMC.

Rajeev Thottappillil (S’88–M’92–SM’06) was born in India in 1958. He received the B.Sc. degree in electrical engineering from the University of Calicut, Calicut, India, in 1981, and the M.S. and Ph.D. degrees in electrical engineering from the University of Florida, Gainesville, in 1989 and 1992, respectively. He became an Associate Professor at Uppsala University, Uppsala, Sweden, in 1996 and was promoted to the rank of Full Professor in 2000 in the area of electricity with a special emphasis on transients and discharges at the Division for Electricity and Lightning Research, which is affiliated with the Department of Engineering Sciences at Uppsala University. His research interests are lightning phenomenon, electromagnetic interference, and electromagnetic-field theory. He has published more than 100 scientific articles, of which 40 are in refereed journals. He has also written a book chapter on lightning electromagnetic field computation. Prof. Thottappillil is the Chairman of the EU project COST action P18 “Physics of Lightning Flash and its Effects,” in which groups from 23 countries are involved. He is also a member of SC 77C of SEK, IEC on High Power Transients and the Technical Committee of Lightning (TC-5) of IEEE-EMC.

M. M. Kissaka received the B.Sc. degree in electrical engineering from the University Dare es Salaam, Dare es Salaam, Tanzania, in 1989, and the Ph.D. degree in telecommunications engineering from the University of Manchester, Manchester, U.K., in 1994. Currently, he is Senior Lecturer in the Department of Telecommunications Engineering, Faculty of Electrical and Computer Systems Engineering, University Dare es Salaam. His research interest includes rural telecommunications and computer networks. Dr. Kissaka is a registered professional engineer with the Engineers Registration Board (ERB) of Tanzania.

N. H. Mvungi received the B.Sc. degree in electrical engineering from the University of Dare es Salaam, Dare es Salaam, Tanzania, in 1978. He received the M.Sc. degree in electronics control from Salford University, Salford, U.K., and the Ph.D. degree from Leeds University, Leeds, U.K. He was with Phillips Center for Technology, Eindhoven, The Netherlands. Currently, he is a Senior Lecturer with the University of Dar es Salaam. His research interest is in control and instrumentation, computer communication and applied electronics, lightning protection, rural access, power-quality aspects, and remote monitoring and control of energy consumption.