64 A. MUQAIBEL, N. IYA, U. JOHAR, WALL COMPENSATION FOR ULTRA WIDEBAND APPLICATIONS Wall Compensation fo Ulta Wideband Applications Ali H. MUQAIBEL, Nuuddeen M. IYA, Uma M. JOHAR Dept. of Electical Engineeing, King Fahd Univ. of Petoleum & Mineals, P.O. Box 1734, Dhahan 3161, Saudi Aabia muqaibel@kfupm.edu.sa, nuuddeeniya@yahoo.com, umjoha@kfupm.edu.sa Abstact. Due to thei low fequency contents, ulta wideband (UWB) signals have the ability to penetate walls and obstacles. As the signal popagates though these obstacles, it gets attenuated, slows down, and gets dispesed. This pape demonstates wall compensation fo thoughwall imaging, localization and communication eceive design puposes by fist chaacteizing wave popagation though vaious building mateials in the UWB fequency ange. Knowledge of the walls obtained fom the wall chaacteization is used to estimate and coect the position accuacy of a taget object located behind the walls using thee poposed methods namely; constant amplitude and delay (CDL), fequency dependent data (FFD), and data fitting methods (FIT). The obtained esults indicated elatively acceptable measue of wall compensation fo the thee methods. Results fom such wok povide insight on how to develop algoithms fo effective taget position estimation in imaging and localization applications. They ae also useful fo channel modeling and link budget analysis. Keywods Wall compensation; localization; UWB; though-wall imaging. 1. Intoduction Electomagnetic waves passing though a medium ae subject to amplitude and phase distotions. These distotions ae attibuted to dispesive and attenuative popeties of the medium of popagation. A popagation path obstuction is defined as a man-made o natual physical object that lies close enough to a adio wave path to cause a measuable effect on the path loss exclusive of eflection effects [1]. Thee is an inceasing need to undestand and model electomagnetic effects associated with wave popagation though obstacles. In addition to mateials that make the wall, the shape of the wall, its composition, multiple eflections within the wall, angles of incidence on the wall, coupling effects, adiation patten, and polaization of tansmit and eceive antennas ae impotant factos that need to be taken into consideation. Seveal studies have been conducted on the electomagnetic chaacteization of building mateials both in the naowband [] [5], and wideband [6] [7] anges of fequencies. Electomagnetic paametes including insetion loss, dielectic constant, loss tangent, eflection and tansmission coefficients wee used to pesent wall chaacteistics. Reseach has been done to popose methods to mitigate the effect of walls on signal popagation and taget detection. Chanda et al. [8] applied a singula value decomposition algoithm to minimize clutte and detect a metallic taget behind plywood and bick wall in the UWB fequency ange. In [9], Guolong et al. illustated the impact of delay in position accuacy and used a thoughthe-wall compensation algoithm to coect the position of the located human to within 4 cm. Ahmad et al. [1] demonstated the imaging poblem with pactical assumptions of unknown wall paametes. They poposed an autofocusing technique based on highe ode statistics that povides high quality images with locations close to tue taget locations. Rovnakova et al. [11] poposed two methods to coectly tace moving tagets behind walls and compensate fo the wall effect. Imaging accuacy, howeve, is not only dependent on signal pocessing but also on the availability of detailed infomation about buildings, which includes mateial constants (pemittivity and conductivity), thicknesses of walls, as well as the stuctues of the buildings themselves [1]. In this pape, we investigate the obstuction effect and popose methods to compensate fo the effect. This is achieved by fist pefoming a wideband electomagnetic chaacteization of typical building wall mateials and assessing thei impacts on localization. Measuements ae caied out on samples of wood, glass and gypsum walls to chaacteize them ove a fequency ange of 1 18 GHz using a vecto netwok analyze. The chaacteization method is based on measuing the insetion tansfe function, defined as the atio of two signals measued in the pesence and in the absence of the wall. The dielectic constant of the wall mateial is elated to the measued insetion tansfe function though a complex tanscendental equation that can be solved using an appoximate one-dimensional oot seach [13]. Tansmission and eflection measuements wee caied out in fequency domain using fee space adiated measuement to extact the insetion loss and dielectic constant fo each wall mateial. Fee space adiated measuements ae contactless and nondestuctive. They also match the configuation of the final application fo which the measuements ae caied out, which is ada and communication.
RADIOENGINEERING, VOL. 1, NO., JUNE 1 641 While penetating though a mateial, an electomagnetic wave fo though wall imaging and detection may change its speed significantly. This is closely elated to the wall thickness and composition, its dielectic constant, and the angle of incidence of the wave. In addition to slowing down, the signal gets attenuated and undegoes efaction as it passes though the wall. It defocuses taget image and displaces the taget fom its tue position. False tagets can also be pesent in the ada images. These effects ae moe ponounced fo walls with highe dielectic constants o in pesence of multiple walls between taget and ada [1]. Wall compensation theefoe, is when effot is being made to coect the advese effect of the wall on the outcome of the detection pocess so that the tue taget position can be obtained. This wok attempts wall compensation by using wall infomation obtained fom the wall chaacteization to coect the position estimation of the taget object. This is achieved though conducting taget expeiments. When compensating fo the wall effect, we eithe use full fequency dependent data o appoximate fits. Thee possibilities ae poposed and compaed. The fist method employs constant amplitude loss and delay suffeed by the signal as obtained fom pevious wall chaacteization measuements to compensate fo the wall effect in the taget measuements. The second method uses peviously known fequency dependent wall infomation epesented by magnitude and phase data, while the thid method which is a tade-off between the fist two uses linea and quadatic fits to magnitude of the insetion loss and dielectic constants of the wall espectively. This pape is oganized as follows. Section summaizes the wall chaacteization pocess and esults. Section 3 descibes the expeimental setup and pocedue fo the taget measuements. In section 4, the poposed wall compensation methods ae explained and the esults fom each method ae illustated. Finally, concluding emaks about the wok ae given in section 5.. Wall Chaacteization Both magnitude and phase infomation ae needed fo accuate chaacteization of walls. Using the fequency domain technique, complex data points epesenting magnitude and phase infomation ae obtained. These data ae obtained using vecto netwok analyze. The insetion tansfe function is obtained as the atio of two tansmit signals given as [5] Xt ( j) H( j) (1) fs X ( j) whee X fs t ( j) and X t ( j ) ae fee-space efeence and though-wall fequency domain signals obtained in the absence and pesence of the wall espectively. The fee-space measuement is used as efeence to take cae of the effects of components othe than the wall. Exta t caution should be taken to make sue that exact setup is used to pefom both expeiments in ode to avoid measuement inconsistencies. Assuming a fictitious laye of fee-space of the same thickness as the wall, then the popagation delay though this laye is τ = d/c whee d is the laye thickness and c is the speed of light in fee space. The scatteing paamete elated to the insetion tansfe function in this case is given as S 1 ( j ) j H ( j ) e. () The un-gated insetion tansfe function is obtained by dividing the though wall S 1 data by that of fee space. A finite impulse esponse filte is then used to emove the noise at the low fequencies and those beyond the antenna bandwidth. The fequency domain signals ae then conveted to time domain using invese fast Fouie tansfom to get the impulse esponses. Zeos ae padded fo highe time domain esolution. The impulse esponses obtained fom fequency-domain measuements ae coelated using a sliding coelato to obtain the fist guess on the delay and effective dielectic constant. An estimate of the aveage dielectic constant could also be obtained though peak-to-peak impulse time delay, Δτ. This aveage dielectic constant, which does not eflect the fequency dependence, is given by ' 1. (3) d c As a function of the unwapped phase, Φ sp (f), the dielectic constant is also given by Insetion tansfe function Dielectic constant ( f ) sp ( f ) ( f ) 1 1 f. (4) Coefficients of linea o quadatic fits to extacted paametes af + b o af + bf + c, f (GHz) Wood Glass Gypsum a b c a b c a b c -.45 -.785 -- -.44-1.687 -- -.175 -.46 --.16 -.56 3.157 -.91 8.396 --.8 -.57.83 Tab. 1. Coefficients fo extacted paametes. In the absence of an anechoic chambe, multipath components, multiple eflections in the wall, and eflections fom the floo, ceiling and othe stuctues become a theat to the measuement accuacy. In ode to educe this effect of multipath, time gating is used to selectively
64 A. MUQAIBEL, N. IYA, U. JOHAR, WALL COMPENSATION FOR ULTRA WIDEBAND APPLICATIONS emove o include the undesied esponses in time. The emaining time domain esponses can then be tansfomed back to the fequency domain with the effect of the gatedout esponses being emoved. The delayed signals can be filteed out by imposing a window function ove the dominating signal that is identified to be the desied one. The gated insetion tansfe function is then obtained fom which the dielectic constant is calculated using the multiple pass technique fom [13]. Fig. 1 shows the esults fo the insetion tansfe function and dielectic constants espectively fo wood, glass and gypsum walls. The coefficients fo these esults ae shown in Tab. 1. A tansmitted signal fom the tansmitte is eflected by the aluminum and eceived by the eceiving antenna. This is the efeence measuement o Taget Only. A wall is then inseted between the antennas and the taget with some distance on both sides as shown in Fig. and anothe measuement is caied out. The steps above wee epeated fo wood, glass, gypsum, and some multiple wall combinations. In each case, fequency domain esponses ae acquied. Fig.. Though taget measuements setup. Fig. 3 shows the pocessed bandlimited time-domain impulse esponse fom the taget both with and without a wood wall. Thee ae two main eflections in the case of the taget behind the wall. The fist (ealy) esponse (solid line) indicates a eflection fom the wall suface. We can clealy see a delayed and attenuated esponse fom the taget when it is behind the wall epesented by the second eflection. This delay tanslated into distance, coesponds to shift in the actual position of the taget behind the wall. The dotted line epesents the Taget Only measuement. Fig. 1. Magnitude of insetion tansfe function (uppe). Dielectic constant (lowe). 3. Measuement Setup and Pocedue An HP851C vecto netwok analyze with two-pot S-paamete test set was used to pefom taget measuements within the band of inteest. The output of the netwok analyze pot 1 is fed though a 1.5 m cable to a wideband powe amplifie; anothe 1.5 m cable connects the amplifie output to a wideband hon antenna mounted on a tipod. An identical antenna mounted on a simila tipod stand is used as a eceiving antenna. The eceive antenna output is fed to the netwok analyze pot. An aluminum sheet is used as the taget object because it appoximately eflects all the enegy impinging on its suface. The pocedue is as follows. The aluminum sheet (taget) is put in font of both antennas at a given distance. Fig. 3. Reflections fom taget object with and without obstuction. 4. Wall Compensation Methods The objective is to compensate fo the taget displacement by emoving the effect of the wall on the esponse obtained fom the taget. This is achieved in thee diffeent methods: (1) Using constant amplitude and delay (CDL), () Using full fequency dependent aw data (FFD), and (3) Using fitted dielectic constant and fits to the magnitude of fequency dependent data (FIT).
RADIOENGINEERING, VOL. 1, NO., JUNE 1 643 4.1 Constant Amplitude and Delay Compensation In the constant amplitude and delay compensation, it is assumed that thee is constant amplitude attenuation due to the wall. It is also assumed that the delay is constant and thus, thee is no fequency dependence. Fo each wall, we use the amplitude attenuation it incued duing wall chaacteization as constant amplitude to compensate fo the amplitude loss the same wall suffeed in the taget measuements. A simila appoach is pefomed on the delay, and the time is coected by a value equal to the delay the wall offeed duing the wall chaacteization measuements. Tab. shows the constant values used fo amplitude and delay compensation. Fig. 4 shows the compensated signal using this appoach compaed with the esponse obtained without a wall fo the case of wood, glass, and gypsum. The constant amplitude used infomation fom tansmission measuements, while the constant delay used is twice the value of the delay fom tansmission measuements. This is because the eflection measuement equies the signal to popagate twice though the wall. 4. Fequency Dependent Data Method Fo moe accuate detemination of the taget s position, the fequency dependent effect of the wall has to be emoved. This is achieved by dividing total tansfe function by that of the wall. This is achieved using ou pevious knowledge of the wall obtained fom wall chaacteization. Fig. 5 shows the esult fo wood both in fequency and time domain. Simila esults wee obtained fo othe walls but ae not pesented hee due to the limited space. It should be noted that because of the aw fequency domain data diectly obtained fom measuements, the pesence of noise is a majo concen. Noise clealy manifests in the time domain. Theefoe, noise eduction is pefomed by filteing the noisy egions aound 1 GHz and 15.5 GHz as in Fig. 5. Wall Constant amplitude Constant delay (ns) Wood 1.44.8513 Glass 1.4.864 Gypsum 1.14.69 Tab.. Constant amplitude and constant delay values used. 4 x 1-4 Taget Only Taget + With Wall Compensated Amplitude, V - -4 Amplitude, V 6 6.5 7 7.5 8 8.5 9 9.5 Time, sec 4 x 1-4 - -4 x1-9 Taget Only Taget + With Wall Compensated -6 7 7.5 8 8.5 9 9.5 Time, sec x 1-9 x 1-4 5 Taget Only Taget + With Wall Compensated Fig. 5. Wall compensation using aw data fo wood sample in fequency domain (uppe), and time domain (lowe). Thee is a close ageement between the compensated esponses and the esponses obtained fom measuements pefomed without the wall. The advantage of this method ove compensating with a constant delay and amplitude is that the pulse shape is coected. In application like matched filte o coelation eceives, pulse shape coection is vey impotant. The cost will be to povide a full data about the insetion tansfe function of the wall. Amplitude, V -5 7 7. 7.4 7.6 7.8 8 8. 8.4 8.6 8.8 9 Time, sec x 1-9 Fig. 4. Illustating compensation using constant amplitude and constant delay, wood (uppe), glass (middle), gypsum (lowe). 4.3 Data Fitting Method As a tade-off between the above two methods, one can povide the paametes that model the magnitude of the insetion loss of the wall using a linea equation to eflect the fequency dependence of the magnitude of the tansfe function. Highe ode fits ae also possible but not needed especially if the walls used ae having some vaiability.
644 A. MUQAIBEL, N. IYA, U. JOHAR, WALL COMPENSATION FOR ULTRA WIDEBAND APPLICATIONS In this method, the esult of the dielectic constant given in Tab. 1 is used to obtain an un-wapped phase Φ W fom (4). The delay τ o is still modeled with a constant value as in the case with constant amplitude and delay method. The phase obtained is then used togethe with the esults of the magnitude of insetion tansfe function (having coefficients in Tab. 1, to calculate a complex insetion tansfe function HW ( f ) HW ( f ) expjw ( f ) epesenting the wall. The esult fo compensation using this method fo wood and glass is shown in Fig. 6. Wood shows good wall compensation as indicated by the similaity between the Taget Only and Compensated esponses. The compensation fo glass did not show good esults. wall of dielectic constant ', the speed of light, c = 3. 1 8 and Δτ is the peak-to-peak time delay incued by the signal passing though the wall. The dielectic constant ' can be a value chosen at the mid-fequency ange fo the wall mateial unde consideation. Tab. 3 povides eos in delay coesponding to appoximate eos in the position of the taget due to the thee diffeent walls. It also shows appoximate eos afte compensation has been done fo the thee methods. Fo the pupose of Tab. 3, we efe to the thee methods with the following abbeviations in paenthesis: Constant Amplitude and Delay (CDL), Fequency Dependent Data (FFD), Fitted Dielectic constants and Fits to Magnitude of Fequency Dependent Data (FIT), and No Compensation as (NC). Peak-to-peak eo (ns) Due to wall Afte compensation Wall Thickness (cm) ' NC CDL FFD FIT Wood 1.8 3..794 -.1489 -.645.1161 Glass.8 6.5.1577.1936.193 -.66 Gypsum 1..4.816.15.34 -.116 Fig. 6. Wall compensation using fit to data fo wood (uppe), and glass (lowe) 4.4 Pefomance Evaluation fo the Suggested Wall Compensation Methods In ode to assess the accuacy of the methods mentioned above in achieving wall compensation, we will make compaison between esults fo No Wall and that of Compensated (estimate). We also compae these esults with the case whee thee is no compensation. The compaison will be in tems of the delay (peak-to-peak) and enegy captue. The peak-to-peak delay is useful in positioning applications and can be used to give appoximate eos in the tue position of detected objects. Since time and distance ae elated by the expession whee ' s v (5) v c is the velocity of the wave though the Tab. 3. Constant amplitude and constant delay values used. Since time domain esponses ae used to demonstate wall compensation, the peak-to-peak delay was used to find the eo due to the wall by compaing the Taget Only and Taget +Wall esponses. Similaly, the peakto-peak delay fo the Taget Only and Compensated esponses ae used to get the eo afte compensating fo the wall. An equivalent eo in displacement is also obtained. The negative sign in the eos indicate instances whee we have ove-compensation ; theefoe, the peakto-peak diffeence will be negative. Note that these figues depend on the dielectic constant and thickness of the walls. Studying Tab. 3 closely, we can see that elative to no compensation at all (NC), wood pefom bette afte FFD compensation as indicated by.64 ns eo in delay compaed to.148 ns and.116 ns fo CDL and FIT espectively. CDL gives.193 ns eo afte compensation fo glass which is less compaed to that of FFD and FIT. Because peak-to-peak delay is used, changes in pulse shapes of the compensated esponses fo FFD and FIT will affect the amount of eo both in time and space. We also attibute this to the high dielectic constant of glass, even though the thickness of ou glass wall is.8 cm. Recall that walls with highe dielectic constant tend to defocus taget images intoducing eos in taget positions [9]. This will yield up to.71 cm eo FIT compensation. Fo gypsum, FFD has the highest amount of eo afte compensation with.34 ns elative to NC, followed by CDL with.15 ns. FIT gives the least. Ove-
RADIOENGINEERING, VOL. 1, NO., JUNE 1 645 all, CDL compensation seems to give less eo fo the thee walls and wood gives a bette eo pefomance fo all thee methods. Enegy captue is useful when analyzing signal wavefoms. Fo ou pupose, we employ the enegy captue equation descibed in [13] to be EC ( t) ( t) c 1 ( t) 1% whee (t) is the Taget Only signal called the tue signal, and c (t) is the Compensated signal and it is called the estimate of the tue signal. A window size of 1 ns aound the main pulses is assumed. To achieve full compaison between the shapes of the wavefoms, they ae fist synchonized, and then the estimate is subtacted fom the tue signal to get the diffeence. If the diffeence is geate than the tue signal (t), the pecentage enegy captue will be negative. Tab. 4 shows how simila, in pecentage, the wall compensation esults ae to the esults obtained fom measuements without the wall fo wood, glass and gypsum, using Constant Amplitude and Delay Method, Fequency Dependent Data Method, and Data Fitting Method. We have added a column epesenting when thee is no compensation (between Taget Only and Taget + Wall ). It should be noted that the accuacy of these figues depends on the size of the window aound the main pulse ove which the compaison is made. Wood and gypsum showed good compensation paticulaly in the data fitting method. This is indicated by the 99.3% similaity to the No Wall case fo wood and 98.7% fo gypsum. Fo glass, we have up to 16% impovement in the esults fom the fequency dependent data method (96.45%) ove constant amplitude and delay s 8.68%. Howeve, in the data fitting method, compensated wavefom fo glass suffeed consideable distotion. Oveall, the esults fo enegy captue in fequency dependent data method show elatively bette similaity between Taget Only and Compensated esponses. (6) Geneally, the basis of compaison depends on how the eceive is inteested in the signal (time, enegy, etc). Fig. 7 shows the thee diffeent methods compaed fo ou wood wall. As with the matched filte system, the use of pulse peak-to-peak delay to measue the coelation between the two esults might be an easie appoach to take howeve; in this case, the shape of the wavefom is affected duing the compensation pocess. Theefoe, finding the coect peak of the signal will not be easy and the peakto-peak delay will yield an uneliable outcome which cannot epesent coelation o similaity. Consequently, wall compensation fo double and tiple walls suffeed sevee change in wavefom, thus, no eliable esults wee obtained fo them. Fig. 8 shows an example fo the case of double wall: wood-gypsum. Fig. 7. Compensation using the thee methods fo wood wall. Method Wall No Compensation Constant Amplitude and Delay Fequency Dependent Data Data Fitting Wood 83.94 % 93.5 % 95.61 % 99.3 % Glass 71.5 % 8.68 % 96.45 % 76.8 % Gypsum 94.85 % 96.7 % 98.98 % 98.7 % Tab. 4. Pecentage similaity of wall compensation esults to No Wall esults. Fig. 8. Wall compensation fo double wall (wood-gypsum). Fequency domain (uppe), time domain (lowe).
646 A. MUQAIBEL, N. IYA, U. JOHAR, WALL COMPENSATION FOR ULTRA WIDEBAND APPLICATIONS 5. Conclusion This pape poposes a way of compensating fo the effect the obstuction has on the tue position of a taget in though wall detection applications. Thee methods fo wall compensation wee discussed. The fist one used estimated constant amplitude loss and delay values suffeed by the walls in the pevious wall chaacteization expeiments to pefom compensation, while the second uses full fequency dependent data fom pevious knowledge of the wall. The thid uses a quadatic and linea fit to peviously obtained dielectic constant and magnitude of insetion loss espectively, while assuming a constant delay. Results obtained show a good level of wall compensation fo the diffeent walls used. Based on enegy captue, esults obtained show a good level of wall compensation fo the diffeent walls used. With the fist method, up to 93.5% similaity was ecoded between compensated and no wall esponses fo wood. Similaly, 98.7% similaity was obtained with the thid method fo gypsum wall. Acknowledgment The authos would like to acknowledge the suppot povided by the Deanship of Scientific Reseach at King Fahd Univesity of Petoleum & Mineals (KFUPM) and the King Abdul-Aziz City of Science and Technology (KACST) unde Reseach Gant # NSTP: 8-ELE44-4-1. The authos would like also to thank D. Mohamed Adnan Landolsi, and D. Abdullah Al-Ahmai fo thei valuable input. Refeences [1] TESSERAULT, G., MALHOUROUX, N., PAJUSCO, P. Detemination of mateial chaacteistics fo optimizing WLAN adio. In Poceedings of the 1th Euopean Confeence on Wieless Technology. Munich (Gemany), 7, p. 5 8. [] OUSLIMANI, H. H., ABDEDDAIM, R., PRIOU, A. Fee-space electomagnetic chaacteization of mateials fo micowave and ada applications. Pogess in Electomagnetics Reseach Symposium. Hangzou (China), 5, p. 18 13. [3] AKUTHOTA, B., ZOUGHI, R., KURTIS, K. E. Detemination of dielectic popety pofile in cement-based mateials using micowave eflection and tansmission popeties. In Instumentation and Measuement Technology Confeence. Como (Italy), 4, p. 37 33. [4] CUINAS, I., SANCHEZ, M. Measuing, modelling, and chaacteizing of indoo adio channel at 5.8 GHz. IEEE Tansactions on Vehicula Technology, 11, vol. 5, no., p. 56 535. [5] MUQAIBEL, A., SAFAAI-JAZI, A., BAYRAM, A., ATTIYA, A., RIAD, S. Ultawideband though-the-wall popagation. IEE Poceedings on Micowave Antennas Popagation, 5, vol. 15, no. 6, p. 581 588. [6] LIU, C., HUANG, C., CHIU, C. Channel capacity fo vaious mateials of patitions in indoo ulta wideband communication system with multiple input multiple output. In 3d IEEE UZ Regional Chapte Intenational Confeence in Cental Asia on Intenet, Taskent (Uzbekistan), 7, p. 1 5. [7] CHANDRA, R., GAIKWAD, A., SINGH, D., NIGAM, M. An appoach to emove the clutte and detect the taget fo ultawideband though-wall imaging. Jounal of Geophysics and Engineeing, 8, vol. 5, p. 41-419. [8] CUI, G., KONG, L., YANG, J., WANG, X. A new wall compensation algoithm fo though-the-wall ada imaging. In 1st Asian and Pacific Conf. on Synthetic Apetue Rada, 7, p. 393 396. [9] AHMAD, F., AMIN, M. G., MANDAPATI, G. Autofocusing of though-the-wall ada imagey unde unknown wall chaacteistics. IEEE Tansactions on Image Pocessing, 7, vol. 16, p. 1785-1795. [1] ROVNAKOVA, J., KOCUR, D. Compensation of wall effect fo though-wall tacking of moving tagets. Radioengineeing, 9, vol. 18, no., p. 189 195. [11] YI, H., PARSONS, D. A time domain appoach fo measuing the dielectic popeties and thickness of walls of a building. In IEE Colloquium on Popagation aspects of Futue Mobile Systems. 1996, p. 7/1 7/7. [1] MUQAIBEL, A., SAFAAI-JAZI, A. A new fomulation fo chaacteization of mateials based on measue insetion tansfe function. IEEE Tansactions on Micowave Theoy and Techniques, 3, vol. 51, no. 8, p. 1946 1951. [13] WIN, M., SCHOLTZ, R. Enegy captue vs coelato esouces in ulta-wide bandwidth indoo wieless communications channels. In MILCOM 97 Poceedings. 1997, vol. 3, p. 177 181. About Authos... Ali MUQAIBEL was bon in Dammam, Saudi Aabia in 1974. He eceived his B.Sc. and M.Sc. fom King Fahd Univesity of Petoleum & Mineals (KFUPM) in 1996 and 1999, espectively. He obtained his Ph.D. fom Viginia Polytechnic Institute and State Univesity (Viginia Tech) in 3. He is cuently an Associate Pofesso at the Electical Engineeing Depatment at KFUPM. His eseach inteests include Ulta Wideband (UWB) communications, channel modeling, positioning, and compessive sensing techniques. He is the autho o co-autho of about 4 papes published in scientific jounals o confeence poceedings. Nuuddeen IYA was bon in Yola, Nigeia. He obtained his M.Sc. fom King Fahd Univesity of Petoleum & Mineals (KFUPM) in 1996. His eseach inteests include ulta wideband (UWB), and though wall popagation. Uma JOHAR was bon in Febuay 1967. He eceived his B.Sc. and M.Sc. degees fom the King Fahd Univesity of Petoleum & Mineals (KFUPM), Saudi Aabia in 199 and 1993 espectively. He woked as a Reseach Assistant in the Electical Engineeing Depatment at KFUPM fom Novembe 199 to Januay 1993. In Febuay 1993, he joined the same depatment to wok as a Lectue whee he is still employed. His eseach inteest includes electomagnetics, fibe optics, micowave engineeing, wieless communications and UWB.