Radio Path Loss and Penetration Loss. Measurements in and around Homes. and Trees at 5.85 GHz. Mobile and Portable Radio Research Group

Transcription

1 1 Radio Path Loss and Penetration Loss Measurements in and around Homes and Trees at 5.85 GHz Greg Durgin, Theodore S. Rappaport, Hao Xu Mobile and Portable Radio Research Group Bradley Department of Electrical and Computer Engineering Virginia Polytechnic Institute and State University 432 New Engineering Building Blacksburg, VA Phone: (540) EDICS No.: CL1.2.6 Indexing Terms: residential wireless communications, in-building propagation, building penetration, path loss Abstract This paper contains measured data and empirical models for 5.85 GHz radio propagation path loss in and around residential areas for the newly allocated National Information Infrastructure (NII) band in the U.S. Three homes and two stands of trees were studied for outdoor path loss, tree loss, and house penetration loss in a narrowband measurement campaign that included 270 local area path loss measurements and over 276,000 instantaneous power measurements. The data will aid the development of futuristic outdoor-to-indoor wireless unlicensed NII systems (in the U.S.) and HIPERLAN systems (in Europe) for home internet access, telecommunications, and wireless local loops.

2 I. Introduction 2 Residential and campus-wide wireless communication networks may soon proliferate due to recently allocated spectrum for NII systems in the U.S. and HIPERLAN radio spectrum in Europe. Both NII and HIPERLAN frequency bands are in the 5-6 GHz range. Path loss increases at frequencies higher than PCS (1.9 GHz) or cellular (0.9 GHz) systems for propagation into homes and through indoor environments [1], [2]. This paper summarizes an experimental campaign and resulting measurements and models for outdoor path loss, tree and house shadowing loss, and home penetration loss for residential wireless links operating at 5.85 GHz [3]. All path loss values reported in this paper are with respect to 1m free space path loss, which is independent of receiver, transmitter, and antenna gains and losses. Path loss with respect to 1m free space ts into the link budget of Eqn (1): P R = P T + G T + G R ; [Path Loss w.r.t. 1m FS] + 20 log 10 (1) 4 where is wavelength (0.05m at 5.85 GHz), G T and G R are transmitter and receiver antenna gains in db, and P T and P R are transmitter and receiver powers in dbm. II. Experimental Setup Outdoor and indoor path loss measurements were taken at three houses: the Rappaport, Woerner, and Tranter homes. Local area averages of received power, each measured over a 1m area, were used to calculate path loss values in order to eliminate the inuence of small-scale fading. Repeated calibrations of hardware were made at each site to ensure the stability of the measurement system. At each home the outdoor transmitter antenna was placed 30-45m from the house at a height of 5.5m { a typical utility pole height. About 24 measurements were taken along the front and back ofeach house with receivers at heights of 1.5m (head level) and 5.5m. Path loss measurements were then recorded in every room of the house, on the rst, second, and basement levels of each home. Then the outdoor transmitter antenna was moved to a distance of m from the same house and kept at a height of 5.5m and the sequence of measurements outside and inside of the house was repeated. Figure 1 demonstrates the

3 3 dierent receiver-transmitter congurations. A stand of deciduous beech trees and a stand of coniferous pine trees were also measured to determine tree shadowing loss at 5.85 GHz. The transmitter was placed transverse to the line of trees and about 20 measurements were taken along the front and back of the tree line using receiver heights of 1.5m and 5.5m. III. Path Loss Exponents Path loss with respect to 1m free space can be described by a simple distance-dependent model as PL(d) [db] = 10n log 10 (d) (2) where PL(d) is the average path loss value in db at a TR separation of d and n is the path loss exponent that characterizes how rapidly the path loss increases with increasing TR separation [4]. The n in Eqn (2) is the calculated value that minimizes the squared error between measured and predicted path loss in db for a large group of data points. The n value provides a quick estimate of path loss as a function of TR separation for wireless link design. Table I shows dierent n values and measured vs. predicted standard deviations (in db) calculated from data collected at the homes. The path loss exponent for the indoor locations (n=3.4) is clearly larger than for the outdoor locations (n=2.9), due to additional penetration loss into the home. Interestingly, there is no statistically signicant dierence between path loss exponents for 1.5m and 5.5m outdoor receiver heights and for rst and second level indoor locations. IV. Home Penetration Loss We dene aggregate penetration loss (APL) into a home as the ratio between the averaged outdoor and indoor local area powers for a single house with the same transmitter location [1]. Eqn (3) expresses this relationship: 2 1 N Aggregate Penetration Loss [db] = 10 log M NP i=1 P(outside) i MP j=1 P(inside) j (3)

4 4 The summation in the numerator is over the N interior local area power measurements, each denoted as P i, while the summation in the denominator is over the M exterior local area power measurements, each denoted as P j. All powers are in absolute power scale (not db values). Table II provides APL calculated from the three homes. The APL for the Woerner home was nearly 8 db less than for the Tranter and Rappaport homes, implying that homes with brick exteriors induce 8 db more penetration loss than homes with wood siding. Also, the Tranter home exhibited, on average, 4 db more APL than the Rappaport home. Both are brick homes, but the Tranter home has aluminum foil-backed insulation around the entire exterior the insulation at the Rappaport home is paper-backed and less lossy. The linear average value of 16.3 db compares favorably to the median value of 16.1 db reported by [1]. V. Building Shadowing Loss For precise site planning of wireless links, it is often useful to isolate the eects of single shadowing elements. Table III presents rule-of-thumb values for additional path loss induced by specic shadowing objects. The loss values were calculated by studying path loss at locations directly in front of and behind single obstructions, such as trees, houses, or interior walls. For example, a receiver directly behind a house at head level (1.5m height) can expect to experience 24 db more path loss than if it were located on the transmitter-side of the home. Wherever possible, Table III reports the db average of several calculations from the same type of shadowing element for a more reliable estimate. VI. Tree Shadowing Loss Deciduous trees, such as beeches or maples, can be potent shadowers at 5.85 GHz. The wavelength at 5.85 GHz is 5 cm { less than the largest dimension of most leaves. Tree shadowing becomes critical in older neighborhoods, where the canopy is thick and developed and concentrated at rooftop level. In many cases it is easier to propagate underneath the canopy to ground level receivers. This behavior suggests that deciduous trees appear to be \oating masses" and typically introduce 10 to 13 db of loss in excess of free space path

5 5 loss. Thick stands of coniferous trees, such as pines, attenuate a propagating radio wave at 5.85 GHz every bit as much as their deciduous counterparts. Unless intentionally pruned, pine trees grow much thicker at the base than leaf-bearing trees. The measurement results show comparable loss in excess of free space at all receiver heights with typical values ranging from 11 to 16 db. VII. Conclusions This paper presents results of path loss and building penetration loss measurements in residential areas and homes. Detailed measurements were performed in the 5.85 GHz NII band for three typical middle to upper-middle class houses and for deciduous and coniferous stands of trees. Specic eects of foliage shadowing, house shadowing, TR separation, and receiver height were quantied for outdoor path loss in residential areas. The data may be used when designing commercial wireless communications links to the home. References [1] S. Aguirre, L.H. Loew, Lo Yeh. \Radio Propagation into Buildings at 912, 1920, and 5990 MHz Using Microcells". Proceedings of 3rd IEEE ICUPC, pp , Oct [2] P. Nobles, D. Ashworth, F. Halsall. \Propagation Measurements in an Indoor Radio Environment at 2, 5, and 17 GHz". IEE Colloquium on `High Bit Rate UHF/SHF Channel Sounders { Technology and Measurements', London UK, pp 4/1-4/6, [3] G.D. Durgin, H. Xu, T.S. Rappaport. Path Loss and Penetration Loss Measurements in and around Homes and Trees at 5.85 GHz. MPRG TR-97-10, Virginia Tech, 115 pages, 20 June [4] S.Y. Seidel and T.S. Rappaport. \914 MHz Path Loss Prediction Models for Indoor Wireless Communications in Multioored Buildings". IEEE Transactions on Antennas and Propagation, vol 40, no 2, pp , Feb 1992.

6 6 Rx m m 5.5 m 1.5 m Tx1 Tx2 Fig. 1. Transmitters and receivers at dierent heights and separation distances.

7 7 TABLE I Summary of path loss exponents for various transmitter-receiver congurations at 5.85 GHz using 5.5m transmitter height. # of Meas. #of TR Conguration n (db) Locations Homes Indoor Overall First Floor Receiver Second Floor Receiver Outdoor Overall m Receiver m Receiver Rappaport First Floor Receiver Second Floor Receiver m Receiver m Receiver Woerner First Floor Receiver Second Floor Receiver m Receiver m Receiver Tranter First Floor Receiver Second Floor Receiver m Receiver m Receiver

8 8 TABLE II Aggregate penetration loss (APL) values (in db) for all homes at 5.85 GHz using 5.5m transmitter height. Home Exterior Insulation Type TR sep APL (db) Rappaport Brick Paper-backed 30m m 16.4 Woerner Wood Siding Paper-backed 30m m 7.2 Tranter Brick Foil-backed 48m m 15.3 Linear Average 16.3 db Average 14.4

9 TABLE III Rule-of-thumb attenuation values. 9 Loss in excess Shadowing Element of free space Brick house exterior 14.5 db Wood siding exterior 8.8 db Cinderblock wall 22 db Subterranean basement loss 31 db Interior wall 4.7 db Small deciduous tree 3.5 db Large deciduous tree 11 db Large coniferous tree 14 db Close-in house shadowing (1.5m RX height) 24 db Close-in house shadowing (5.5m RX height) 16 db