IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 9, 2010

867

Path Loss Characteristics for IMT-Advanced Systems in Residential and Street Environments Han-Shin Jo and Jong-Gwan Yook, Member, IEEE

Abstract—This letter presents the measured path loss characteristics in microcellular residential and street environments at 3.4-, 5.3-, and 6.4-GHz band signals. Path loss characteristics on the different residential areas are thoroughly investigated and compared to a modified Hata model, and a new path loss model based on the modified Hata model is developed from the measurement of the residential area. A two-ray model is applied to analyze the path loss characteristics in a line-of-sight (LOS) street. Moreover, in a non-line-of-sight (NLOS) street, the power level decreases due to corner loss, and path loss exponents were investigated. Index Terms—International Mobile Telecommunications (IMT)Advanced band, macrocellular residential area, path loss, street microcell.

I. INTRODUCTION

C

URRENT mobile communication systems have evolved by adding numerous system capabilities and enhancements, and users will see a significant increase in capability through the future development of third-generation mobile communication systems [1]. Taking into account sufficient mobility and an acceptable tradeoff between cost and full area coverage, it is suggested that the suitable frequency range for International Mobile Telecommunications (IMT)-Advanced services is below 6 GHz [2]. In this letter, signal strength of 3.4, measurements were conducted at the frequencies 5.3, and 6.4 GHz in order to characterize the behavior of narrowband signals in outdoor microcell residential and street environments. The characteristics of path loss on the different residential areas were thoroughly investigated and then compared to a modified Hata model (COST-231 Hata model) [3]. We propose a new path loss model, which is revised from the modified Hata model and is suitable for the residential area and the frequency range considered in this letter. A two-ray model was applied to analyze the breakpoint as well as the upper and lower bounds of the line-of-sight (LOS) path loss. In

Manuscript received April 08, 2010; revised June 03, 2010 and July 19, 2010; accepted August 14, 2010. Date of publication August 26, 2010; date of current version September 23, 2010. This work was supported by the Ministry of Knowledge Economy (MKE), Korea, under the Information Technology Research Center (ITRC) support program supervised by the National IT Industry Promotion Agency (NIPA) [NIPA-2010-(C1090-1011-0006)]. H.-S. Jo is with the Department of Electrical and Computer Engineering, The University of Texas at Austin, Austin, TX 78712-0240 USA (e-mail: han-shin. [email protected]). J.-G. Yook is with the Department of Electrical and Electronic Engineering, Yonsei University, Seoul 120-749, Korea (e-mail: [email protected]). Color versions of one or more of the figures in this letter are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/LAWP.2010.2070482

Fig. 1. Measurement environment. (a) Residential area A. (b) Residential area B. (c) LOS street. (d) NLOS street.

a non-line-of-sight (NLOS) street, the power level drops due to corner loss, and path loss exponents were investigated. II. MEASUREMENT SYSTEM AND SCENARIOS The transmitter is composed of a continuous wave (CW) generator, power amplifier, and antenna. The receiver comprises an antenna, low noise amplifier (LNA), and spectrum analyzer. The receiving antenna is connected to the LNA, whose output is connected to the spectrum analyzer. The transmit and receive antenna is formed by identical half-wavelength dipoles featuring vertical polarization. A GPS unit with maximum error of 5 m is used to establish the measurement location. Measurements are conducted in residential and LOS and NLOS street environments in Seoul and Bun-dang in South Korea. Fig. 1 shows the measurement area, where the transmitter position and measurement route are represented by a point marked with “BS” and a dashed line, respectively. For the residential area Bun-dang (area A), the transmit antenna with a height of 19 m is surrounded by houses with a height of 15 m. In the residential area of Seoul (area B), the transmit antenna with a height of 50 m is surrounded by apartment buildings with a height of 46 m. The transmit antennas are located at a higher position than neighboring buildings for both the residential areas. On the other hand, for the measurement in the LOS and NLOS street environments, the transmit antenna height (3 m) is lower than the height

1536-1225/$26.00 © 2010 IEEE

868

IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 9, 2010

of surrounding buildings. For all measurement scenarios, a receive antenna is mounted on the roof of a vehicle at a height of 2 m. The receiving vehicle traveled on streets with an average speed of approximately 6 m/s, and six receiving position data are recorded per 1 s at the GPS unit. Signal power measured satisfying the Lee’s guidelines for sample averaging (50 points per 40 wavelengths) [4]. The recorded data are streamed to a computer hard disk for later processing. The measurement is repeated three times and averaged in order to get a uniform number of position data with respect to a distance from the transmitter as well as to reduce the GPS location error. The measured data are time-averaged to remove the effect of short-term fading. III. PATH LOSS CHARACTERISTICS A. Residential Area In the wireless radio propagation channel, received signal power fluctuates because signal propagates through multipaths due to reflection, diffraction, and scattering at objects. When the received power is averaged over a sufficiently long time period so that the fading is averaged out, path loss can be modeled as (1) where denotes the path loss exponent indicating the rate at which the path loss increases with respect to distance, denotes the close-in reference distance given from a measurement close to the transmitter, and represents the distance between transmitter and receiver. The shadowing is modeled as in decibels with a a zero-mean Gaussian random variable standard deviation in decibels. The measured path loss characteristics for areas A and B are presented in Figs. 2 and 3. The path loss distance dependency can be obtained by fitting the measured data with a least-square (LS) regression curve. The modified Hata models of the suburban are plotted with dashed line (1) as compared and , which to the measured path loss. The parameters characterize the path loss in tested areas, are given in Table I. No frequency dependence of and was found. The path loss exponents in area A are 3.1–3.2, which are smaller than those (3.5–3.9) in area B. This is clearly explained from the , COST-231 Walfish–Ikegami model [3]; a smaller denotes the building height and represents the where difference between the building height and transmit antenna height, results in a steeper slope (path loss versus distance). The standard deviation is observed between 6.6 and 7.7 dB, which is similar to those typically encountered in cellular radio [5]. As the modified Hata model is designed for the 1500–2000 MHz frequency range, it does not match the LS regression curve as shown in Figs. 2 and 3; it overestimates path loss for all cases. Furthermore, [3] shows that path loss is highly dependent on local topography. In particular, the lower the building density, the larger the difference between the building height and transmit antenna height, or a smaller difference between the building height and receive antenna height can cause the measured path loss to be less than that predicted by the modified Hata model. Thus, we proposed a new path loss model that is revised from the modified Hata model as follows:

Fig. 2. Measured path loss in area A. The dashed line (2) is given by Eq. (2). (a) f : GHz. (b) f : GHz. (c) f : GHz.

=34

=53

=64

(2)

for suburban areas

(3)

and are respectively the height (in meters) of the where transmit and receive antennas. is the carrier frequency in megahertz, and is the distance (in kilometers) between the transmitter and receiver. , , , and are conventional paand are additional rameters of the modified Hata model. parameters for the proposed model, which is given as for Area A for Area B for Area A for Area B

(4)

JO AND YOOK: PATH LOSS CHARACTERISTICS FOR IMT-ADVANCED SYSTEMS IN RESIDENTIAL AND STREET ENVIRONMENTS

Fig. 3. Measured path loss in Area B. The dashed line (2) is given by Eq. (2). : GHz. (b) f : GHz. (c) f : GHz. (a) f

=34

=53

=64

TABLE I PATH LOSS EXPONENT n AND STANDARD DEVIATION 

Fig. 4. Measured path loss in LOS street. (a) f (c) f : GHz.

= 64

869

= 3:4 GHz. (b) f = 5:3 GHz.

model [6]–[9]. Specifically, [7] proposed upper and lower bounds based on a two-ray model, which are validated in the 1850–1900 MHz range and given as for for for for

Here, and respectively compensate the path loss differences caused by transmit antenna height and carrier frequency. less than zero indicates that the modified Hata model The overestimates path loss at the frequency band considered in this study. As shown in Figs. 2 and 3, our models well fit the LS regression curve. B. LOS Street Some empirical measurements have shown that LOS propagation along city streets is accurately described by the two-ray

(5) (6)

where breakpoint distance is given as a function of wavelength and the height of transmit and receive antennas and as follows: (7) is given by Additionally, the loss at . Fig. 4 depicts the measured path loss and the bounds given in (5) and (6). It is observed that the measured path loss is mostly

870

IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 9, 2010

TABLE II BREAKPOINT DISTANCE, R

TABLE III PATH LOSS EXPONENT n AND CORNER LOSS

Fig. 6. Reflected (solid line) and diffracted (dashed line) rays and reflected rays entering the NLOS street.

Fig. 5. Measured path loss in NLOS street. (a) : GHz. : GHz. (c) f

53

=64

f = 3:4 GHz. (b) f =

bounded by the bounds for all frequencies. Measured and theoretic breakpoints are compared in Table II. The measured breakpoint is given by calculating minimum square error between measured path loss and broken regression line [8]. The measured breakpoint is maximally 6% larger than the theoretic one given by (7). This is because of a rising reflection area due to pedestrians and vehicles [9]. The measurement results indicate that the conventional two-ray model is still valid for candidate bands of IMT-Advanced. C. NLOS Street: Around Street Corners In this measurement scenario, the receiver communicates via LOS with the transmitter until it reaches the street corner,

and the NLOS propagation is obtained after rounding the street corner. The LOS distance between the transmitter and the corner, , is 80 and 130 m. Fig. 5 shows the scatter plot of measured path loss and the straight line that represents the LS regression fit. The slope of the linear curve corresponds to the path loss exponent , which is presented in Table III. Increasing increases , which is analyzed using the effect of reflection and diffraction [10]–[13]. As shown in Fig. 6, the incidence angle is larger for a larger , which results in the rays being reflected more frequently along the NLOS distance . These multiple reflections increase . Moreover, Fig. 6 shows four diffracted rays (dashed line). Erceg et al. [11] demonstrate that the corner-1 diffracted ray is stronger contribution than other rays at small , and the corner-4 diffracted ray becomes dominant at large . In [12], diffraction loss of the corner-1 diffracted rays first increases and then decreases as increase, i.e., as increases within . Furthermore, in [12], diffraction loss of the corner-4 diffracted rays decreases increase, i.e., as increases within . as Therefore, at larger , where the corner-4 diffracted ray is more dominant than the diffracted rays at other corners, the received power of the diffracted rays decreases more rapidly with . of 80 and From Table III, is 7.2–8.1 and 12.1–13.6 for 130 m, respectively. Moreover, a previous study [13] shows of 64 and 429 m, respecthat is 4.1–4.9 and 12–28 for tively. The comparison between the two measurement results increases. Fig. 5 also confirms that the increases as the shows a sudden power level decrease after turning the corner increases. This is because (corner loss), which increases as the number of rays entering the out-of sight street decreases as increases. No frequency dependence of path loss exponent and corner loss was found; street geometry is dominant.

JO AND YOOK: PATH LOSS CHARACTERISTICS FOR IMT-ADVANCED SYSTEMS IN RESIDENTIAL AND STREET ENVIRONMENTS

IV. CONCLUSION In this letter, signal strength measurements were conducted at the frequencies of 3.4, 5.3, and 6.4 GHz in order to characterize the behavior of narrowband signals in outdoor microcell residential and street environments. We developed a new path loss model, which is revised from the modified Hata model and suitable for the residential area at the frequency range of the 3–6 GHz band. The measurement results for LOS street indicate that the conventional two-ray model is still valid for candidate bands of IMT-Advanced. In a NLOS street, corner loss and path loss exponents are 13–19 and 7.2–13.6 dB, respectively. The findings presented here will help make link budget in outdoor microcell residential and street environments for IMT-Advanced service. REFERENCES [1] Framework and Overall Objectives of the Future Development of IMT2000 and Systems Beyond IMT-2000, ITU-R Rec. M. 1645, Jun. 2003. [2] Preliminary Draft New Report on Radio Aspects for the Terrestrial Component of IMT-2000 and Systems Beyond IMT-2000, ITU-R WP 8F/TEMP/290, Oct. 2005. [3] COST Action 231, “Digital mobile radio towards future generation systems, final report,” Tech. Rep., European Communities, EUR 18957, 1999. [4] W. C. Jakes, Microwave Mobile Communications. New York: IEEE Press, 1974.

871

[5] T. S. Rappaport, Wireless Communications Principles and Practice. Upper Saddle River, NJ: Prentice-Hall, 1996. [6] H. Xia, H. L. Bertoni, L. R. Maciel, A. Lindsay-Stewart, and R. Rowe, “Radio propagation characteristics for line-of-sight microcellular and personal communications,” IEEE Trans. Antennas Propag., vol. 41, no. 10, pp. 1439–1447, Oct. 1993. [7] L. B. Milstein, D. L. Schilling, R. L. Pickholtz, V. Erceg, M. Kullback, E. G. Kanterakis, D. S. Fishman, W. H. Biederman, and D. C. Salerno, “On the feasibility of a CDMA overlay for personal communications networks,” IEEE J. Sel. Areas Commun., vol. 10, no. 4, pp. 655–668, May 1992. [8] H. Masui, R. Ishii, K. Sakawa, H. Shimizu, T. Kobayashi, and M. Akaike, “Upper- and lower-bound evaluations in microwave urban LOS propagation,” in Proc. 11th IEE Conf. Antennas Propag., 2001, vol. 2, pp. 419–422. [9] Y. Oda and K. Tsunekawa, “Advanced LOS path-loss model in microcellular mobile communications,” IEEE Trans. Veh. Technol., vol. 49, no. 6, pp. 2121–2125, Nov. 2000. [10] V. Erceg, S. Ghassemzadeh, M. Taylor, D. Li, and D. L. Schilling, “Urban/suburban out-of-sight propagation modeling,” IEEE Commun. Mag., vol. 30, no. 6, pp. 56–61, Jun. 1992. [11] V. Erceg, A. J. Rustako, Jr., and R. S. Roman, “Diffraction around corners and its effects on the microcell coverage area in urban and suburban environments at 900 MHz, 2 GHz, and 6 GHz,” IEEE Trans. Veh. Technol., vol. 43, no. 3, pp. 762–766, Aug. 1994. [12] D. N. Schettino, F. J. S. Moreira, and C. G. Rego, “Novel UTD coefficients for lossy conducting wedges,” in Proc. IEEE IMOC, Oct. 2007, pp. 270–274. [13] H. Masui, M. Ishii, K. Sakawa, H. Shimizu, T. Kobayashi, and M. Akaike, “Microwave path-loss characteristics in urban LOS and NLOS environments,” in Proc. 53rd IEEE VTC 2001 Spring, pp. 395–398.