Samimi, M., Wang, K., Azar, Y., Wong, G. N., Mayzus, R., Zhao, H., Schulz, J. K., Sun, S., Gutierrez, F., Rappaport, T. S., “28 GHz Angle of Arrival and Angle of Departure Analysis for Outdoor Cellular Communications using Steerable Beam Antennas in New York City,” to appear in the 2013 IEEE Vehicular Technology Conference (VTC), June 2~5, 2013. 28 GHz Angle of Arrival and Angle of Departure Analysis for Outdoor Cellular Communications using Steerable Beam Antennas in New York City Mathew Samimi, Kevin Wang, Yaniv Azar, George N. Wong, Rimma Mayzus, Hang Zhao, Jocelyn K. Schulz, Shu Sun, Felix Gutierrez, Jr., and Theodore S. Rappaport NYU WIRELESS Center Polytechnic Institute of New York University, Brooklyn, NY 11201 tsr@nyu.edu Abstract— Propagation measurements at 28 GHz were conducted in outdoor urban environments in New York City using four different transmitter locations and 83 receiver locations with distances of up to 500 m. A 400 mega-chip per second channel sounder with steerable 24.5 dBi horn antennas at the transmitter and receiver was used to measure the angular distributions of received multipath power over a wide range of propagation distances and urban settings. Measurements were also made to study the small-scale fading of closely-spaced power delay profiles recorded at half- wavelength (5.35 mm) increments along a small-scale linear track (10 wavelengths, or 107 mm) at two different receiver locations. Our measurements indicate that power levels for small-scale fading do not significantly fluctuate from the mean power level at a fixed angle of arrival. We propose here a new lobe modeling technique that can be used to create a statistical channel model for lobe path loss and shadow fading, and we provide many model statistics as a function of transmitter- receiver separation distance. Our work shows that New York City is a multipath-rich environment when using highly directional steerable horn antennas, and that an average of 2.5 signal lobes exists at any receiver location, where each lobe has an average total angle spread of 40.3° and an RMS angle spread of 7.8°. This work aims to create a 28 GHz statistical spatial channel model for future 5G cellular networks. Keywords—28 GHz, 5G, millimeter wave, RF propagation, channel sounder, statistical spatial channel model, AOA, AOD, lobe, angle spread, path loss, shadow fading, polar plot. I. INTRODUCTION Recent work has suggested the viability of mm-wave spectrum (i.e. at 28 GHz and 38 GHz) to satisfy the increased bandwidth demand for cellular and backhaul; these bands provide much more spectrum and allow the use of miniature high gain antennas [1][2][3]. Fig. 1 shows the relationship between atmospheric attenuation at sea level and frequency. The additional propagation attenuation at 28 GHz is about 0.06 dB/km; however, since the maximum cell size is projected to be 200 m at 28 GHz [4], the vertical axis of Fig. 1 can be divided (in dB) by a factor of five. Thus, atmospheric attenuation is 0.012 dB/200 m at 28 GHz, a negligible value like cellular bands used today. When using directional antennas, urban propagation and Fig. 1. Air attenuation at sea level versus frequency, showing the additional path loss due to atmospheric oxygen [2]. The white circle highlights the 28 GHz and 38 GHz frequency bands that will be used for future 5G wireless. The green circles (solid border) highlight frequencies that have comparable free space (air) characteristics to modern cellular frequencies. The blue circles (dotted border) show frequencies with greater attenuation, which are therefore ideal for short-range indoor communications. rain attenuation are also not major concerns at 28 GHz [4]. Using the mm-wave spectrum for future cellular and backhaul networks will not only alleviate the current spectrum shortage, but also offer multi-gigabit per second data for each mobile user . Recent analysis has shown that the relative power consumption of wireless devices decreases as the RF bandwidth increases [5], implying that future smartphones will not only have faster data rates, but also be more energy efficient. The commercialization of mm-wave cellular can be realized with high-gain steerable antennas, which are necessary to overcome the propagation challenges of higher carrier frequencies, and to direct energy towards optimal directions that can exploit multipath and successfully complete a link [6]. Future smart antenna arrays, possibly on-chip [1], will algorithmically determine the optimal angle of arrival (AOA) and angle of departure (AOD). One possible beam steering method is to use narrowband pilot tones that enable the prediction of the spatial location of multipath based on narrowband