Uplink Interference Mitigation Techniques for Coexistence of 5G mmWave Users with Incumbents at 70 and 80 GHz

1 Uplink Interference Mitigation T echniques for Coe xistence of 5G mmW a v e Users with Incumbents at 70 and 80 GHz Ghaith Hattab, Student Member , IEEE, Eugene V isotsk y , Member , IEEE, Mark Cudak, Member , IEEE, and Amitav a Ghosh, F ellow , IEEE Abstract The millimeter wa ve spectra at 71-76GHz (70GHz) and 81-86GHz (80GHz) ha ve the potential to endow fifth-generation ne w radio (5G-NR) with mobile connectivity at gigabit rates. Howe ver , a pressing issue is the presence of incumbent systems in these bands, which are primarily point-to-point fixed stations (FSs). In this paper , we first identify the key properties of incumbents by parsing databases of existing stations in major cities to devise se veral modeling guidelines and characterize their deployment geometry and antenna specifications. Second, we de velop a detailed uplink interference framework to compute the aggregate interference from outdoor 5G-NR users into FSs. W e then present several case studies in dense populated areas, using actual incumbent databases and building layouts. Our simulation results demonstrate promising 5G coexistence at 70GHz and 80GHz as the majority of FSs experience interference well below the noise floor thanks to the propagation losses in these bands and the deployment geometry of the incumbent and 5G systems. F or the few FSs that may incur higher interference, we propose se veral passive interference mitigation techniques such as angular-based e xclusion zones and spatial power control. Simulation results sho w that the techniques can effecti vely protect FSs, without tangible de gradation of the 5G coverage. This paper was presented in part at the IEEE Global communications Conference (Globecom), Singapore, December 2017 [1]. G. Hattab was with Nokia Bell Labs. He is currently with the Department of Electrical Engineering, Uni versity of California, CA, USA. E. V isotsk y , M. Cudak, and A. Ghosh are with Nokia Bell Labs in Naperville, IL, USA. (email: ghattab@ucla.edu, { eugene.visotsky , mark.cudak, amitava.ghosh } @nokia-bell-labs.com). 2 Index T erms 5G, coe xistence, interference management, spectrum sharing, mmW ave, wireless backhaul. I . I N T RO D U C T I O N Fifth-generation new radio (5G-NR) is en visioned to be the first cellular standard with mil- limeter wav e (mmW av e) spectrum access [2], [3]. Such paradigm shift to wards mmW av e access is necessary to scale with the explosi v e growth of mobile traffic and to provide unparalleled network capacity , with peak data rates reaching tens of Gbps [4]. Indeed, the mmW av e spectrum has attracted significant attention from standard bodies, industry , and the academic community , culminating recently when the Federal Communications Commission (FCC) has opened up 3.85GHz of licensed spectrum for cellular services, and specifically at 28GHz (27.5-28.35GHz) and 39GHz (37-40GHz) [5]. Nev ertheless, there is still an additional 10GHz of licensed spectra at 70GHz (71-76GHz) and 80GHz (81-86GHz) that are left for future consideration as candidate bands for mmW a ve mobile networks [5], [6]. The advantages of using 70GHz and 80GHz bands, also kno wn as the e-band , are twofold. First, each band can easily pro vide a contiguous high bandwidth, e.g., 2GHz, in contrast to 28GHz and 39GHz, where each provides a maximum of 850MHz and 1.6GHz, respecti vely . Second, the e-band is av ailable worldwide, enabling economies of scale through uni versal adoption of common mmW av e de vices. Equally important, operating at the higher end of the mmW av e spectrum is not significantly different from operating at 28GHz as the channel models are the same [7], and the increase in path loss can be compensated by using an array with a larger number of antenna elements. In addition, sev eral prototypes hav e shown the feasibility of mmW av e systems over 70GHz. For instance, Nokia and Hua wei have already demonstrated experimental 5G systems designed to operate at 73.5 GHz [8]–[10]. One ke y challenge of using the 70GHz and 80GHz bands is the presence of existing in- cumbents, which are primarily fixed stations (FSs) that provide point-to-point services such as wireless backhaul. Per FCC regulations, these incumbents must be protected from harmful interference. Thus, our objectiv e is to study the feasibility of the coexistence of 5G systems with existing FSs and to dev elop interference mitigation techniques that ensure harmonious spectrum sharing. 3 A. Related work Se veral works ha ve studied spectrum sharing paradigms for mmW av e networks [11]–[13]. Ho wev er , these works ha ve solely focused on sharing among dif ferent mobile operators, e.g., sharing frequenc y channels, infrastructure, etc. Spectrum sharing of 5G systems and other services has recently attracted attention. For instance, the work in [14]–[16] focus on the 5G coexistence with radar systems, whereas the work in [17] studies the coexistence with W iFi. While these aforementioned works are limited to sub-6GHz, the mmW ave access paradigm has also spurred interest in coe xistence studies. For e xample, the work in [18] and [19] study the feasibility of 5G coexistence with incumbents at 28GHz, which are satellite systems, and the coexistence with fixed service at 39GHz. In [20], the impact of FSs interference on the throughput of UEs operating at 28GHz is studied. A more relev ant work to this paper is the one in [21], which studies the coe xistence of 5G with FSs at 70GHz. Howe ver , the work makes se veral modeling assumptions, e.g., only a single FS is assumed to e xist at a fixed distance from the 5G system and only a fixed portion of links are assumed to be non-line-of-sight (NLOS). In addition, the work in [21] focuses on the 5G downlink (DL) interference. T o mitigate the uplink (UL) interference, a probing device is proposed to be installed on the FS to report excessi ve interference to the 5G system. In this work, howe ver , we focus on UL passi ve interference mitigation techniques, i.e., we propose techniques that do no require an y coordination between the 5G system and the incumbents or require probing devices. B. Contributions The main contrib utions of this paper are summarized as follo ws. • Characterizing incumbents: W e analyze databases of existing FSs in four major areas in the United States, characterizing their deployment geometry and key antenna specifications. The analysis pro vides insights on the feasibility of 5G coexistence and gi v es benchmarks for accurate modeling of FSs, which can be of interest to the academic community . • 5G uplink interfer ence analysis: W e present a detailed interference analysis frame work to compute the aggre gate uplink (UL) interference from 5G users into FSs. W e also present random models for user’ s azimuth and elev ation antenna directions to help reduce the simulator complexity without de grading the simulation’ s accuracy . • Passiv e interference mitigation: W e propose se veral passi ve interference techniques that do not require any coordination between the 5G system and the incumbent systems. Specifically , 4 we propose sector-based and beam-based exclusion zones where 5G base stations (gNBs) switch off certain beams to protect victim FS recei vers. While these techniques are shown to be effecti ve, they can af fect the 5G do wnlink (DL) coverage. Thus, we propose spatial po wer control, defining quiet beams where associated users transmit at lower po wer . W e discuss the implementation of such techniques for 5G-NR. The coexistence fe asibility and the ef fectiv eness of the proposed mitigation techniques are v alidated via three case studies, where we deploy 5G systems in dense urban and suburban areas. The studies use the databases of existing FSs and actual b uilding layouts for accurate interference analysis. Our results hav e shown that the majority of FSs are protected from harmful interference due to the high propagation losses at 70GHz and 80GHz, the high attenuation due to the misalignment between the user and the FS’ s antenna boresight, and the deployment geometry of FSs and 5G systems. For the few FSs that experience higher interference, the proposed mitigation techniques provide significant protection, and they are more effecti ve than switching of f gNBs that are in vicinity of FSs. Finally , as a by-product of the simulation set-up, we validate the performance of 5G networks in 70GHz and 80GHz and show the distribution of the beams used by the gNB and the user , making design insights for mobile network operators and vendors. C. P aper Or ganization The rest of the paper is or ganized as follo ws. The system model is presented in Section II. The study of FSs’ deployment and the interference analysis framew ork are presented in Section III and Section IV, respecti vely . The proposed mitigation techniques are discussed in Section V. Simulation results are presented in Section VI, and the conclusions are drawn in Section VII. I I . S Y S T E M M O D E L A. 5G base stations (gNBs) W e consider a street-lev el deployment of gNBs such that each one is deployed at a street corner at height h g and the inter -site distance (ISD) between e very site is approximately d ISD . 1 Each site consists of four sectors, i.e., each sector covers an area of 90 ◦ . 1 In the simulation set-up, we first deploy gNBs in a grid with a fixed ISD of d ISD , covering the entire simulated area. Then, we look at the location of each dropped gNB to check if it lies at a street corner . If the gNB does not lie at a corner , we move it from its initial location to the nearest street corner , gi ven that there are no gNBs located there. Such deployment strategy is followed by mobile operates, where gNBs (or small cells) are deployed at street corners ev ery few blocks. 5 2 T xRUs ܰ ୥ǡ୦ ܰ ୥ǡ୴ Cov ers a se ctor Fig. 1: An illustrativ e example of a 5G gNB site. Each sector is equipped with a large-scale cross-polarized antenna array of size N g , h × N g , v × 2 . The antenna array is assumed to be mechanically tilted downw ard at angle φ g (fe w degrees) as the majority of outdoor UEs are at a ground-le vel, whereas the gNB is fe w meters abov e the ground. Each antenna element has a gain of G g and a transmit po wer of P g and is half-wa velength apart from the nearest antenna element. An illustrati ve e xample of a gNB site is giv en in Fig. 1 [8]. B. 5G Users W e only consider outdoor user equipment terminals (UEs), that are randomly deployed ov er space, as FSs are outdoors and the attenuation due to penetration losses for indoor UEs is very high at 70GHz and 80GHz. Each UE is equipped with a cross-polarized antenna array of size N u , h × N u , v × 2 , where each antenna element has a gain and a transmit po wer of G u and P u , respecti vely . The UE array height is assumed to be h u , and it is titled upward at angle φ u 2 . The UE is also assumed to have two panels, i.e., two sectors, with each one co vering 180 ◦ . Thus, the user can sense beams in all directions, but only one panel will be acti ve after user and beam association. During cell selection and association, the UE measures the receiv ed power of reference signals sent ov er dif ferent beams from gNBs in vicinity of the UE. Then, the UE connects to the beam with the highest receiv ed power (other beam association algorithms or criteria can be considered [22]–[26]). 2 The actual mechanical tilt will depend on the UE, yet assuming an upward one can be considered as a worst case scenario. W e note that we also consider a randomized tilt in Section IV.C 6 C. Incumbent F ixed Stations W e consider FSs that operate in the 71-76GHz and 81-86GHz bands, and they are currently registered in the FCC’ s database as incumbents are required to be in the database for operating in these bands [27]. Thus, their e xact three-dimensional locations are used. Similarly , we extract their antenna specifications, e.g., beamwidth, gain, azimuth orientation, and tilt. While dif ferent FSs may operate at dif ferent center frequencies in the aforementioned bands, we assume in this paper that all of them share the same spectrum with the 5G system, as a worst case scenario. D. Antenna P atterns For beam association and data communications, the gNB can use one of the 4 N g , h N g , v av ailable beams, where we assume the number of beams per dimension is twice the number of antennas in that dimension. 3 The azimuth (or elev ation) beam pattern beamwidth is approximately θ BW g , BP ≈ 102 / N g , h (or φ BW g , BP ≈ 102 / N g , v ) [28]. W e further assume a parabolic element pattern such that the normalized azimuth and elev ation attenuations are, in dB, [29] A g , EP ( θ ) = 12 θ θ BW g , EP ! 2 and A g , EP ( φ ) = 12 φ φ BW g , EP ! 2 , (1) where θ BW g , EP and φ BW g , EP are the element pattern 3dB beamwidths in azimuth and ele v ation, respecti vely . The same definitions are applied for the UE side, replacing the subscript g with u . Fig. 2a and Fig. 2b show the antenna patterns of 5G gNBs and UEs, respectiv ely , where it is assumed that the gNB and UE arrays are, respecti vely , of size 16 × 8 × 2 and 4 × 4 × 2 . For the incumbent system, we assume all FSs have antenna patterns that, at least, meet the FCC’ s regulation as specified in [30]. Essentially , the re gulation specifies the minimum radiation suppression for a giv en angle from the centerline of the main beam. Fig. 2c sho ws the normalized antenna gain for a gi ven of f-axis angle. Due to the high directivity of the FS’ s antenna, it is sho wn that a slight misalignment with the main boresight is enough to incur significant signal attenuation. A summary of the main parameters used is provided in T able I. 3 The number of beams, or directions to sweep, is a design parameter that also depends on the type of antenna used. Sweeping the angular domain with more beams improves the cov erage of the 5G system as narrower beams, with higher gain, are used. Howe ver , finer sweeping typically increases the search space, increasing the complexity and delay of initial access. 7 -20 0 20 40 60 80 100 120 Azimuth angles (degrees) -40 -30 -20 -10 0 Antenna gain (dBi) gNB Antenna Pattern (Steering orientation=45 ° ) -60 -40 -20 0 20 40 60 Elevation angles (degrees) -40 -30 -20 -10 0 Antenna gain (dBi) gNB Antenna Pattern (Steering orientation=-6 ° ) (a) gNB beams -90 -45 0 45 90 135 180 225 270 Azimuith angles (degrees) -60 -40 -20 0 Antenna gain (dBi) UE Antenna Pattern (Steering orientation=90 ° ) -60 -40 -20 0 20 40 60 80 Elevation angles (degrees) -60 -40 -20 0 Antenna gain (dBi) UE Antenna Pattern (Steering orientation=6 ° ) (b) UE beams 0 1 02 03 04 05 06 07 08 09 0 Off-axis elevation/azimuth angles (degrees) -60 -50 -40 -30 -20 -10 0 Normalized Antenna Gain (dBi) (c) FS beams Fig. 2: Antenna beam patterns of: (a) the gNB; (b) the UE, and (c) the FS T ABLE I: Main parameters and their v alues if applicable Symbol Description V alue(s) if applicable h ( · ) Height of gNB or UE h g = 6 m; h u = 1 . 5 m d ISD Inter-site distance d ISD = 200 m N ( · ) , h Number of columns in an array N g , h = 16 ; N u , h = 4 N ( · ) , v Number of rows in an array N g , v = 8 ; N u , v = 4 φ ( · ) Antenna tilt φ g = − 6 ◦ ; φ u = 6 ◦ G ( · ) Antenna gain G g = G u = 5 dBi P ( · ) Antenna transmit power P g = 7 dBm; P u = 1 dBm θ BW ( · ) , ( · ) 3dB beamwidth of beam/element patterns in azimuth θ BW g , BP = 6 ◦ ; θ BW u , BP = 25 ◦ ; θ BW g , EP = θ BW u , EP = 65 ◦ φ BW ( · ) , ( · ) 3dB beamwidth of beam/element patterns in elev ation φ BW g , BP = 12 ◦ ; φ BW u , BP = 65 ◦ ; φ BW g , EP = φ BW u , EP = 65 ◦ A ( · ) , FTBR Front-to-back ratio loss A f , FTBR = 55 dB; A g , FTBR = A u , FTBR = 30 dB F ( · ) Noise figure F u = 9 dB B Channel bandwidth B = 1 GHz f c Carrier frequency f c = { 73 . 5 , 83 . 5 } GHz x a ( x, y ) -coordinates of a d a → b 2D distance from a to b (m) PL a → b Path loss from a to b (dB) X ( · ) Log-normal shadowing with standard deviation of σ ( · ) σ LOS = 4 dB; σ NLOS = 7 . 82 dB β Indicator variable that denotes a blockage event Blockage: β = 1 ; No blockage: β = 0 G ( · ) , max Maximum antenna gain (dBi) I I I . A N A LY S I S O F F S S D E P L OY M E N T In this section, we study the deployment of FSs to get some guidelines on their deployment geometry and features. The insights help understand ho w the deployment of FSs affects the coexistence with 5G systems. Equally important, they can be also used as a benchmark for modeling FSs using stochastic-based approaches [31]. W e parse the databases of FSs deployed in four major metropolitan areas: Chicago, New Y ork, Los Angeles, and San Fransisco [27]. Each database cov ers an area of radius 300km. T able II summarizes the analyzed databases. A link is defined as a two-w ay communication between two FSs, whereas a pair is defined as a link with unique spatial coordinates of the FSs. Thus, the 8 T ABLE II: Current number of links and pairs in each database Database No. of links No. of pairs Chicago 1743 512 New Y ork 5303 1685 Los Angeles 1013 911 San Francisco 1892 1801 same pair could hav e multiple links, each over a different channel in 70GHz and/or 80GHz. A. Spatial Distrib ution W e first analyze the spatial distribution of these FSs.In Fig. 3a, we sho w the density of FSs around a city center with variations of the re gion radius, i.e., we compute the number of FSs in an area of a gi ven radius, where the area is centered around one of the city’ s main hubs (e.g., W illis T o wer for Chicago, the Empire State Building for Ne w Y ork, and the financial districts of Los Angeles and San Fransisco). It is evident that FSs are non-uniformly distributed over space, and specifically the y tend to hav e higher density near city centers while they become very sparsely deployed in suburban areas as city centers have higher density of people, buildings, and attractions, elev ating the need for denser fixed backhaul links. Overall, FSs hav e low density relati ve to existing cellular networks. For each FS density sho wn in Fig. 3a, we also compute the a verage height among those FSs deployed in a gi v en area, and sho w in Fig. 3b these heights for the different densities of FSs. It is sho wn that, e xcept for San Francisco, the av erage height generally increases in denser areas compared to lightly dense areas, showing that the deployment height appears to be correlated with the av erage building heights in these areas. From the 5G coexistence perspectiv e, this implies that the density of FSs in urban areas should not be worrisome as these stations tend to be deployed at altitudes that are abov e 5G cell sites. In contrast, FSs are likely to be deployed at relativ ely low heights in sub urban areas, yet their density is very low in such re gions. Fig. 3c and Fig. 3d sho w the cumulati ve density function (CDF) and the probability density function (PDF) of the FSs’ deployment height. The a verage and median heights are at least 34m and 19m, respectiv ely . More importantly , 95% of FSs are deployed abo ve 12m for most metropolitan areas. Note that for LA, the fifth percentile is 2m, b ut this is relativ e to ground, i.e., many of FSs in LA are actually deployed on hills. Since 5G sites are e xpected to be deployed 9 0 50 100 150 200 250 300 Radius (km) 10 -3 10 -2 10 -1 10 0 10 1 10 2 Density (per km 2 ) Chicago New York Los Angeles San Francisco (a) Density with variations of region’ s radius 10 -3 10 -2 10 -1 10 0 10 1 10 2 Density (per km 2 ) 35 40 45 50 55 60 65 70 Average height (m) Chicago New York Los Angeles San Francisco (b) A verage height for a gi ven density 0 20 40 60 80 100 120 140 160 180 200 Height (m) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 CDF Chicago New York Los Angeles San Francisco Avg.= 63.3m median=43.5m 5th=12.4m 95th=179.3m Avg.= 48.7m median=31.7m 5th=13.5m 95th=133.5m Avg.= 34.6m median=22.6m 5th=17.3m 95th=110.4m Avg.= 34.1m median=19.2m 5th=2m 95th=121.8m (c) CDF of FSs’ height 0 200 400 600 Height (m) 0 0.005 0.01 0.015 0.02 Histogram Chicago 0 200 400 600 Height (m) 0 0.01 0.02 0.03 Histogram New York 0 100 200 300 400 Height (m) 0 0.01 0.02 0.03 0.04 Histogram Los Angeles 0 100 200 300 Height (m) 0 0.01 0.02 0.03 Histogram San Francisco (d) PDF of FSs’ height Fig. 3: FSs’ spatial deployment. at heights of four to six meters, gNBs will be belo w the majority of FSs, limiting the 5G interference on FSs and vice versa. B. Antenna Specifications Another critical aspect of FSs’ deployment is their physical antenna orientation. Fig. 4b shows the histogram of the antenna’ s tilt, verifying that the v ast majority of FSs hav e their tilt angles pointing horizontally . For instance, more than 93% of FSs hav e their tilt angles within [ − 10 , 10] degrees. There are only few FSs with high ne gati ve tilts, i.e., they point to the street lev el. These FSs, ho we ver , are typically deployed on top of high-rise buildings as verified in Fig. 4b. In other words, there is a correlation between the deployment height and the neg ativ e tilt. Thus, although these FSs will hav e a higher chance to experience UE interference, as they point to the ground, 5G signals will typically experience a lar ger path loss giv en the height of these FSs. Another key feature of FSs is their high antenna gain. Indeed, as sho wn in Fig. 4c, the antenna gain is typically from 40dBi to 55dBi. Such high gains are necessary for long-range coverage at millimeter wa ve frequencies, but can be troublesome for other transmitter-recei ver pairs in 10 -40 -20 0 20 40 Tilt (degrees) 0 0.05 0.1 0.15 0.2 Histogram Chicago -40 -20 0 20 40 Tilt (degrees) 0 0.1 0.2 0.3 0.4 Histogram New York -40 -20 0 20 40 Tilt (degrees) 0 0.1 0.2 0.3 0.4 Histogram Los Angeles -40 -20 0 20 40 Tilt (degrees) 0 0.05 0.1 0.15 0.2 0.25 Histogram San Francisco (a) T ilt histograms -60 -40 -20 0 Tilt (degrees) 0 200 400 600 Height (m) Chicago -60 -40 -20 0 Tilt (degrees) 50 100 150 200 250 300 Height (m) Tilt (degrees) New York -60 -40 -20 0 0 100 200 300 400 Height (m) Tilt (degrees) Los Angeles -60 -40 -20 0 Tilt (degrees) 0 50 100 150 200 Height (m) San Francisco (b) A verage height for a gi ven tilt 30 40 50 60 Max Antenna Gain (dBi) 0 0.02 0.04 0.06 0.08 0.1 Histogram Chicago 30 40 50 60 Max Antenna Gain (dBi) 0 0.02 0.04 0.06 0.08 0.1 Histogram New York 30 40 50 60 Max Antenna Gain (dBi) 0 0.05 0.1 0.15 Histogram Los Angeles 30 40 50 60 Max Antenna Gain (dBi) 0 0.05 0.1 0.15 0.2 Histogram San Francisco (c) Antenna gain histograms 0 0.5 1 1.5 Beamwidth (degrees) 0 1 2 3 4 Histogram Chicago 0 0.5 1 1.5 Beamwidth (degrees) 0 1 2 3 4 5 Histogram New York 0 0.5 1 1.5 Beamwidth (degrees) 0 2 4 6 Histogram Los Angeles 0 0.5 1 1.5 Beamwidth (degrees) 0 2 4 6 8 Histogram San Francisco (d) Beamwidth histograms Fig. 4: FSs’ antenna information vicinity . For this reason, the maximum 3dB beamwidth, per FCC regulations, should be less than or equal to 1 . 2 ◦ [30]. This is verified in Fig. 4d, where the vast majority of FSs ha ve beamwidths at 1 ◦ . From a 5G coexistence perspectiv e, the UE must be tightly aligned with the FS for it to cause tangible interference. Otherwise, most 5G signals will be highly attenuated, falling outside the FS recei ver’ s beam (cf. Fig. 2c). C. Comments on incumbent modeling and 5G coexistence The aforementioned analysis of the different incumbents’ databases helps provide sev eral modeling guidelines of incumbent FSs. For instance, using the popular homogeneous Poisson Point Process (HPPP) [31] to model the locations of FSs may not be practical if the region of interest is large, as FSs tend to be non-uniformly distrib uted ov er space. In addition, due to the disparities between the height of FS deplo yment and the 5G mmW av e deployment, it is more meaningful to consider three-dimensional stochastic processes (or two-dimensional processes with the third dimension being a constant that reflects the mean height of the buildings in a giv en area). For antenna parameters, it is observed the majority of FSs hav e similar characteristics, and 11 thus it suffices to assume all of them hav e the same antenna gain and beamwidth, and further assume they point horizontally in elev ation. From a coexistence perspectiv e, the deployment strategy of FSs is fa vorable for future 5G deployment ov er 70GHz and 80GHz for the following reasons: • FSs are generally deployed above 12m, whereas 5G cell sites will be only at 4 to 6 meters abov e the ground for street-lev el deployment, and hence they will be well below FSs. • The v ast majority of FSs are oriented horizontally , i.e., the y are directed above 5G deploy- ments. For the fe w FSs that point to the street le vel, these are typically at high altitudes, increasing the path loss between the UE and the FS. • The ultra-narrow beamwidths of FSs can help significantly attenuate UE interfering signals when they fall outside the main lobe. I V . A N A L Y S I S O F U E I N T E R F E R E N C E O N F S S In this section, we present our framew ork to compute the aggreg ate interference from the 5G system into incumbent systems. The approach used is applicable to the coexistence of any two wireless communication systems that rely on directional beams. W e focus on the 5G system operating in the uplink mode, i.e., we study the UE interference into FSs, for the follo wing reasons. First, UEs typically hav e positi ve tilt angles compared to 5G gNBs, and thus the former are more likely to interfere with FSs. Second, the mobility of UEs mak es their locations appear random, while gNBs’ deployment can be optimized to ensure minimal interference on FSs. W e note that in the Simulations Section, we sho w that although gNBs have higher transmit po wer and antenna gains, the 5G DL aggregate interference is not higher than that of the UL, primarily because gNBs tend to hav e beams pointing to the ground. The interference seen at a victim FS is an aggre gation of all UEs transmitting in the UL to their respecti ve gNBs. This interference depends mainly on three components: (i) The path loss between the UE and the FS, (ii) the attenuation due to the FS’ s antenna pattern, and (iii) the attenuation due to the UE’ s antenna pattern. W e describe each one in details next. In what follo ws, x u , x f , and x f , tx denote the ( x, y ) -coordinates of the interfering UE, the victim FS recei ver , and the corresponding FS transmitter , respectiv ely . In addition, d a → b denotes the 2D distance between a and b while a • b denotes the dot product between two vectors a and b , i.e., a T b . 12 A. P ath Loss between a User and a F ixed Station Signals can be significantly attenuated if they are blocked by objects at mmW a ve frequencies, i.e., it is critical to consider whether the link is LOS or NLOS for path loss computations at such high frequencies. T o this end, we use the 3GPP path loss model [29], which is expressed, in dB, as 4 PL u → f = 1 ( β =0) PL LOS ( x u , x f , h u , h f , f c ) + 1 ( β =1) PL NLOS ( x u , x f , h u , h f , f c ) + X ( β ) , (2) where PL LOS ( · ) is the line-of-sight (LOS) path loss, PL NLOS ( · ) is the non-LOS (NLOS) path loss, X is the log-normal shadow fading, β ∈ { 0 , 1 } is a binary v ariable that indicates whether the UE-FS is blocked by a building or not, and 1 ( · ) is the indicator function. W e note that PL LOS and PL NLOS are functions of the distance between the UE and the FS, their heights, and the center frequency f c , as giv en in [29]. Essentially , the path loss is a multi-slope model with dif ferent path loss exponents depending on the distance between the UE and the FS. Also, the standard deviation of the log-normal shado w fading depends on whether the link is LOS or NLOS [29]. In this work, we rely on actual building layouts to determine whether the link is LOS or NLOS. W e intentionally ignore blockage by other objects, e.g., foliage [32], cars, etc., to emulate a w orst case scenario as additional blockage should reduce the interference. W e note that in the Simulations Section, we compare the LOS probability using the actual building layouts with the theoretical LOS probability used by the 3GPP model, which is e xpressed as [29] P LOS ( d u → f ) = min  18 d u → f , 1  ×  1 − exp  − d u → f 36  + exp  − d u → f 36  . (3) As stated earlier , we define a blockage e vent as ha ving the UE-FS blocked by a building. This is computed as follows. Assuming the xy -plan represents the ground, we first check whether the line that connects between the UE and the FS is blocked by a building, which is defined as a 2D polygon. If the polygon does intersect with the line, we then check whether it blocks the line with the 3D v ersion of the polygon, where the third dimension is the building’ s height, h BL . Specifically , let d u → BL be the distance between the UE and the b uilding and d u → f be the distance between the UE and the FS. Then, a blockage e vent occurs if ˜ h + h u ≤ h BL , where ˜ h = d u → BL × tan  tan − 1  h f − h u d u → f  . (4) 4 The model can be generalized to include indoor losses and indoor-to-outdoor penetration losses, when UEs are located indoors, as shown in [29]. 13 ݄ ୤ ݄ ୆୐ ݄ ୳ ݀ ୳՜୤ ݀ ୳՜୆୐ ෨ ݄ Fig. 5: A blockage ev ent in 3D occurs when ˜ h + h u ≤ h BL . This is visualized in Fig. 5. B. Attenuation due to FS Antenna P attern As illustrated in Fig. 2c, a small misalignment between the recei ved signal and the FS’ s antenna boresight results in significant attenuation. Thus, it is critical to accurately compute the interfering signal angle-of-arriv al at the FS antenna. Define the line connecting the UE to the FS as the interfer ence axis . Let the off-axis azimuth angle θ off f → u be the angle between the FS’ s antenna boresight and the interference axis, then we hav e θ off f → u = cos − 1 ( u f → f , tx • u f → u ) , (5) where u f → f , tx = x f , tx − x f k x f , tx − x f k is the unit vector in the azimuth direction of the FS’ s antenna boresight, and u f → u = x u − x f k x u − x f k is the unit vector from the FS’ s antenna tow ards the UE. Similarly , let φ off f → u be the of f-axis elev ation angle, then it can be sho wn that φ off f → u = tan − 1  h f − h u d f → u  + φ f , (6) where φ f is the FS’ s antenna tilt. All these vectors and off-axis angles are sho wn in Fig. 6a. Finally , the combined azimuth and elev ation attenuation at the FS victim receiv er is expressed as G f → u = G f , max − min  A f ( θ off f → u ) + A f ( φ off f → u ) , A f , FTBR  , (7) where G f , max is the maximum antenna gain in dBi, A f , FTBR is the front-to-back ratio loss (FTBR) in dB, and A f ( · ) is the attenuation for a giv en off-axis angle, and it corresponds to the antenna pattern that matches the FCC regulations (cf. Fig. 2c) [30]. 14 C. UE Radiated P ower (EIRP) Into FS Antenna 1) Actual dir ections: The directions of the UE’ s two opposite panels are defined by unit vectors in the direction of the panels’ boresight. W e assume that these directions are random in azimuth such that the boresight of the first one is distrib uted uniformly as U (0 , 180) while the other one is pointing in the opposite direction, i.e., 180 ◦ from the first one. Only one of the UE antenna panels is activ e during data communications. Let u str u denote the unit vector in the azimuth direction of the UE’ s panel that is activ e. Also, let u beam u denote the unit vector in the azimuth direction of the main lobe of the UE’ s beam used in the UL, which corresponds to the beam with the maximum received po wer during user and beam association. W e similarly define v str u and v beam u for the ele vation directions. Then, the total radiated po wer from the UE into the direction of the victim FS is expressed, in dBm, as E u → f = P u + 10 log 10 (2 N u , h N u , v ) + G u , max − ( A u , BP ( θ beam u → f ) + A u , BP ( φ beam u → f )) − min { A u , EP ( θ str u → f ) + A u , EP ( φ str u → f ) , A u , FTBR } , (8) where G u , max is the maximum antenna gain and A u , FTBR is the FTBR loss. The azimuth of f-axis angles are computed as θ beam u → f = cos − 1  u u → f • u beam u  , (9) and θ str u → f = cos − 1  u u → f • u str u  , (10) where u u → f = − u f → u . The ele v ation off-axis angles are computed as φ beam u → f = tan − 1  h f − h u d f → u  − ∠ v beam u , (11) and φ str u → f = tan − 1  h f − h u d f → u  − ∠ v str u , (12) where ∠ · denotes the angle of the vector . All of the relev ant vectors and off-axis angles are illustrated in Fig. 6b. 2) Random dir ections: W e also present a random model for the UE’ s azimuth and ele vation directions. This model does not require the deployment of gNBs, and hence ignores the computa- tional complexity in simulating user and beam association. The model assumes that the UE uses the beam in the direction of the antenna’ s main boresight, i.e., u beam u = u str u and v beam u = v str u . T o 15 Inter ference ax is ߠ ୤՜୳ ୭୤୤ ܠ ୤ ܠ ୤ǡ୲ ୶ ܠ ୳ ܝ ୤՜୳ ܝ ୤՜୤ǡ୲୶ ݄ ୳ ݄ ୤ ݀ ୤՜୳ ߶ ୤՜୳ ୭୤୤ ߶ ୤ (a) W ith respect to the FS Inter ference ax is ܠ ୤ ݄ ୳ ݄ ୤ ݀ ୤՜୳ ܝ ୳՜୤ ܠ ୳ ܝ ୳ ୱ୲୰ ܝ ୳ ୠୣୟ୫ ߠ ୳՜୤ ୱ୲୰ ߠ ୳՜୤ ୠୣୟ୫ ܞ ୳ ୱ୲୰ ܞ ୳ ୠୣୟ୫ ߶ ୳՜୤ ୱ୲୰ ߶ ୳՜୤ ୠୣୟ୫ (b) W ith respect to the UE Fig. 6: Of f-axis azimuth and ele vation angles this end, we model the azimuth direction as a uniform random v ariable ˜ θ u ∼ U (0 , 360) , whereas the elev ation direction is modeled as ˜ φ u = tan − 1  h g − h u d g → u  , (13) where d g → u ∼ U ( d 0 , d ISD / 2) , where d 0 > 0 is some constant, e.g., in this work we consider d 0 = 10 m. These angles are used to compute the unit vectors needed for azimuth and ele vation of f-axis angles. D. UE Aggr e gate Interfer ence The interference caused by the i -th UE on the FS is giv en as I i, dBm = E i → f + G f → i − PL i → f . (14) W e use the interference-to-noise ratio (INR) to determine the effect of 5G interference on the incumbent, which is expressed as INR dB = 10 log 10 ( I agg ) − (10 log 10 ( N 0 B ) + F f ) , (15) where I agg = P i 10 I i, dBm / 10 , N 0 is the noise power spectral density (mW/Hz), B is the bandwidth (Hz), and F f is the noise figure of the FS (dB). V . P A S S I V E I N T E R F E R E N C E M I T I G A T I O N T E C H N I Q U E S In this section, we propose several interference mitigation techniques to protect the incumbent FSs. W e focus on two critical aspects. First, the techniques should be passi ve, i.e., they do not require any coordination with FSs, and second the y should be practical to implement to appeal for mobile operators and vendors. 16 A. Sector-based Mitigation In this technique, we propose to switch off sectors, creating sector -based e xclusion zones. The ke y idea is that the 5G UE beam directions are typically reciprocal to those of 5G gNBs. Thus, if such reciprocal directions point to FSs, then the UE must be discouraged from using them, i.e., the sector with a reciprocal direction pointing to wards the FSs should be switched of f. More formally , let u str ,i g be the unit vector in the direction of the i -th sector boresight and − u str ,i g is its reciprocal direction. Then, the i -th sector is switched off if s l,i =    1 , cos − 1  − u str ,i g • u g → f  ≤ ψ s 0 , otherwise , (16) where ψ s is a predetermined decision threshold. A more relaxed sector exclusion criterion is to switch sectors off if they are not only aligned with the FS’ s location b ut also its antenna orientation. Such criterion can still reduce the interference e xperienced at FSs as a slight mis- alignment with FS’ s antenna incurs significant signal attenuation. More formally , the i -th sector can be switched off if s o,i = 1 ( s l,i =1) × 1 (cos − 1 ( u str ,i g • u f → f , tx ) ≤ ψ s ) . (17) W e refer to (16) as location-based mitigation and (17) as orientation-based mitigation. Both techniques are demonstrated in Fig. 7a. B. Beam-based Mitigation In the sector -based mitigation, only four decisions need to be made a priori for each gNB, making the approach simple to implement. This, ho we ver , may result in tangible cov erage holes, af fecting the performance of the 5G system. T o this end, we can make e xclusion zones at a finer scale, where decisions are made on a beam-by-beam basis instead. Specifically , the i -th beam is switched off if b l,i =    1 , cos − 1  − u beam ,i g • u g → f  ≤ ψ b 0 , otherwise , (18) where ψ b is a predetermined beam decision threshold and u beam ,i g is a unit vector in the direction of the i -th beam. In other words, the same sector could hav e beams switched on and beams switched of f, depending on whether the beam meets the criterion in (18) or not. W e can also make decisions based on the orientation of the FS’ s along with its location, i.e., b o,i = 1 ( b l,i =1) × 1 (cos − 1 ( u beam ,i g • u f → f , tx ) ≤ ψ b ) . (19) 17 െܝ ୥ ୱ୲୰ǡ௜ ܝ ୥ ୱ୲୰ǡ௜ ܠ ୤ ܠ ୥ UEs should n ot connect to this secto r ܝ ୤՜୤ǡ୲୶ െܝ ୥ ୱ୲୰ǡ௜ ܝ ୥ ୱ୲୰ǡ௜ ܠ ୤ ܠ ୥ UEs ca n connect to this secto r ܝ ୥ ୱ୲୰ǡ௜ ൈ ܝ ୥՜୤ ܝ ୥՜୤ Loc ati on- bas ed Ori ent ati on -bas ed (a) Sector -based mitigation െܝ ୥ ୠୣୟ୫ǡ௜ ܝ ୥՜୤ ܠ ୤ ܠ ୥ ܝ ୤՜୤ǡ୲୶ ܝ ୥՜୤ ܠ ୤ ܠ ୥ ൈ ܝ ୥ ୠୣୟ୫ǡ ௜ ܝ ୥ ୠୣୟ୫ǡ ௜ െܝ ୥ ୠୣୟ୫ǡ௜ ܝ ୥ ୠୣୟ୫ǡ ௜ Loc ati on- bas ed Ori ent ati on -bas ed (b) Beam-based mitigation Fig. 7: Illustration of passiv e mitigation techniques Beam-based exclusion zone is sho wn in Fig. 7b. C. Spatial P ower Contr ol The aforementioned techniques can be classified as angular exclusion zones, leading ine vitably to lo wer downlink co verage with higher de gradation if sector-based zones are used instead of beam-based. Alternati ve to switching beams (or sectors) off, we can implement po wer control, where the key idea is to transmit at lo wer po wer for beams that hav e higher alignment with the incumbent receiv er . In this paper , we seek a simple binary power control algorithm, where two power le vels can be used depending on whether the beam is classified as a r e gular beam or as a quiet one. Specifically , if the of f-angle between the beam reciprocal direction and the incumbent receiv er is below a predetermined threshold, then the beam is classified as quiet or almost blank , and thus the UE will transmit at low po wer . If the beam is not aligned with the incumbent, then the UE transmits at the maximum allow able power . T o summarize, we hav e the 18 follo wing UL po wer control P UL , u ,i ? =    P lo , cos − 1  − u beam ,i ? g • u g → f  ≤ ψ b P up , otherwise , (20) where i ? is the index of the gNB beam that the UE connects to, and P lo and P up represent the lo w and high transmit powers, respecti vely . Note that it is natural to extend this approach to sectors or make it with respect to the orientation of the FS instead of its location. Remark: More sophisticated po wer control algorithms can be considered, particularly when a multi-user access scheme is used. For example, one candidate formulation is to optimize the po wer allocated ov er each beam direction such that the aggregate interference on the incumbent recei ver is minimized. D. Implementation Implementation of angular exclusion zones should be straightforward. Indeed, upon the de- ployment of the gNBs in a giv en region, the mobile operator must identify the FSs in vicinity using the FCC’ s database, where the operator can extract their locations and azimuth directions, which will be used to compute the necessary unit vectors. The operator then switch sectors (or beams) depending on the protection criterion used. Thus, UEs cannot find any reference signals from those sectors (or beams), and hence they do not connect to them during user and beam association. Clearly , the operator may need to update the sector-based (or beam-based) decisions if the FS’ s databased is changed, e.g., switch back sectors if an incumbent license is e xpiring, etc., which typically happens at a long-time scale. T o implement spatial binary power control, the operator must tag each beam, from the possible DL gNB beams, with an indicator variable denoting whether the beam is a regular beam or a quiet one, which is determined by computing cos − 1  − u beam ,i ? g • u g → f  . The indicator v alue and the allo wable transmit of the beam are then embedded in the reference signal sent ov er the beam during user association. This is done over the physical broadcast channel (xPBCH or ePBCH), and thus during synchronization, the UE can decode the master and system information blocks (MIB and SIB), identifying the UL transmit power limit over that beam. W e remark that the passi v e mitigation techniques primarily require the design of the angular protection thresholds, e.g., ψ s and ψ b . For instance, using a higher threshold value provides more protection to incumbents, yet this may come at the expense of the 5G system coverage. Since we 19 4.41 4.42 4.43 4.44 4.45 4.46 4.47 4.48 x-coord 10 5 4.6405 4.641 4.6415 4.642 4.6425 4.643 4.6435 4.644 4.6445 4.645 4.6455 y-coord 10 6 (a) Lincoln Park (852 gNBs) 4.474 4.476 4.478 4.48 4.482 4.484 4.486 4.488 x-coord 10 5 4.636 4.6365 4.637 4.6375 4.638 y-coord 10 6 (b) Chicago Loop (108 gNBs) 5.852 5.854 5.856 5.858 5.86 5.862 5.864 x-coord 10 5 4.5112 4.5114 4.5116 4.5118 4.512 4.5122 4.5124 4.5126 y-coord 10 6 (c) Lo wer Manhattan (58 gNBs) Fig. 8: Simulation scenarios. Here ‘ ◦ ’ denotes the FS, ‘ ♦ ’ denotes the gNB, and ‘ × ’ denotes the UE. use the 3GPP channel model, it is difficult to analytically determine the threshold that strikes a good balance between the 5G co verage performance and the INR at the incumbent receiv er . For this reason, a mobile network operator may run preliminary computer simulations, e.g., using the interference frame work used in this paper , to determine ψ s or ψ b . V I . S I M U L AT I O N R E S U L T S W e study the aggregate UE interference on FSs deployed in Lincoln Park, Chicago Loop, and Lo wer Manhattan. W e deploy gNBs on a grid in each of the aforementioned cities, where d ISD = 200 m. Further , we randomly deploy outdoor UEs in each city , and assume an UL instantaneous traffic load of 25%, i.e., each gNB site, which consists of four sectors, serv es one UE in a gi ven time slot. Fig. 8 shows one spatial realization of the three deployment scenarios. Fo the channel model, we use the 3GPP NR-UMi model [29]. All other important simulation parameters are giv en in T able I. W e consider the center frequencies: 73.5GHz and 83.5GHz, and assume that the UE maximum radiated power , without any attenuation, is 33dBm or 43dBm [33], [34]. Per FCC regulations, we consider A f , FTBR = 55 dB [30]. For noise power , we assume B = 1 GHz and N 0 is computed at temperature 290K. Finally , the FS’ s location, height, maximum antenna gain, antenna tilt, and noise figure, are all extracted from the FCC’ s incumbent database [27]. The subsequent results are a veraged out o ver 1000 spatial realizations, where each one has a dif ferent deployment of UEs and different channel realizations. Additionally , unless otherwise stated, the results consider actual UEs direction, where gNB-UE association is performed first. 20 -5 0 5 10 15 20 25 30 35 40 45 50 UE SNR (dB) 0 0.5 1 CDF Lincoln Park 73.5GHz 83.5GHz -5 0 5 10 15 20 25 30 35 40 45 50 UE SNR (dB) 0 0.5 1 CDF Chicago Loop -5 0 5 10 15 20 25 30 35 40 45 50 UE SNR (dB) 0 0.5 1 CDF Lower Manhattan (a) SNR CDF Mean Median Cell-edge 0 10 20 30 SNR (dB Lincoln Park 73.5GHz 83.5GHz Mean Median Cell-edge 0 10 20 30 SNR (dB Chicago Loop Mean Median Cell-edge 0 10 20 30 SNR (dB Lower Manhattan (b) SNR statistics Fig. 9: SNR at 5G UE A. V alidation of the 5G System W e first verify that deployment of gNBs lead to reliable coverage for UEs. Fig. 9a shows the CDF of the signal-to-noise ratio (SNR) at the UE side after beam association, whereas Fig. 9b shows the main SNR statistics. Overall, it is shown that the deployment provides reliable cov erage with positiv e cell-edge SNR v alues. Operating at 83.5GHz has slight SNR degradation due to higher path loss compared to operating at 73.5GHz. Next, we look at the DL and UL beams used by gNBs and UEs, respectiv ely , after user and beam association. This provides insights on which beams are likely to be used by the gNB and the UE for a realistic deployment scenario. In Fig. 10, we sho w the histograms of the beams used in azimuth and elev ation by gNBs and UEs. It is sho wn that, ov erall, each azimuth gNB beam is equally likely to be used, with a similar observation regarding the gNB sectors. More importantly , only fe w ele vation beams are activ e. This suggests that mobile operators should implement only a couple of ele v ation beams to serve outdoor users, which reduces the complexity of user association and codebook design. For the UE, only fe w ele v ation beams are used as well, with the majority of them being less than 10 ◦ . Further , the UE uses the azimuth center beams more frequently because the y hav e higher array gain. This suggests that at the UE side, only fe w candidate directions should be explored during user association, and particularly those that are centered around the physical orientation of the UE antenna boresight. This observ ation helps significantly reduce the beam search space in user association [24], [25]. 21 0 1 02 03 04 05 06 07 08 09 0 Azimuth Beam Angle/sector (degrees) 0 0.05 0.1 Histogram 0 1 02 03 04 05 06 07 08 09 0 Azimuth Beam Angle/sector (degrees) 0 0.05 0.1 Histogram Lincoln Park 0 1 02 03 04 05 06 07 08 09 0 Azimuth Beam Angle/sector (degrees) 0 0.05 0.1 Histogram Lower Manhattan (a) gNB azimuth beams -25 -20 -15 -10 -5 0 5 Elevation Beam Angle (degrees) 0 0.5 Histogram Chicago Loop -25 -20 -15 -10 -5 0 5 Elevation Beam Angle (degrees) 0 0.5 Histogram Lincoln Park -25 -20 -15 -10 -5 0 5 Elevation Beam Angle (degrees) 0 0.5 Histogram Lower Manhattan (b) gNB elev ation beams 1234 Sector Index 0 0.2 0.4 Histogram Chicag Loop 1234 Sector Index 0 0.2 0.4 Histogram Lincoln Park 1234 Sector Index 0 0.2 0.4 Histogram Lower Manhattan (c) gNB sectors 0 20 40 60 80 100 120 140 160 180 Azimuth Beam Angle (degrees) 0 0.1 0.2 Histogram 0 20 40 60 80 100 120 140 160 180 Azimuth Beam Angle (degrees) 0 0.1 0.2 Histogram Lincoln Park 0 20 40 60 80 100 120 140 160 180 Azimuth Beam Angle (degrees) 0 0.1 0.2 Histogram Lower Manhattan (d) UE azimuth beams -5 0 5 10 15 20 25 30 35 Elevation Beam Angle (degrees) 0 0.5 Histogram Chicago Loop -5 0 5 10 15 20 25 30 35 Elevation Beam Angle (degrees) 0 0.5 Histogram Lincoln Park -5 0 5 10 15 20 25 30 35 Elevation Beam Angle (degrees) 0 0.5 Histogram Lower Manhattan (e) UE elev ation beams 0 5 10 15 20 25 30 35 Elevation Angle (degrees) 0 0.02 0.04 0.06 0.08 0.1 0.12 PDF (f) PDF of UE ele vation angle un- der the random model Fig. 10: Distrib ution of used beams and sectors B. Distribution of INR Fig. 11 sho ws the CDF of INR for the different case studies. W e also show a reference INR threshold of − 6 dB, which corresponds to signal-to-noise-plus-interference ratio (SINR) degradation of 1dB, meeting the FCC’ s interference protection criterion [30]. W e hav e the follo wing observations. First, using the random model, i.e., random UE azimuth and ele vation directions, provides accurate results that match well with computing the actual pointing directions of the UE in the presence of gNBs. This follo ws because the deployment of gNBs is agnostic to the locations of FSs, and the distrib ution of used ele v ation directions (cf. Fig. 10e) has a similar PDF to the one used in the random model (cf. (13) and Fig. 10f). Second, the CDFs sho w that the INR is ov erall lo w , with the majority of FSs experiencing INR lev els well below the noise floor . This follows due to the high attenuation at millimeter w av e frequencies, i.e., the networks operate in a noise-limited regime, the stark height difference in deploying FSs and 5G systems, and the v ery low likelihood of UEs being aligned within 1 ◦ of the FS’ s beam. It is also shown 22 -80 -70 -60 -50 -40 -30 -20 -10 0 10 Interference-to-noise ratio (dB) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 CDF Lincoln Park 73.5GHz: Actual 73.5GHz: Random 83.5GHz: Actual 83.5GHz:Random -6dB threshold UE EIRP=33dbm UE EIRP=43dBm (a) Lincoln Park -80 -70 -60 -50 -40 -30 -20 -10 0 10 Interference-to-noise ratio (dB) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 CDF Chicago Loop 73.5GHz: Actual 73.5GHz: Random 83.5GHz: Actual 83.5GHz:Random UE EIRP=43dBm -6dB threshold UE EIRP=33dbm (b) Chicago Loop -80 -70 -60 -50 -40 -30 -20 -10 0 10 Interference-to-noise ratio (dB) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 CDF Lower Manhattan 73.5GHz: Actual 73.5GHz: Random 83.5GHz: Actual 83.5GHz:Random -6dB threshold UE EIRP=43dBm UE EIRP=33dbm (c) Lo wer Manhattan Fig. 11: CDF of INR -80 -70 -60 -50 -40 -30 -20 -10 0 10 Interference-to-noise ratio (dB) 0 0.01 0.02 0.03 0.04 0.05 PDF 73.5GHz UE EIRP= 33dBm UE EIRP= 43dBm -80 -70 -60 -50 -40 -30 -20 -10 0 10 Interference-to-noise ratio (dB) 0 0.01 0.02 0.03 0.04 0.05 PDF 83.5GHz (a) Lincoln Park -80 -70 -60 -50 -40 -30 -20 -10 0 10 Interference-to-noise ratio (dB) 0 0.01 0.02 0.03 0.04 0.05 PDF 73.5GHz UE EIRP= 33dBm UE EIRP= 43dBm -80 -70 -60 -50 -40 -30 -20 -10 0 10 Interference-to-noise ratio (dB) 0 0.01 0.02 0.03 0.04 0.05 PDF 83.5GHz (b) Chicago Loop -80 -70 -60 -50 -40 -30 -20 -10 0 10 Interference-to-noise ratio (dB) 0 0.01 0.02 0.03 0.04 0.05 PDF 73.5GHz UE EIRP= 33dBm UE EIRP= 43dBm -80 -70 -60 -50 -40 -30 -20 -10 0 10 Interference-to-noise ratio (dB) 0 0.01 0.02 0.03 0.04 0.05 PDF 83.5GHz (c) Lo wer Manhattan Fig. 12: PDF of INR that dense urban areas, e.g., downto wn Manhattan, has lower INR due to the increased blockage resulted from the presence of high-rise buildings. Finally , the INR is slightly lower at 83.5GHz compared to 73.5GHz due to the higher path loss in the former . Fig. 12 sho ws the PDF of INR and its 95th percentile for the dif ferent case studies. As it can be seen, only very few FSs may experience high INR v alues, i.e., abo ve the − 6 dB protection threshold, in Lincoln Park, whereas the rest are well protected. This moti v ates implementing the proposed mitigation techniques only to impro ve INR protection at those few FSs, simplifying the 5G coe xistence. W e then look at the INR performance when the 3GPP LOS model is used instead of the actual building layout. In Fig. 13a, we show the CDF of the INR, where the UE EIRP is 43dBm. It is observ ed that for Lincoln Park and Chicago Loop, the INR using the 3GPP LOS model is lo wer than the INR using the actual building layout. This is because the 3GPP LOS model in 23 (a) INR CDF 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2D distance, d u f (m) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Probability of LOS 3GPP probability model Lincoln Park Chicago Loop Lower Manhattan (b) LOS probability Fig. 13: Comparison between the 3GPP LOS model and the actual LOS based on b uildings layouts in terms of: (a) INR CDF and (b) LOS probability . (3) underestimates the LOS probability for larger distances in these cities, as sho wn in Fig. 13b. This is not the case for Lower Manhattan due to the dense deployment of high-rise buildings, i.e., the 3GPP LOS model is shown to be more suitable for areas with denser high-rise buildings. W e remark that we expect the LOS probability to be lower when blockage due to other objects is included, e.g., foliage, cars, etc., making FSs e ven better protected. W e also study the INR performance for other set-ups, where we consider the UE EIRP to be 43dBm. For instance, in Fig. 14a, we sho w the CDF of the INR in the DL, comparing it with that achieved in the UL. Here, the gNB EIRP is 57dBm. It is observ ed that although gNBs hav e higher EIRP , the y do not incur higher interference, compared to UEs, primarily because their antenna tilts point to the ground. In Fig. 14b, we sho w the INR’ s 95th percentile for the multi-user case, where the number of spatially multiplex ed UEs is varied from one to four . It sho wn, as expected, that increasing the number of multiplex ed UEs increases the INR, yet it remains relativ ely low , particularly for the denser areas, e.g., Chicago Loop and Manhattan. C. Impact of Sector-based and Beam-based Mitigation W e focus on a particular FS in Lincoln Park, which has relati v ely high INR in comparison with other FSs. The FS of interest is deplo yed at height of 34m with a wide open-space in its vicinity , making it more susceptible to interference from the 5G system. Here, we only consider operating at 73.5GHz with UE maximum radiated po wer of 43dBm, as this set-up leads to the highest interference. 24 (a) INR in the DL and the UL (73.5GHz) 1234 No. of multiplexed UEs per gNB -30 -25 -20 -15 -10 -5 0 5 INR 95th percentile (dB) 73.5GHz 83.5GHz Lincoln Park Chicago Loop Lower Manhattan (b) 95th percentile for different number of multiplex ed UEs Fig. 14: The INR performance under additional set-ups. Fig. 15 shows the a verage INR on the FS in the presence of sector-based and beam-based mitigation. W e ha ve the following observ ations. First, using lo w protection thresholds, i.e., ψ s and ψ b , does not result in tangible reduction in the 95th percentile of the INR. This follows because UEs tend to point randomly over space and e ven if the main lobe is not aligned, there is still a chance to ha ve high interference from the side lobes. For this reason, larger thresholds provide much better protection. Second, it is sho wn that location-based protection is more reliable than orientation-based. This implies that to get very low INR, it is not enough to protect the boresight of the FS, i.e., signal attenuation due to FS pattern may not be sufficient if the UE effecti v e radiated power is very high. Third, beam-based mitigation slightly outperforms sector -based mitigation, particularly for high thresholds. Equally important, the former also enables better 5G DL coverage, as it makes decisions at higher angular resolution compared to sector-based. The cost of using beam-based mitigation is the increased number of decisions needed to be made for each gNB in vicinity of the FS. Fig. 16 shows the main INR statistics with v ariations of the angular protection threshold. W e sho w the INR performance in the absence of mitig ation for reference. W e also sho w the INR in the presence of spatial exclusion zones with radii 200m and 500, i.e., no gNBs are deployed inside these zones. As expected, the INR is significantly deceased for high protection thresholds. For instance, the 95th percentile decreases by approximately -5.5dB and -13dB when location- based beam mitigation is used with ψ b = 45 ◦ and ψ b = 90 ◦ , respectiv ely . Angular exclusion zones are more ef fectiv e than spatial exclusions as the latter leads to coverage holes in the 5G 25 -70 -60 -50 -40 -30 -20 -10 0 10 20 INR (dB) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 CDF No mitigation 22.5deg: Location 22.5deg: Orientation 45.0deg: Location 45.0deg: Orientation 90.0deg: Location 90.0deg: Orientation -6dB threshold (a) Sector -based -70 -60 -50 -40 -30 -20 -10 0 10 20 INR (dB) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 CDF No mitigation 22.5deg: Location 22.5deg: Orientation 45.0deg: Location 45.0deg: Orientation 90.0deg: Location 90.0deg: Orientation -6dB threshold (b) Beam-based Fig. 15: INR CDF in the presence and absence of passi ve mitigation 0 50 100 Threshold (degrees) -40 -38 -36 -34 -32 -30 -28 -26 INR (dB) Mean No mitigation Exc. zone (200m) Exc. zone (500m) S: Location-based S: Orientation-based B: Location-based B: Orientation-based 0 50 100 Threshold (degrees) -42 -40 -38 -36 -34 -32 -30 -28 INR (dB) Median 0 50 100 Threshold (degrees) -22 -20 -18 -16 -14 -12 -10 -8 INR (dB) 95% Percentile Beam-based Sector-based Fig. 16: INR statistics with v ariations of the angular protection threshold. system. This also emphasizes that the interference is not dominated by UEs that are close to the FS but rather by UEs that ha ve beams directed towards the FS’ s boresight. Fig. 17 shows one snapshot of the FS of interest and the 5G system in vicinity of the FS with and without the mitigation techniques. In the snapshot, we sho w the UE’ s beam used for data communication with its associated gNB as well as the interference generated from the UE into the FS (in dBm). In Fig. 17a, the INR is high as it is dominated by a UE with an interference of − 62 dBm (the noise floor at the FS is approximately − 77 dBm). By using the location-based sector mitigation with ψ s = 45 ◦ , it is sho wn in Fig. 17b that this particular UE switches to a dif ferent gNB, reducing its interference by 66dB! A similar observation is made for the beam- based approach, illustrated in Fig. 17c, where we use ψ b = 22 . 5 ◦ , showing that angular e xclusion zones at a finer scale are suf ficient to protect the FS without compromising the 5G DL coverage. 26 4.4565 4.457 4.4575 4.458 4.4585 4.459 4.4595 x-coord 10 5 4.6423 4.6424 4.6425 4.6426 4.6427 4.6428 4.6429 4.643 4.6431 y-coord 10 6 -469 -174 -109 -147 -62 -235 -220 (a) No mitigation 4.4565 4.457 4.4575 4.458 4.4585 4.459 4.4595 x-coord 10 5 4.6423 4.6424 4.6425 4.6426 4.6427 4.6428 4.6429 4.643 4.6431 y-coord 10 6 -497 -149 -129 -149 -128 -211 -203 (b) Sector -based 4.4565 4.457 4.4575 4.458 4.4585 4.459 4.4595 x-coord 10 5 4.6423 4.6424 4.6425 4.6426 4.6427 4.6428 4.6429 4.643 4.6431 y-coord 10 6 -476 -149 -129 -149 -211 -203 -128 (c) Beam-based Fig. 17: One simulation snapshot of an FS that experiences high INR in the absence of mitigation. The beams used by UEs are shown, and the interference generated by each one into the FS is gi ven in dBm: (a) No mitigation; (b) Sector-based; (c) Beam-based. D. Impact of spatial power contr ol W e set P lo and P up such that the UE EIRP is 33dBm and 43dBm, respecti vely . Fig. 18 sho ws the INR’ s CDF at the FS when the power control (PC) in (20) is used. It is evident that for higher protection thresholds, ψ b , power control can be effecti ve to reduce the INR. F or instance, the 95th percentile reduces from -8dB to -15dB when ψ b = 45 ◦ . Finally , Fig. 19 shows the main INR statistics with v ariations of the protection threshold. It is sho wn that the 95th percentile can 27 -60 -50 -40 -30 -20 -10 0 10 20 INR (dB) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 CDF No mitigation PC: 22.5deg PC: 90.0deg Fig. 18: INR CDF in the presence and absence of po wer control 0 50 100 Threshold (degrees) -35 -34 -33 -32 -31 -30 -29 -28 -27 INR (dB) Mean No mitigation Exc. zone (200m) Exc. zone (500m) PC 0 50 100 Threshold (degrees) -38 -36 -34 -32 -30 -28 INR (dB) Median 0 50 100 Threshold (degrees) -20 -18 -16 -14 -12 -10 -8 -6 INR (dB) 95% Percentile Fig. 19: INR statistics when spatial power control is used be reduced by approximately 10dB without the need to shut of f any beams. E. Comparison of Mitigation T echniques The aforementioned techniques hav e sho wn the ef fectiv eness in mitigating interference at the FS. In this section, we compare them in terms of their impact on the DL cov erage of the 5G system. Using the gNB antenna parameters, it can be sho wn that the maximum radiated po wer is 57dBm. Fig. 20 sho ws a comparison between the dif ferent techniques in terms of the DL coverage. W e only consider location-based protection. Due to the angular exclusion zones created, using larger thresholds, i.e., ψ s and ψ b , inevitably af fect the DL coverage. This is not the case in spatial power control as all beams and sectors are acti ve. Fig. 21 shows the SNR-INR curves of the different mitigation techniques. The curves highlight the different possible operating points of the coexisting 5G and incumbent systems, i.e., the interference lev el expected on the incumbent for a target 5G DL coverage. W e hav e the following observ ations. Comparing location-based beam angular zones with sector angular zones, it is e vident that the former presents more operating points, making it more flexible and effecti v e. 28 0 50 100 Threshold (degrees) 17 18 19 20 21 22 23 SNR (dB) Mean No Mitigation Exc. zone (500m) Sector-based Beam-based PC 0 50 100 Threshold (degrees) 15 16 17 18 19 20 21 SNR (dB) Median 0 50 100 Threshold (degrees) -2 0 2 4 6 8 SNR (dB) Cell-edge Fig. 20: SNR statistics when spatial power control is used 15 16 17 18 19 20 21 Median DL SNR (dB) -22 -20 -18 -16 -14 -12 -10 -8 -6 95th Percentile INR (dB) No mitigation Exclusion Zone Sector-based Beam-based PC (a) Median SNR vs 95th percentile INR - 2 - 1 01234567 Cell-edge SNR (dB) -22 -20 -18 -16 -14 -12 -10 -8 -6 95th Percentile INR (dB) No mitigation Exclusion Zone Sector-based Beam-based PC (b) Cell-edge SNR vs 95th percentile INR Fig. 21: Comparison of SNR-INR curves of different mitigation techniques Ho wev er , both hav e an inevitable trade-of f: higher INR protection incurs 5G coverage de gra- dation. In addition, spatial exclusion zones are not as ef fectiv e as angular e xclusion zones. For instance, using a first-order approximation of the simulated curves, it can be shown that slopes for the beam-based and spatial e xclusion zone techniques are approximately 3 and 1. In other words, by reducing the median SNR by 1dB, the INR reduces by 3dB when using beam-based angular exclusion zones and by 1dB when using spatial exclusion zones. Finally . spatial po wer control enables the reduction of INR with negligible coverage loss. V I I . C O N C L U S I O N The 10GHz of spectrum in the 70GHz and 80GHz bands hav e the potential to enable true mobile connectivity at gigabit speeds. A key obstacle to the 5G deployment of these bands is the presence of incumbent FS systems that require protection from harmful interference. T o this end, we ha ve thoroughly analyzed existing databases to understand the key features and properties of the incumbent system, including the spatial distrib ution and the antenna specifications. In addition, we ha ve analyzed the aggregate interference from 5G UEs using realistic channel 29 models and actual building layouts for accurate results. Our analysis and results have rev ealed that 5G coexistence beyond 70GHz is feasible thanks to the high propagation losses at millimeter wa ve frequencies, the high attenuation due misalignment with the FS antenna boresight, and the deployment geometry of FSs as they tend to be above 5G systems. For FSs that are deployed at relati vely lo w altitudes, we ha ve proposed se veral passiv e miti- gation techniques, including angular exclusion zones and spatial power control. Such techniques require minimal effort from mobile operators and do not require any coordination with the incumbents. W e hav e sho wn that exclusion zones in the angular domain are more effecti v e than spatial exclusion zones. Howe ver , the 5G DL cov erage can be degraded due to the coverage holes introduced by exclusion zones. This can be o vercome via po wer control, where the transmit po wer of the UE depends on the beam’ s direction with respect to the victim FS. R E F E R E N C E S [1] G. Hattab, E. V isotsky , M. Cudak et al. , “Coexistence of 5G mmwa ve users with incumbent fixed stations ov er 70 and 80 GHz, ” in IEEE GLOBECOM , Dec. 2017. [Online]. A v ailable: https://arxi v .org/abs/1801.05393 [2] M. Xiao, S. Mumtaz, Y . Huang et al. , “Millimeter wav e communications for future mobile networks, ” IEEE J. Sel. Areas Commun. , vol. 35, no. 9, pp. 1909–1935, Sep. 2017. [3] “IMT vision- frame work and overall objectiv es of the future dev elopment of IMT for 2020 and beyond, ” ITU-R, M.2083-0, Sep. 2015. [4] J. G. Andrews, S. Buzzi, W . Choi et al. , “What will 5G be?” IEEE J. Sel. Areas Commun. , vol. 32, no. 6, pp. 1065–1082, Jun. 2014. [5] “Use of spectrum bands abo ve 24 GHz for mobile radio services, ” Federal Communications Commission (FCC), T ech. Rep. GN Docket No. 14-177, July 2016. [6] ITU, “W orld radiocommunication conference 2015, ” Geneva Switzerland, Nov . 2015. [7] S. Rangan, T . S. Rappaport, and E. Erkip, “Millimeter -wav e cellular wireless networks: Potentials and challenges, ” Pr oceedings of the IEEE , vol. 102, no. 3, pp. 366–385, Mar . 2014. [8] M. Cudak, T . Ko varik, T . A. Thomas et al. , “Experimental mm wave 5G cellular system, ” in Proc. IEEE Globecom W orkshops , Dec. 2014, pp. 377–381. [9] Y . Inoue, S. Y oshioka, Y . Kishiyama et al. , “Field experimental trials for 5G mobile communication system using 70 GHz- band, ” in Pr oc. IEEE W ir eless Communications and Networking Conf. W orkshops (WCNCW) , Mar . 2017, pp. 1–6. [10] L. Huawei T echnologies Co. (2016, Feb .) Huawei to bring 73GHz mmW av e Mu-MIMO live demo to Deutsche T elekom. [Online]. A vailable: http://www .huawei.com/en/news/2016/2/73GHzmm- W ave- Mu- MIM- live- demo [11] A. K. Gupta, A. Alkhateeb, J. G. Andrews et al. , “Gains of restricted secondary licensing in millimeter wa ve cellular systems, ” IEEE J ournal on Selected Areas in Communications , vol. 34, no. 11, pp. 2935–2950, Nov . 2016. [12] M. Rebato, F . Boccardi, M. Mezzavilla et al. , “Hybrid spectrum sharing in mmW av e cellular networks, ” IEEE T rans. on Cogn. Commun. Netw . , vol. 3, no. 2, pp. 155–168, Jun. 2017. [13] F . Boccardi, H. Shokri-Ghadikolaei, G. Fodor et al. , “Spectrum pooling in MmW ave networks: Opportunities, challenges, and enablers, ” IEEE Commun. Mag. , vol. 54, no. 11, pp. 33–39, Nov . 2016. 30 [14] M. Ghorbanzadeh, E. V isotsky , P . Moorut et al. , “Radar inband and out-of-band interference into L TE macro and small cell uplinks in the 3.5 GHz band, ” in Pr oc. IEEE W ireless Communications and Networking Conf. (WCNC) , Mar . 2015, pp. 1829–1834. [15] ——, “Radar interference into L TE base stations in the 3.5 GHz band, ” Physical Communication , vol. 20, pp. 33 – 47, 2016. [Online]. A v ailable: http://www .sciencedirect.com/science/article/pii/S1874490716300088 [16] A. Khawar , A. Abdelhadi, and T . C. Clancy , “Coexistence analysis between radar and cellular system in LoS channel, ” IEEE Antennas W ir eless Pr opag. Lett. , vol. 15, pp. 972–975, 2016. [17] F . Beltran, S. K. Ray , and J. A. Guti ´ errez, “Understanding the current operation and future roles of wireless networks: Co-existence, competition and co-operation in the unlicensed spectrum bands, ” IEEE Journal on Selected Areas in Communications , vol. 34, no. 11, pp. 2829–2837, Nov . 2016. [18] J. Kim, L. Xian, A. Maltsev et al. , “Study of coexistence between 5G small-cell systems and systems of the fixed service at 39 GHz band, ” in Pr oc. IEEE MTT -S Int. Micr owave Symp , May 2015, pp. 1–3. [19] J. Kim, L. Xian, and A. S. Sadri, “Numerical simulation study for frequency sharing between micro-cellular systems and fixed service systems in millimeter -wav e bands, ” IEEE Access , vol. 4, pp. 9847–9859, 2016. [20] F . Guidolin and M. Neko vee, “In vestigating spectrum sharing between 5g millimeter wave networks and fixed satellite systems, ” in Pr oc. IEEE Globecom W orkshops (GC Wkshps) , Dec. 2015, pp. 1–7. [21] S. Kim, E. V isotsky , P . Moorut et al. , “Coexistence of 5G with the incumbents in the 28 and 70 GHz bands, ” IEEE J. Sel. Ar eas Commun. , vol. 35, no. 6, pp. 1254–1268, Jun. 2017. [22] Y . Liu, X. Fang, M. Xiao et al. , “Decentralized beam pair selection in multi-beam millimeter-wa ve networks, ” IEEE T ransactions on Communications , p. 1, 2018. [23] Y . Cui, X. Fang, Y . F ang et al. , “Optimal non-uniform steady mmwav e beamforming for high speed railway , ” IEEE T ransactions on V ehicular T echnology , p. 1, 2018. [24] A. Alkhateeb, Y . H. Nam, M. S. Rahman et al. , “Initial beam association in millimeter wa ve cellular systems: Analysis and design insights, ” IEEE T rans. W ir eless Commun. , vol. 16, no. 5, pp. 2807–2821, May 2017. [25] M. Giordani, M. Mezzavilla, and M. Zorzi, “Initial access in 5G mmwave cellular networks, ” IEEE Commun. Mag. , vol. 54, no. 11, pp. 40–47, Nov . 2016. [26] C. N. Barati, S. A. Hosseini, M. Mezzavilla et al. , “Initial access in millimeter wa ve cellular systems, ” IEEE T rans. W ireless Commun. , vol. 15, no. 12, pp. 7926–7940, Dec. 2016. [27] Comsearch. (2018) Comsearch online database: 70-90 GHz registration database. [28] M. I. Skolnik, Introduction to Radar Systems . McGraw-Hill, 2001, vol. 3. [29] 3GPP , “Study on channel model for frequency spectrum above 6GHz, ” 3rd Generation Partnership Project (3GPP), T ech. Rep., 2017. [30] “Title 47 of the code of federal regulations, ” Federal Communications Commission (FCC), T ech. Rep. Section 101.115, July 2017. [31] M. Haenggi, Stochastic Geometry for W ireless Networks . Cambridge Univ ersity Press, 2012. [32] T . A. Thomas and F . W . V ook, “System level modeling and performance of an outdoor mmwa ve local area access system, ” in Pr oc. IEEE PIMRC , Sep. 2014, pp. 108–112. [33] Y . Huo, X. Dong, and W . Xu, “5g cellular user equipment: From theory to practical hardware design, ” IEEE Access , vol. 5, pp. 13 992–14 010, 2017. [34] FCC, “Report and order and further notice of proposed rulemaking: Use of spectrum bands above 24 GHz for mobile radio services, ” GN Docket No. 14-177, Jul. 2016.