Guard Band Protection Scheme to Facilitate Coexistence of 5G Base Stations and Radar Altimeters

Abstract

Reformation of the 3.7–4.0 GHz band to expand 5G communication deployment poses a risk of 5G signals disrupting radar altimeter operation, leading to data loss or inaccuracies. Thus, this paper proposes a guard band protection method to facilitate the coexistence of 5G base stations and radar altimeters operating in the 4.2–4.4 GHz band. To enhance the adjacent channel leakage ratio (ACLR), we implemented spectral regrowth on an oversampled waveform using a high-power amplifier model, filtering out-of-band spectral emissions. The results demonstrated that a 150 MHz guard band enables coexistence, except in the case of the 16-by-16 antenna array in rural environments. Notably, for the 4-by-4 antenna array in urban environments, coexistence can be achieved using a 50 MHz guard band. The proposed mitigation techniques may also be extended to promote coexistence between non-terrestrial networks and 5G communication systems, including satellites, unmanned aerial vehicles, and hot air balloons.

1. Introduction

3GPP defines two frequency ranges for 5G new radio (NR) technologies: mid-frequency FR1 (4.1 to 7.125 GHz) and high-frequency FR2 (24.25 to 52.6 GHz) [1]. While the high-frequency band offers advantages such as a wide absolute band and narrow waves, it exhibits several limitations, including high path loss and significant atmospheric attenuation. In contrast, the mid-frequency band is superior in terms of transmission range and cost-effectiveness. Consequently, many countries are accelerating the deployment of 5G mid-frequency bands, typically ranging from 3.4 to 4.0 GHz, as shown in Figure 1. Notably, the frequencies in the 3.7 to 4.0 GHz mid-band are being considered and reallocated for 5G usage, particularly in South Korea [2,3].

However, in 2020, the Radio Technical Commission for Aeronautics reported that the frequency ranging from 3.7 GHz to 4.2 GHz may cause significant interference with radar altimeters [5]. These frequency bands closely align with those used by radar altimeters, leading to radio-frequency interference from 5G communication. Figure 2 illustrates the relationship between the fundamental and spurious emissions from 5G and the radar altimeter band. Interference from the 5G communication system on the radar altimeter consists of two aspects: (1) in-band interference, where out-of-band spurious signals from the 5G system fall within the radar altimeter frequency band ranging from 4.2 GHz to 4.4 GHz, and (2) out-of-band interference, where 5G in-band signals block the radar altimeter.

Various researchers have explored the interference from ground-based communication systems on radar altimeters. For example, in [6], an interference analysis method was proposed to evaluate the potential for spectrum sharing between the long-term evolution system and radar altimeters in the 4.2 GHz to 4.4 GHz band. However, the study did not conclusively determine the feasibility of such sharing. In another study [7], the interference effects of 5G base stations on radar altimeters were analysed, focusing on suburban environments. Previously, in [8] we developed a fundamental mathematical model for calculating the interference received by radar altimeters from 5G base stations. Building on this foundation, we extended the mathematical model to more accurately evaluate received interference by incorporating real environmental parameters, including a detailed path loss model across different environments as described in [9]. In [9], we performed an in-depth analysis of the radar altimeter signal path loss at varying altitudes (from 1500 ft to 50 ft) in rural, suburban, and urban environments. Additionally, we introduced a power control-aided distance protection method to enable the coexistence of 5G base stations and radar altimeters. Exclusion and restriction zones were defined based on factors such as the base station transmitted power, channel gain, and protection criteria. 5G base stations could not be deployed in the exclusion zone. Many countries currently use this strategy to introduce geographic separation between base stations and airport runways to mitigate interference. For instance, in Japan, 5G base stations are deployed 200 m away from airplane approach routes [10]. In the United States, operators have set a 1.5-mile 5G exclusion zone at airports [11]. Reference [12] proposed an adaptive beamforming scheme to mitigate interference between 5G base stations and radar altimeters, demonstrating that the scheme effectively eliminates sidelobes and ensures accurate altitude measurements, with potential applicability to coexistence with other non-terrestrial networks. Reference [13] investigated the coexistence of 5G systems and radar altimeters, finding that interference issues arise at low altitudes and short separation distances. Reference [14] examined potential interference between 5G networks and aviation altimeters, highlighting concerns over safety and operational disruptions, and recommends enhanced collaboration among regulators, and industry stakeholders to ensure safe and effective coexistence through informed technical and policy measures. Reference [15] addressed aviation safety concerns related to 5G interference with radar altimeters, proposing a comprehensive approach that includes redesigning the altimeter system with a bandpass filter and cognitive radio-based dynamic filtering to mitigate interference while ensuring compliance with FAA regulations and minimal disruption to existing systems. Reference [16] highlighted the challenges 5G poses to various wireless systems, proposing that future wireless technologies, including 6G, should prioritise adaptive and reconfigurable spectrum-sharing approaches, supported by parallel technological and policy developments, to prevent interference and enhance coexistence with critical systems. Reference [17] conducted a detailed interference analysis between 5G systems and radar altimeters, revealing that 5G networks in the C-band pose a higher risk of interference at lower altitudes, a crucial phase for flight safety, and suggests methods to mitigate this issue. Reference [18] presents a deployment protection method to safeguard aeronautical radar altimeters from 5G interferences, using an integrated interference model to define prohibited zones and employing power control and angle shutoff techniques, which are validated across various scenarios, offering practical guidelines for safe base station deployment near airports.

In this paper, we introduce a protection method inspired by [19], named the guard band protection scheme, to facilitate the coexistence of 5G systems and radar altimeters. Compared to other proposed protection methods, such as redesigning the radar altimeter hardware or imposing restrictive measures like establishing prohibited zones or shutting off certain beam sweeping angles, our proposed guard band protection scheme maintains full deployment freedom for 5G infrastructure while ensuring robust protection for radar altimeters. This approach not only preserves the integrity and functionality of both technologies but also minimises the need for extensive modifications, making it a more practical and cost-effective solution for the coexistence of 5G and radar altimeter systems. The main contributions of this work can be summarised as follows:

• A guard band protection scheme to enable the coexistence of 5G base stations and radar altimeters has been proposed;
• A filter to mitigate the out-of-band spectral emissions of the 5G waveform has been developed to improve the performance of ACLR;
• The Monte Carlo method has been employed to validate the effectiveness of the proposed guard band protection scheme.

The remaining paper is organised as follows. Section 2 introduces the proposed guard band protection method and outlines the calculation of relevant parameters. Section 3 describes the analysis of the feasibility of guard band protection using Monte Carlo simulations. Section 4 presents the concluding remarks.

2. Modelling for Guard Band Protection and Analysis

We develop an interference analysis model for the guard band protection method, as illustrated in Figure 3. The guard band, an additional unused frequency, is established between 5G base stations and radar altimeters to prevent interference. With guard band protection, the interference signal consists of two components [20]. The first component consists of emissions from 5G systems in their assigned channel, which are received by the radar altimeter in its adjacent channel. The second component involves emissions from 5G systems in their adjacent channel, which the radar altimeter receives in its assigned channel. In the former case, the sensitivity of the radar altimeter to interference is determined by its adjacent channel selectivity (ACS), while the quantity of interfering emissions in the latter is influenced by the adjacent channel leakage ratio (ACLR) of 5G base stations.

2.1. Mathematical Model for Guard Band Protection

The power of the interference signal received by the radar altimeter from the ith 5G base station is calculated as [20]

$${\mathrm{I}}_{\mathrm{i}}={\mathrm{P}}_{\mathrm{t},\mathrm{i}}+{\mathrm{G}}_{\mathrm{BS},\mathrm{i}}+{\mathrm{G}}_{\mathrm{RA}}-{\mathrm{L}}_{\mathrm{total},\mathrm{i}}-\mathrm{ACIR},$$

(1)

where $P_{\mathrm{t},\mathrm{i}}$ is the transmitted power of a single 5G base station, $G_{\mathrm{BS},\mathrm{i}}$ is the antenna gain of the ith 5G base station, and $G_{\mathrm{RA}}$ is the antenna gain of the radar altimeter. $ACIR$ is the adjacent channel interference ratio, and $L_{\mathrm{total},\mathrm{i}}$ is the total loss.

The gain of a 5G base station is determined by its antenna gain and beamforming gain, expressed as

$${\mathrm{G}}_{\mathrm{BS},\mathrm{i}}={\mathrm{A}}_{\mathrm{element}}+{\mathrm{A}}_{\mathrm{array}},$$

(2)

where ${\mathrm{A}}_{\mathrm{element}}$ and ${\mathrm{A}}_{\mathrm{array}}$ denote the gains of the antenna element and beamforming, respectively [21]. The gain of the antenna element is given by:

$${\mathrm{A}}_{\mathrm{element}}={\mathrm{G}}_{\mathrm{element}}-\min \left\{-\lfloor {\mathrm{A}}_{\mathrm{e},\mathrm{h}}\left(\mathsf{\phi }\right)+{\mathrm{A}}_{\mathrm{e},\mathrm{v}}\left(\mathsf{\theta }\right)\rfloor ,{\mathrm{A}}_{\mathrm{m}}\right\},$$

(3)

where ${\mathrm{G}}_{\mathrm{element}}$ is the maximum antenna element gain, and ${\mathrm{A}}_{\mathrm{m}}$ is the front-to-back ratio. ${\mathrm{A}}_{\mathrm{e},\mathrm{h}}\left(\mathsf{\phi }\right)$ represents the antenna radiation pattern in horizontal direction, which can be calculated as:

$${\mathrm{A}}_{\mathrm{e},\mathrm{h}}\left(\mathsf{\phi }\right)=-\min \left[12{\left({\displaystyle \frac{\mathsf{\phi }}{{\mathsf{\phi }}_{3\mathrm{dB}}}}\right)}^2,{\mathrm{A}}_{\mathrm{m}}\right],$$

(4)

where $\mathsf{\phi }$ denotes the azimuth angle, and ${\mathsf{\phi }}_{3\mathrm{dB}}$ is the 3-dB beamwidth of the horizontal pattern.

${\mathrm{A}}_{\mathrm{e},\mathrm{v}}\left(\mathsf{\theta }\right)$ denotes the antenna radiation pattern in vertical direction, which can be expressed as:

$${\mathrm{A}}_{\mathrm{e},\mathrm{v}}\left(\mathsf{\theta }\right)=-\min \left[12{\left({\displaystyle \frac{\mathsf{\theta }-90}{{\mathsf{\theta }}_{3\mathrm{dB}}}}\right)}^2,{\mathrm{SLA}}_{\mathrm{v}}\right],$$

(5)

where $\mathsf{\theta }$ is the elevation angle, ${\mathsf{\theta }}_{3\mathrm{dB}}$ is the 3-dB beamwidth of the vertical pattern, and ${\mathrm{SLA}}_{\mathrm{v}}$ denotes the front-to-back ratio.

The antenna array gain, ${\mathrm{A}}_{\mathrm{array}}$, is expressed as:

$${\mathrm{A}}_{\mathrm{array}}=10{\log }_{10}\left({\left|{{\displaystyle \sum }}_{\mathrm{m}=1}^{{\mathrm{N}}_{\mathrm{H}}}{{\displaystyle \sum }}_{\mathrm{n}=1}^{{\mathrm{N}}_{\mathrm{V}}}{\mathrm{w}}_{\mathrm{n},\mathrm{m}}·{\mathrm{v}}_{\mathrm{n},\mathrm{m}}\right|}^2\right),$$

(6)

where ${\mathrm{N}}_{\mathrm{H}}$ and ${\mathrm{N}}_{\mathrm{V}}$ are the number of antenna elements in the horizontal rows and vertical columns of the antenna array, respectively. ${\mathrm{v}}_{\mathrm{n},\mathrm{m}}$ is the super position vector, which is given by:

$${\mathrm{v}}_{\mathrm{n},\mathrm{m}}=\exp \left\{\sqrt{-1}·2\mathsf{\pi }\left[\left(\mathrm{n}-1\right){\displaystyle \frac{{\mathrm{d}}_{\mathrm{V}}}{\mathsf{\lambda }}}\cos \mathsf{\theta }+\left(\mathrm{m}-1\right){\displaystyle \frac{{\mathrm{d}}_{\mathrm{H}}}{\mathsf{\lambda }}}\sin \mathsf{\theta }\sin \mathsf{\phi }\right]\right\},$$

(7)

where ${\mathrm{d}}_{\mathrm{V}}$ and ${\mathrm{d}}_{\mathrm{H}}$ are the radiating antenna element spacings, and $\mathsf{\lambda }$ is the wavelength.

${\mathrm{w}}_{\mathrm{n},\mathrm{m}}$ represents the beamforming weighting, which can be calculated as:

$${\mathrm{w}}_{\mathrm{n},\mathrm{m}}={\displaystyle \frac{1}{\sqrt{{\mathrm{N}}_{\mathrm{H}}{\mathrm{N}}_{\mathrm{V}}}}}\exp \left\{\sqrt{-1}·2\mathsf{\pi }\left[\left(\mathrm{n}-1\right){\displaystyle \frac{{\mathrm{d}}_{\mathrm{V}}}{\mathsf{\lambda }}}\sin \left({\mathsf{\theta }}_{\mathrm{etilt}}\right)+\left(\mathrm{m}-1\right){\displaystyle \frac{{\mathrm{d}}_{\mathrm{H}}}{\mathsf{\lambda }}}\cos \left({\mathsf{\theta }}_{\mathrm{etilt}}\right)\cos \left({\mathsf{\phi }}_{\mathrm{escan}}\right)\right]\right\},$$

(8)

where ${\mathsf{\theta }}_{\mathrm{etilt}}$ denotes down-tilt angle, and ${\mathsf{\phi }}_{\mathrm{escan}}$ is the direction of the beamforming.

The gain of the radar altimeter depends on the incident angle and 3-dB beamwidth of the altimeter antenna [22]:

$${\mathrm{G}}_{\mathrm{RA}}=-{\displaystyle \frac{12}{{{\varnothing }_{3\mathrm{dB}}}^2}}·{\varnothing }^2+{\mathrm{G}}_{\mathrm{RA},\max },$$

(9)

where $\varnothing$ is the incident angle, ${\varnothing }_{3\mathrm{dB}}$ is the 3-dB beamwidth of the radar altimeter antenna, and $G_{\mathrm{RA},\max }$ is the maximum antenna gain of the radar altimeter.

Furthermore, the total loss between the 5G base station and radar altimeter is calculated considering the losses attributable to system components. Notably, the various propagation losses that occur when the interfering signal of the 5G base station reaches the radar altimeter are expressed as [20]

$${\mathrm{L}}_{\mathrm{total},\mathrm{i}}={\mathrm{PL}}_{\mathrm{i}}+{\mathrm{L}}_{\mathrm{body}}+{\mathrm{L}}_{\mathrm{polar}}+{\mathrm{L}}_{\mathrm{cable}},$$

(10)

where $P{L}_{\mathrm{i}}$ is the path loss [23], which depends on the distance between the ith 5G base station and radar altimeter, antenna height of the 5G base station, and building height. $L_{\mathrm{body}}$, $L_{\mathrm{polar}}$, and $L_{\mathrm{cable}}$ denote the body loss, polarisation loss, and cable loss, respectively, assumed to be constants.

ACIR is the ratio of the total interference received by the radar altimeter between ad-jacent channels, defined as [20]

$$\mathrm{ACIR}=-10{\log }_{10}\left({\displaystyle \frac{1}{{10}^{\frac{\mathrm{ACLR}}{10}}}}+{\displaystyle \frac{1}{{10}^{\frac{\mathrm{ACS}}{10}}}}\right),$$

(11)

where the calculations of ACLR and ACS are outlined in Section 2.2 and Section 2.3, respectively.

The aggregate received interference by the radar altimeter is expressed as

$${\mathrm{I}}_{\mathrm{agg}}=10{\log }_{10}\left({{\displaystyle \sum }}_{\mathrm{i}=1}^{{\mathrm{N}}_{\mathrm{BS}}}{10}^{{\displaystyle \frac{{\mathrm{I}}_{\mathrm{i}}}{10}}}\right),$$

(12)

where $N_{\mathrm{BS}}$ is the number of 5G base stations within the coverage of the radar altimeter, calculated as

$${\mathrm{N}}_{\mathrm{BS}}=\lceil {\left({\displaystyle \frac{{\mathrm{h}}_{\mathrm{RA}}·\tan \left(\frac{{\varnothing }_{3\mathrm{dB}}}{2}\right)}{{\mathrm{R}}_{\mathrm{BS}}}}\right)}^2\rceil ,$$

(13)

where $h_{\mathrm{RA}}$ is the altitude of the airplane, and $R_{\mathrm{BS}}$ is the cell radius of the 5G base station.

2.2. ACLR

ACLR indicates the ratio of the mean power centred on the assigned channel frequency to the mean power centred on an adjacent channel frequency:

$$\mathrm{ACLR}=\overline{{\mathrm{P}}_{\mathrm{input}}-{\mathrm{P}}_{\mathrm{channel}}},$$

(14)

where $P_{\mathrm{input}}$ is the input signal power, and $P_{\mathrm{channel}}$ is the channel power in dBm, defined as [24]

$${\mathrm{P}}_{\mathrm{channel}}=10{\log }_{10}\left[{\displaystyle \frac{{\mathrm{BW}}_{\mathrm{channel}}}{{\mathrm{BW}}_{\mathrm{re}}·{\mathrm{k}}_{\mathrm{n}}}}·{\displaystyle \frac{1}{\mathrm{N}}}·{{\displaystyle \sum }}_{\mathrm{i}=1}^{\mathrm{N}}{10}^{{\displaystyle \frac{{\mathrm{P}}_{\mathrm{i}}-{\mathrm{a}}_{\mathrm{i}\left(\mathrm{RRC}\right)}}{10}}}\right],$$

(15)

where $B{W}_{\mathrm{channel}}$ is the channel bandwidth (in kHz), $B{W}_{\mathrm{re}}$ is the resolution bandwidth, $k_{\mathrm{n}}$ is the correction factor for noise bandwidth of a specific resolution, $N$ is the number of pixels within the channel, $P_{\mathrm{i}}$ is the level represented by pixel $i$ of the trace (in dBm), and $a_{\mathrm{i}\left(\mathrm{RRC}\right)}$ represents the attenuation of the 3GPP root-raised cosine (RRC) filter at pixel $i$ (in dB).

We assume that 5G downlinks are allocated to the 3.7–4.0 GHz band with a bandwidth of 100 MHz. The subcarrier spacing is set as 30 kHz, and the duplexing mode is time division duplexing (TDD). To generate the 5G NR test model (TM) waveforms, we use the MATLAB R2024a class hNRReferenceWaveformGenerator, with the parameters outlined in

Table 1. 5G NR-TM waveform parameters.

Parameters	Unit	Value
Model		DL-FRC-FR1-QPSK
Channel bandwidth	MHz	100
Subcarrier spacing	kHz	30
Duplexing mode		TDD
Cell identity		1
Windowing	%	0
Sampling rate	Hz	614,400,000

.

Subsequently, we use the helper function hACLRParametersNR to acquire the necessary parameters for measuring ACLR from the generated waveform. Finally, the helper function hACLRMeasurementNR is adopted to calculate the input signal power in the main channel and ACLR using a square window on adjacent channels [25]. Through these calculations, the ACLR is determined to be 64.166 dB. Notably, the orthogonal frequency division multiplexing (OFDM) signal is derived through the inverse continuous-time Fourier transform, which can be challenging for many hardware implementations. Therefore, the modulation process of OFDM is accomplished using the inverse discrete Fourier transform. As depicted in Figure 4a, excessive out-of-band powers, attributable to extreme side lobes, results from implicit windowing in the OFDM time domain. To mitigate interference in adjacent channels, we design a filter to enhance the ACLR [26].

Figure 5 illustrates the process flow for ACLR measurement with filter design. The design process is initiated by generating 5G NR-TM waveforms, specifying parameters as outlined in Table 1. Subsequently, a low-pass filter is devised with its stopband centred at 3.95 GHz and a bandwidth of 50 MHz. The transition band starts at the edge of the occupied bandwidth and stops at the edge of the channel bandwidth. Moreover, the stopband bandwidth is set as 49.14 MHz, accompanied by an attenuation of 80 dB. To ensure perfect reconstruction of the signal, oversampling of the filtered 5G NR-TM waveforms is performed at a sampling frequency twice the highest frequency of the input signal. Following the oversampling, a high-power amplifier model is applied to introduce out-of-band distortion, thereby generating adjacent channel carrier signals. This filter design aims to shape the spectrum within the original bandwidth of the waveform. The transition band begins at the edge of the occupied transmission bandwidth and ends at the edge of the overall channel bandwidth. Unlike other advanced filter technologies that rely on narrowband high-Q filters made with specialized materials [27], our approach focuses on efficiently managing spectral leakage within the existing transmission system without altering the sampling rate or requiring additional complex hardware.

Figure 4b illustrates the ACLR performance following waveform filtering. Power leakage in the first adjacent channel is significantly reduced. This improvement is attributable to the implementation of a low-pass filter in the frequency domain, effectively cancelling out a specific amplitude of the OFDM reference signal and achieving peak cancellation. This peak-cancellation technique serves to mitigate undesirable out-of-band interference. Furthermore, complex calculations for ACLR are facilitated using MATLAB helper functions, specifically hACLRParametersNR and hACLRMeasurementNR. In the first adjacent channels, ACLR increases by 23.1785 dB and 22.8709 dB. Ultimately, after waveform filtering, ACLR is optimised to 75.6678 dB.

2.3. ACS and ACIR

The ACS of radar altimeters is detailed in [4], indicating that the filter attenuation of the radar altimeter system is proportional to 24 dB per octave, reaching a maximum of 40 dB. Figure 6 depicts the ACS of radar altimeters against frequency.

Given that the guard band represents the frequency separation between the edges of the radar altimeter channel and base station channel, the offset frequency is used to compute the ACS of radar altimeters with guard band protection as

$${\mathrm{f}}_{\mathrm{offset}}={\displaystyle \frac{1}{2}}{\mathrm{BW}}_{\mathrm{BS}}+\mathrm{GB}.$$

(16)

where $B{W}_{\mathrm{BS}}$ is the channel bandwidth of the 5G base station, and $GB$ denotes the bandwidth of the guard band.

Leveraging the calculations of ACLR and ACS, we investigate the ACIR vs. the guard band, ranging from 0 to 200 MHz in increments of 5 MHz (Figure 7). According to the spectrum allocation of South Korea, 200 MHz is the maximum buffering frequency. The ACIR increases with the expansion of the guard band. Notably, the ACIR exhibits abrupt changes when the guard band reaches 50 MHz and 150 MHz but not at other values. Contrary to expectations, the ACIR values with and without filtering exhibit no significant differences, as highlighted in red. This is because most radar altimeter designs originated in the 1990s, a period when mobile communication primarily operated within the 2 GHz range, with only minimal satellite downlink signals above 3 GHz. Consequently, radar altimeters were not initially designed to contend with the widespread presence of high-power mobile communications. During that era, high-performance filters were not commonly integrated into the receivers of radar altimeter systems. The subpar out-of-band power emission performance of radar altimeters led to a series of nonlinear responses. Therefore, although the ACLR performance is improved by using a filter, the improvement is not significant.

3. Numerical Results

We use the aforementioned interference analysis model to evaluate the interference received by radar altimeters from 5G base stations, leveraging the guard band protection method based on Monte Carlo simulation. The received interference is determined by the random placement of 5G base stations within the radar altimeter coverage, random antenna pointing and beamforming directions of the 5G base station, and random channel conditions. The simulation results are obtained by repeating 1000 snapshots with randomness, maintaining the same environment and the same antenna pattern at the same radar altimeter height. Introducing new random variables in each simulation snapshot allows us to derive the variation in interference signal power from 5G base stations on the radar altimeter. Subsequently, the simulation results are analysed and expressed as a cumulative distribution function (CDF), which is compared with the threshold of the radar altimeter at −6 dB to explore the possibility of coexistence between 5G base stations and the radar altimeter after using the guard band protection method.

As various factors influence the airplane approach speed, a uniform descent speed is assumed. The simulation uses data for the Boeing 737-800. 5G downlinks are assigned to the 3.7–4.0 GHz band with a 100 MHz bandwidth, while the radar altimeter operates in the 4.2–4.4 GHz band with a 196 MHz channel bandwidth. While most airports are situated on the outskirts of a city, some are located in urban areas, such as heliports. Hence, we assume 5G base stations are distributed in the rural, suburban, and urban environments. The 5G base station is modelled using parameters from

Table 2. 5G base station parameters [20].

Parameters	Unit	Value
Carrier frequency	GHz	3.7–4.0
BS channel bandwidth	MHz	100
UE channel bandwidth	MHz	20
BS antenna height	m	Rural: 35 Suburban: 25 Urban: 20
Antenna array		4 × 4, 8 × 8, and 16 × 16
Vertical 3-dB beamwidth	◦	65
Horizontal 3-dB beamwidth	◦	90
Vertical element spacing		0.5λ
Horizontal subarray spacing		2.1λ
Building height average	m	5
Body loss	dB	4
Polarisation loss	dB	3
Noise temperature	K	290
Modulation		QPSK [28]
Centre frequency	GHz	3.95
Front-to-back ratio	dB	30
Elevation angle	◦	[−90, 90]
Azimuth angle	◦	[−180, 180]
Polarisation	◦	45
Duplexing		TDD
Mechanical down tilt angle	◦	Rural: 3 Suburban: 6 Urban: 10
Cell radius	m	Rural: 1200 Suburban: 600 Urban: 300
Transmitted power	dBm	Rural: 71 [29] Suburban: 68 Urban: 68
Element gain	dBi	6.4
UE antenna height	m	1.5

, and the simulation involves parameters from

Table 3. Parameters of the radar altimeter [4].

Parameters	Unit	Value
Nominal centre frequency	GHz	4.3
Channel bandwidth	MHz	196 [30]
Carrier frequency	GHz	4.2–4.4
Maximum antenna gain	dB	10
Noise figure	dB	6
Cable loss	dB	6
−3 dB beamwidth	◦	55
Modulation		FMCW
Landing downward angle	◦	15 [31]
Landing speed average	m/s	74.588 [32]
Fuselage length	m	39.5 [33]
Interference protection criteria	dB	−6

for the radar altimeter.

According to our previous work [9], the interference received by the radar altimeter from 5G base stations increases as the airplane altitude decreases. The received interference peaks at altitudes of 116 ft, 87 ft, and 75 ft in rural, suburban and urban environments, respectively, as outlined in

Table 4. Peak received interference without any protective measure [9].

Environment	Antenna Array	Received Interference (dB)
Rural	4 × 4 8 × 8 16 × 16	34.3446 40.3652 46.3858
Suburban	4 × 4 8 × 8 16 × 16	22.3548 28.3754 34.396
Urban	4 × 4 8 × 8 16 × 16	12.0633 18.0839 23.6751

. We perform a detailed analysis by simulating the airplane at these altitudes in the rural, suburban, and urban environments with 50 MHz and 150 MHz guard bands, as well as without a guard band.

3.1. Rural Areas

Figure 8 displays the simulation results for different antenna arrays, i.e., 4-by-4, 8-by-8, and 16-by-16 antenna arrays. The solid blue line represents the scenario without a guard band, indicating no protection. The solid red line corresponds to the case with a 50 MHz guard band, while the solid yellow line represents the scenario with a 150 MHz guard band. The simulation results are compared with the protection threshold, set as −6 dB and indicated by the red-dotted line. This threshold serves as a reference point, functioning as the interference limit.

Figure 8 shows a clear trend that as the guard band width increases, the received INR values are significantly reduced, resulting in a leftward shift of the CDF curves. As expected, the protection threshold of the radar altimeter is not exceeded when using a 150 MHz guard band for the 4-by-4 and 8-by-8 antenna arrays. In this case, the maximum received interference is −12.2302 dB and −7.9293 dB with the 4-by-4 and 8-by-8 antenna arrays, respectively. In contrast, with the 16-by-16 antenna array, the maximum received interference reaches 2.0897 dB, with a threshold-exceeding probability of 0.03%. When the 50 MHz guard band is used, the maximum received interference is 8.4409 dB, 16.7643 dB, and 24.5277 dB with the 4-by-4, 8-by-8, and 16-by-16 antenna arrays, respectively. The probabilities of exceeding the protection threshold are 98.9%, 96.6%, and 94.8%, with the 4-by-4, 8-by-8, and 16-by-16 antenna arrays, respectively. Furthermore, the maximum received interference without the protection method is 33.6218 dB, 38.4036 dB, and 44.1662 dB with the 4-by-4, 8-by-8, and 16-by-16 antenna arrays, respectively. These values closely align with those summarised in Table 4. The probabilities of exceeding the protection threshold are 76.8%, 63.3%, and 49.9% with the 4-by-4, 8-by-8, and 16-by-16 antenna arrays, respectively. The probability of exceeding the threshold decreases with an increase in the number of antenna arrays and a decrease in the guard band span. In summary, the 150 MHz guard band effectively protects radar altimeters when 5G base stations use 4-by-4 and 8-by-8 antenna arrays, keeping the received interference levels below the −6 dB threshold. However, the proposed guard band protection method may not be entirely effective for larger antenna arrays, such as the 16-by-16 antenna array, in rural environments. Given the importance of ensuring aviation safety, deploying 5G base stations with 16-by-16 or larger antenna arrays near airports in rural areas is not recommended under the proposed guard band protection scheme.

3.2. Suburban Areas

In the suburban environment (Figure 9), all curves using a 150 MHz guard band remain below the red-dotted line. The maximum received interference is −15.0214 dB, −13.8884 dB, and −11.7434 dB for the 4-by-4, 8-by-8, and 16-by-16 antenna arrays, respectively. This indicates that the proposed guard band protection scheme effectively mitigates interference to acceptable levels in suburban environments. In contrast, in rural environments, a 5G base station equipped with a 16-by-16 antenna array presents a significant interference risk, even with a 150 MHz guard band. This discrepancy arises because signal propagation in suburban areas experiences higher path loss compared to rural areas. Conversely, when using a 50 MHz guard band, all curves surpass the threshold. Specifically, the maximum received interference is −2.4096 dB, 6.2689 dB, and 12.6532 dB for the 4-by-4, 8-by-8, and 16-by-16 antenna arrays, respectively. In these antenna configurations, the probabilities of exceeding the threshold are 99.4%, 98.1%, and 95.6%, respectively. Additionally, in the absence of a guard band, the maximum received interference reaches 21.6080 dB, 26.8996 dB, and 34.0605 dB with the 4-by-4, 8-by-8, and 16-by-16 antenna arrays, respectively. The probabilities of meeting the threshold are 74.7%, 59.8%, and 41.5%, respectively. These values closely align with the peak values without protection, as summarised in Table 4.

3.3. Urban Areas

The trends observed in the urban environment are similar to those in the suburban environment (Figure 10). All interference values when using a 150 MHz guard band meet the protection threshold, regardless of the antenna array configuration (4-by-4, 8-by-8, or 16-by-16). The maximum received interference is −20.7453 dB, −16.9017 dB, and −14.2304 dB for the 4-by-4, 8-by-8, and 16-by-16 antenna arrays, respectively. This indicates that a 150 MHz guard band provides robust interference mitigation, even when using larger antenna arrays in densely built environments. Notably, the curve for the 50 MHz guard band with the 4-by-4 antenna array approaches but does not exceed the −6 dB threshold, with a maximum received interference of −6.7601 dB. This confirms that a 50 MHz guard band is sufficient for the urban environment with the 4-by-4 antenna array. With a 50 MHz guard band, the maximum received interference is −0.0126 dB and 6.1357 dB for the 8-by-8 and 16-by-16 antenna arrays, respectively. The corresponding probabilities of meeting the protection threshold are 99.8% and 98.7%, respectively. Without protection, the maximum received interference is 11.1755 dB, 17.8717 dB, and 22.7011 dB with the 4-by-4, 8-by-8, and 16-by-16 antenna arrays, respectively. These values match those outlined in Table 4. The probabilities of meeting the protection threshold are 87.3%, 74.2%, and 55%, respectively. Urban environments typically involve more physical obstructions, such as skyscrapers, compared to suburban and rural areas. These obstacles can absorb, reflect, and scatter 5G signals, resulting in higher path loss and reducing direct interference with radar altimeters. This natural attenuation explains why narrower guard bands will still be effective with smaller antenna arrays in urban areas.

Regarding practical implementation, the variability of the simulation results underscores the challenge of establishing a fixed guard band that effectively balances spectrum efficiency and interference mitigation. A one-size-fits-all approach could either under-protect radar altimeters or lead to inefficient spectrum use. Static guard band assignments do not account for dynamic factors such as weather conditions and varying network loads. To address these limitations, sensors and monitoring systems could be deployed in strategic locations, such as areas near flight paths, to provide real-time data on interference levels. Network management systems could then use this real-time feedback to dynamically adjust the guard band width.

4. Conclusions

This paper proposes an interference-analysis model based on the guard band protection method to investigate the possibility of coexistence between 5G base stations and radar altimeters. For accurate simulation, we use the Monte Carlo method, model the 5G base station using 4-by-4, 8-by-8, and 16-by-16 antenna arrays, distribute the interference scenario in different environments, including rural, suburban, and urban areas, and consider various types of losses. The results indicate that using a 150 MHz guard band can facilitate the coexistence of 5G base stations and the radar altimeter, except in the case of the 16-by-16 antenna array in the rural environment. Notably, in the urban environment, coexistence can be achieved with only a 50 MHz guard band for the 5G base station using the 4-by-4 antenna array. These simulation results can provide a valuable reference for researchers, and the results without protection align well with our previous findings.

Motivated by these promising results, future work focused on interference mitigation can be aimed at shutting off the partial vertical beamforming angle, limiting the deployment location of the 5G base stations, and decreasing the antenna height of the 5G base stations. Additionally, the proposed protection methods may be confirmed with using practical measured data.