What are the beamforming architectures?

Analog beamforming: each element has an RF phase shifter (and optionally a variable gain amplifier). All elements are combined into a single RF chain before the ADC. Produces one beam at a time. Simple, low power, but limited flexibility. Used in 5G FR2 user equipment and legacy radar. Digital beamforming: each element has its own complete RF chain (LNA, mixer, ADC or PA, mixer, DAC). Beamforming is performed in the digital domain, allowing simultaneous formation of multiple beams in any direction, adaptive nulling, and MIMO spatial multiplexing. Maximum flexibility but highest cost, power, and data rate. Used in military AESA radar and base station massive MIMO. Hybrid beamforming: combines analog and digital. Elements are grouped into sub-arrays; analog phase shifters steer each sub-array, and digital processing combines sub-arrays. Offers a practical trade-off: fewer RF chains than full digital but multiple simultaneous beams. The dominant architecture for 5G NR gNB at mmWave.

How does the array factor work?

The array factor describes the combined radiation pattern of the array, separate from the individual element pattern. For a uniform linear array of N elements with spacing d: AF(theta) = sum from n=0 to N-1 of w_n times exp(j n k d sin theta), where w_n is the complex weight (amplitude and phase) of element n, k = 2 pi / lambda, and theta is the angle from broadside. The total pattern is the product of the element pattern and the array factor (pattern multiplication). To steer the beam to angle theta_0, set the phase of each element to: phi_n = minus n k d sin theta_0. This creates constructive interference at theta_0. The beamwidth narrows as N increases: BW approximately equals 0.886 lambda / (N d cos theta_0). The array gain equals N times the element gain (in linear), or 10 log N plus element gain in dB. Grating lobes appear when d exceeds lambda / 2.

What is massive MIMO beamforming?

Massive MIMO uses arrays with 64 to 256 or more antenna elements at the base station to simultaneously serve multiple users on the same time-frequency resource through spatial multiplexing. Unlike conventional beamforming which forms a single beam, massive MIMO forms one beam per user, creating spatial channels that are nearly orthogonal when the number of antennas is much larger than the number of users. The capacity scales linearly with the minimum of the number of antennas and users. A 64-element array serving 8 users simultaneously achieves approximately 8 times the spectral efficiency of a single-antenna system. Channel estimation is critical: in TDD, uplink pilots enable downlink channel estimation through reciprocity. In FDD, the overhead scales with the number of antennas, making massive MIMO impractical for FDD above 32 elements. 5G NR supports up to 256 antenna ports for massive MIMO.

Phased Array

Beam Forming

/beem form-ing/

Electronic beam steering via phase/amplitude weighting across N antenna elements. AF(θ) = Σw_nexp(jnkd sinθ). Steer: φ_n = −nkd sinθ₀. Gain = 10log(N) + G_element. BW = 0.886λ/(Nd cosθ). Analog: RF phase shifters, 1 beam. Digital: per-element ADC, multi-beam. Hybrid: sub-arrays. 5G mMIMO: 64-256 elements, 8-64 beams.

Gain: 10log(N)

Spacing: λ/2

5G: 64-256 el

Understanding Beamforming

Beamforming is the enabling technology for 5G mmWave, modern radar, and satellite communications. By controlling the relative phase (and amplitude) of signals at each antenna element, the array creates constructive interference in the desired direction and destructive interference elsewhere, electronically steering the beam in microseconds without mechanical motion.

The revolution in 5G is massive MIMO beamforming: instead of illuminating an entire cell with one broad beam (wasting power and creating interference), the base station forms narrow beams pointed directly at each user, dramatically increasing spectral efficiency and reducing interference. A 64-element array provides 18 dB of beamforming gain, extending mmWave range from impractical to viable.

Array Factor Equations

Uniform linear array:
AF(θ) = Σ_n=0^N-1 w_n e^{jnkd sinθ}
k = 2π/λ, d = element spacing

Beam steering:
φ_n = −nkd sinθ₀
Progressive phase for direction θ₀

Beamwidth:
BW_3dB ≈ 0.886λ/(Nd cosθ₀)
N=64, d=λ/2, broadside:
BW = 0.886/32 = 1.6°

Array gain:
G_array = N × G_element (linear)
= 10log(N) + G_el(dBi)
64 elements: +18 dB

Beamforming Architecture Comparison

Architecture	RF Chains	Beams	Flexibility	Application
Analog	1	1	Low	UE, legacy radar
Digital (full)	N	N	Maximum	AESA, mMIMO
Hybrid	K<N	K	Medium	5G gNB mmWave
Sub-array	M	M	Medium	Satellite, radar
Lens (Butler)	N	N (fixed)	Low	Switched beam

Common Questions

Frequently Asked Questions

Architectures?

Analog: RF phase shifters, 1 beam, simple, low power. Digital: per-element ADC/DAC, multi-beam, adaptive null, MIMO, highest cost/power. Hybrid: sub-arrays analog + digital combine, practical 5G mmWave trade-off. Choice: cost/power vs flexibility vs beam count.

Array factor?

AF = Σw_n×exp(jnkd sinθ). Steer: progressive phase φ_n = −nkd sinθ_0. BW = 0.886λ/(Nd cosθ). Gain = 10log(N). Pattern = element × AF. Grating lobes if d>λ/2. N=64: BW=1.6°, gain=+18dB. Amplitude taper for sidelobe control (costs 1-2 dB gain).

Massive MIMO?

64-256 elements, simultaneous per-user beams on same time-freq resource. Capacity scales with min(antennas, users). 64-el, 8 users: ~8x spectral efficiency. TDD: channel reciprocity for CSI. FDD: overhead scales with N, impractical above 32. 5G NR: up to 256 antenna ports.

Related Terms

← Beam Efficiency Beam Hopping →

Array Design

Request a Quote

Need phased array design, beamformer ICs, or antenna element integration? Contact our team.

Get in Touch