How do beam ports work in a Butler matrix?

A Butler matrix is an N×N passive network (N = power of 2) with N beam ports and N array ports. Each beam port produces a unique progressive phase shift across the array ports. For an N×N Butler matrix, beam port m (m = 1 to N) produces phase gradient: Δφ_m = (2m - 1 - N)π/N between adjacent array ports. Example: 4×4 Butler matrix has 4 beam ports producing phase gradients of -135°, -45°, +45°, +135° (beam directions at -48.6°, -14.5°, +14.5°, +48.6° for λ/2 spacing). Each beam port sees 50Ω input impedance regardless of which other ports are terminated. Unused beam ports must be terminated in matched loads to prevent reflections and pattern distortion. The network consists of log₂(N) stages of hybrid couplers and fixed phase shifters.

What is the difference between beam ports and array ports?

In any beamforming network, two sets of ports exist: Beam ports (input side): each beam port corresponds to one beam direction. Exciting beam port 1 steers the beam to direction θ₁, beam port 2 to θ₂, etc. Connected to the transmitter/receiver. The number of beam ports equals the number of available beam directions. Array ports (output side): each array port connects to one antenna element. All array ports are excited simultaneously when any beam port is driven. The phase/amplitude relationship across array ports determines the beam direction. The BFN provides a fixed mapping between beam ports and array ports. This is in contrast to digital/analog phased arrays where the mapping is programmable. The advantage of passive BFNs: no active components, no power consumption, no control latency. Disadvantage: beam directions are fixed at design time and cannot be changed.

Can multiple beam ports be excited simultaneously?

Yes, exciting multiple beam ports simultaneously creates multiple simultaneous beams, but with power sharing. For K simultaneous beams from a network with total input power P: each beam receives P/K power, reducing the per-beam EIRP by 10log₁₀(K) dB. For 2 beams: -3 dB each. For 4 beams: -6 dB each. The beams are orthogonal (non-interfering) in a Butler matrix because the DFT-based phase relationships ensure zero coupling between beam ports. In a Rotman lens, adjacent beams have some sidelobe interaction. Multi-beam operation is used in: SATCOM (serving multiple ground terminals simultaneously), radar (searching multiple directions), 5G NR (multi-user MIMO via analog beams). The practical limit is typically 4-8 simultaneous beams before the per-beam gain reduction becomes unacceptable for the link budget.

Beamforming Networks

Beam Port

/beem port/

The input/output terminal of a beamforming network (BFN) corresponding to a specific beam direction. In a Butler matrix, beam port m produces progressive phase Δφ_m = (2m−1−N)π/N across N array ports. In a Rotman lens, each beam port sits on the focal arc at a position mapping to the desired beam angle. Exciting multiple ports simultaneously creates simultaneous beams with −10log₁₀(K) dB power sharing per beam. Passive BFN beam ports require matched load termination on unused ports to prevent pattern distortion.

Butler: N = 2^k

Multi-beam: −3 dB/beam (K=2)

Type: Passive, fixed

Understanding Beam Ports

A beamforming network translates the problem of beam steering from "adjust N phase shifters" to "select the right beam port." Each beam port is pre-wired through the network's hybrid couplers and phase shifters (Butler matrix) or through the lens geometry (Rotman lens) to produce the correct phase gradient across the array elements for its designated beam direction.

The elegance of passive BFNs is that beam switching requires no computation and no active control: simply route the RF signal to the desired beam port. This makes the switching time limited only by the RF switch speed (nanoseconds), far faster than any digital beamformer. The tradeoff is that beam directions are fixed at fabrication and cannot be changed.

Butler Matrix Beam Port Phase Gradients

4×4 Butler Matrix:
Port 1: Δφ = −135° → θ = −48.6°
Port 2: Δφ = −45° → θ = −14.5°
Port 3: Δφ = +45° → θ = +14.5°
Port 4: Δφ = +135° → θ = +48.6°

General Formula:
Δφ_m = (2m − 1 − N)π/N
θ_m = arcsin(Δφ_mλ/(2πd))

Multi-beam Power Sharing:
K beams: EIRP_per-beam = EIRP_total − 10log₁₀(K)
2 beams: −3 dB each
4 beams: −6 dB each

BFN Beam Port Comparison

BFN Type	Beam Ports	Bandwidth	Advantage
Butler Matrix	N = 2^k	10–20%	Compact, low loss
Rotman Lens	Arbitrary	40–60%+	Wideband, true time delay
Blass Matrix	Arbitrary	Narrowband	Flexible beam placement
Nolen Matrix	M ≤ N	10–15%	Fewer beams than elements

Common Questions

Frequently Asked Questions

How do beam ports work?

Each port pre-wired through hybrids/phase shifters to produce unique phase gradient. 4×4 Butler: ports produce −135°, −45°, +45°, +135° gradients. Rotman: physical position on focal arc determines beam angle.

Beam ports vs. array ports?

Beam ports: input side, each = one beam direction, connected to Tx/Rx. Array ports: output side, each = one antenna element. BFN maps beam → array with fixed phase/amplitude. Passive, no control latency, no power consumption.

Simultaneous beams?

Multiple beam ports excited = multiple beams. Power sharing: −10log(K) dB per beam. Butler: orthogonal (zero inter-port coupling). Practical limit: 4 to 8 beams before gain reduction exceeds link budget margin.

Related Terms

← Beam Management Beam Recovery →

Beamforming Networks

Precision Waveguide Components

RF Essentials provides precision terminations and custom waveguide assemblies for Butler matrices, Rotman lenses, and multi-beam antenna feed networks.

Request a Quote