A 5G base station uses a massive-MIMO panel — dozens to hundreds of small antenna elements — to combine their signals in phase toward each device. This "beamforming" concentrates energy into a narrow lobe instead of spraying it in every direction, which is what lets a single mast serve many phones at once with a strong link.
Higher carrier frequencies pack more available bandwidth (more capacity per Hertz of spectrum), but their short wavelength means the same physical antenna panel produces a much narrower beam and the signal attenuates far faster with distance — the classic 5G low-band vs. mmWave trade-off.
λ = c / f
θ_beam ≈ (70·λ / D) degrees (D = 0.4 m panel aperture)
FSPL(dB) = 20·log10(d_km) + 20·log10(f_MHz) + 32.44
SNR(dB) = P_tx + G_beam − FSPL − (−174 + 10·log10(B_Hz) + NF)
C = B · log2(1 + SNR) (Shannon capacity per device)
- Carrier frequency — sets wavelength λ, beamwidth and how quickly the link budget degrades with range.
- Channel bandwidth — the B term in Shannon's law; wider channels (common at mmWave) directly raise throughput.
- Connected devices — how many independent beams the massive-MIMO array must steer simultaneously.
- Beamforming ON/OFF — toggles the ~10·log10(N) array gain from coherently combining N antenna elements versus broadcasting omnidirectionally.
This is why real 5G deployments pair long-range low/mid-band macro cells with dense mmWave small cells in stadiums and city centres — and why 6G research is pushing into sub-terahertz bands for even more bandwidth at even shorter range.