Few-Mode Optical Fiber

Updated 17 January 2026

Few-mode optical fiber is a guided wave medium that supports a limited number of spatial modes, enabling precise spatial division multiplexing and controlled modal propagation.
Engineered with tailored refractive-index profiles—such as ring-core, annulus-core, and multi-core designs—FMFs optimize modal dispersion, amplification, and intermodal spacing.
FMFs facilitate advanced applications in high-capacity telecommunications, microwave photonics, and quantum multiplexing by managing nonlinear effects and enabling efficient mode-selective coupling.

Few-mode optical fiber (FMF) is a class of guided wave transmission medium that supports a discrete, small number of transverse spatial modes—typically 2–30 linearly polarized (LP) groups—enabling multiplexed information channels, distributed photonic signal processing, and enhanced control over modal dispersion, nonlinearity, and noise. Whereas traditional single-mode fibers (SMFs) allow transmission only in the fundamental LP₀₁ mode and standard multimode fibers accommodate hundreds of modes often in the regime of strong coupling and complex modal mixing, FMFs occupy a regime in which modal propagation is engineered and can be precisely harnessed for advanced optical communication, microwave photonics, quantum multiplexing, spectroscopy, and high-power laser applications.

1. Fiber Structures, Mode Sets, and Refractive-Index Engineering

FMFs are realized using tailored step-index, graded-index, and ring-core refractive index profiles to support a prescribed finite set of modes at operational wavelengths. Exemplary structures include:

Ring-core step-index FMFs: Piecewise constant profile $n(r) = \{ n_1,\, 0\leq r<d_1;\ n_2,\, d_1\leq r\leq a_2;\ n_{cl},\, r>a_2 \}$ , with engineered index steps $\Delta n_{eff}>5\times10^{-4}$ to maximize intermodal spacing and suppress unwanted coupling, and parametric control over core radii and doping (e.g., $d_1=3\ \mu$ m, $a_2=10\ \mu$ m) (Garcia et al., 2019).
Annulus-core FMFs: Double-core (cylindrical or ring) geometries involving silica and GeO₂-doped layers supporting four or more LP modes, with tunable index contrast $\Delta$ and optional extra-annulus erbium doping for gain equalization (Gaur et al., 2016).
Multi-core FMFs: Hexagonal arrays of uncoupled few-mode cores in a common cladding, enabling scalable space-division multiplexed amplification with minimal inter-core crosstalk (Chen et al., 2017).
Graded-index FMFs: Parabolic or quasi-parabolic index profiles to produce equidistant modal group velocities, often used in high-power and soliton studies (Zhu et al., 2016).

Guided mode groups (LP $_{lm}$ ) are determined by generalized $V$ -number criteria, $V = (2\pi/\lambda)a\sqrt{n_{core}^2-n_{clad}^2}$ , with modal cutoffs controlled to achieve the desired number of spatial channels at the telecom wavelength ( $\lambda_0 \sim 1.55\ \mu$ m), or in visible/near-IR, as required for specific applications.

Each spatial mode $m$ is characterized by its propagation constant $\Delta n_{eff}>5\times10^{-4}$ 0, group delay per unit length $\Delta n_{eff}>5\times10^{-4}$ 1, and chromatic dispersion $\Delta n_{eff}>5\times10^{-4}$ 2. Key reported metrics include:

Modal group delays (ps/km): For a 7-mode ring-core FMF at 1550 nm, $\Delta n_{eff}>5\times10^{-4}$ 3 {3489.1, 8182.3, 13022.3, ...} ps/km for LP $\Delta n_{eff}>5\times10^{-4}$ 4 LP $\Delta n_{eff}>5\times10^{-4}$ 5 ... LP $\Delta n_{eff}>5\times10^{-4}$ 6 (Garcia et al., 2019).
Dispersion parameters (ps/(km·nm)): $\Delta n_{eff}>5\times10^{-4}$ 7 19–29 ps/(km·nm) for typical LP modes (Garcia et al., 2019), supporting accurate true-time-delay line construction via tailored mode-to-mode differences and spatially distributed long-period gratings (LPGs).
Differential Modal Group Delay (DMGD): In low-DMG designs, DMGD remains $\Delta n_{eff}>5\times10^{-4}$ 82.2 dB over the C-band for four-mode spans, supporting robust SDM transmission (Gaur et al., 2016).

Control over $\Delta n_{eff}>5\times10^{-4}$ 9 and $d_1=3\ \mu$ 0—via core/ring geometry, doping, and LPG placement—enables distributed delay taps for RF photonics, tuneable delay lines, and stable modal multiplexing.

3. Amplification, Loss Equalization, and Mode-Selective Couplers

FMF amplifiers are essential for extending practical SDM links and supporting high-capacity transmission. Notable approaches include:

Annulus-core and cladding-pumped FM-EDFAs: Uniform and extra-annulus erbium doping yields nearly equal mode gain $d_1=3\ \mu$ 10.5 dB for up to four spatial channels, with overall DMG $d_1=3\ \mu$ 22.2 dB over the C-band, $d_1=3\ \mu$ 320–22 dB modal gain at 250 mW pump, and noise figure $d_1=3\ \mu$ 43.3 dB (Gaur et al., 2016, Chen et al., 2017).
Integrated mode-selective couplers (MSCs): All-fiber phase-matched tapers and silicon photonic grating-based multiplexers efficiently launch/pick LP modes with insertion loss $d_1=3\ \mu$ 52–5 dB and cross-talk $d_1=3\ \mu$ 6–20 dB (Wang et al., 2020, Zhou et al., 2023).
Cladding-pumped multi-core FM-EDFA: Simultaneous amplification of 18 spatial channels (6 cores × 3 modes) with $d_1=3\ \mu$ 720 dBm output per core, noise figure $d_1=3\ \mu$ 87 dB, and negligible inter-core/pump depletion crosstalk (Chen et al., 2017).
Advanced integrated mode MUX: Compact silicon devices with Mikaelian lens mode size converters and MMGCs achieve $d_1=3\ \mu$ 90.25 dB mode-conversion loss and $a_2=10\ \mu$ 0–30 dB cross-talk, enabling dense SDM front-end fabrication (Zhou et al., 2023).

Equalization of modal gain and minimization of DMGD/crosstalk are critical for high-fidelity telecom, quantum, and sensing networks.

FMFs support unique distributed photonic signal processing functionalities:

Sampled true-time-delay lines (TTDL): Engineering $a_2=10\ \mu$ 1 and $a_2=10\ \mu$ 2, a sequence of LPGs and mode conversions enables RF signal replica generation with accurately controlled delay taps $a_2=10\ \mu$ 3, operated as multi-tap delay filters (Garcia et al., 2019). $a_2=10\ \mu$ 4 tunability: 50–150 ps/km over 20 nm wavelength span.
Frequency-domain monitoring: CAZAC-sequence-based probing of MIMO FMF channels achieves in-service, real-time extraction of mode-dependent loss (MDL) and DMGD with $a_2=10\ \mu$ 50.3 dB and $a_2=10\ \mu$ 60.3 ps error, scalable to arbitrarily many spatial dimensions without added hardware (Fan et al., 30 May 2025).
Modal control/switching: Piezo-mechanical "fiber piano" and polarization optics can dynamically excite, convert, and separate individual modes (LP $a_2=10\ \mu$ 7, LP $a_2=10\ \mu$ 8, LP $a_2=10\ \mu$ 9) with $\Delta$ 085% purity and insertion loss of 30–60%, extending mode shaping down to the single-photon level (Wu et al., 2024).

These capabilities extend FMF utility into microwave photonics, optical beamforming, adaptive optics compensation, and quantum signal routing.

5. Nonlinear Effects: Raman, Brillouin, Kerr, and Soliton Dynamics

Nonlinear propagation in FMFs is governed by multimode GNLSE frameworks, with key nonlinear phenomena including:

Intermodal Raman scattering: Leads to power-dependent depletion of LP $\Delta$ 1 (fundamental) mode; transfer to HOMs reaches $\Delta$ 280% in high-power visible regime, imposing upper limits for beam delivery and modal cross-talk control (Gemechu et al., 16 Oct 2025).
Spontaneous Brillouin scattering: Forward/backward SBS resolved for intra-/inter-modal interactions via heterodyne detection; backward SBS gain up to 160 W $\Delta$ 3km $\Delta$ 4 for LP $\Delta$ 5, forward SBS in MHz–GHz range relevant for optomechanics/quantum phononics (Kikuchi et al., 10 Jan 2026).
Multimode solitons and nonlinear mixing: Energy-volume scaling transition between single-mode and bulk solitons; at high energy, FMF solitons expand spatially rather than temporally, resulting in robust high-energy pulse delivery (Zhu et al., 2016). Coupled nonlinear envelopes exhibit strong intermodal interference and XPM/FWM in time-resolved experiments (Dacha et al., 2020).

Control or mitigation of modal nonlinearities is required for high-power transmission, advanced modulation, and quantum networking.

FMFs provide the basis for joint quantum-classical multiplexing in capacity-constrained networks:

Mode-wavelength dual multiplexing in quantum key distribution (QKD): Weakly-coupled FMFs with MSCs achieve simultaneous transmission of 100 Gb/s classical data and real-time secure QKD over 86 km with 86% Raman reduction compared to SMF at identical launch power (Wang et al., 2020).
Modal division multiplexing of quantum/classical signals: MPLC-based systems with 8 km graded-index FMFs support 15 channels, achieving $\Delta$ 610% quantum cross-talk and SNR $\Delta$ 710 dB with 20 nW classical power per channel (Zia et al., 2024).
Quantum SDM with time-bin and phase encoding: Detector dead-time gating is exploited to suppress modal cross-talk, supporting multimode single-photon transmission at >1 Mqubit/s over 8 km (Zitelli, 5 Sep 2025).

Modal isolation, low loss ( $\Delta$ 8– $\Delta$ 9 dB/km), and advanced MDM technologies underpin SDM-enhanced secure networks.

FMFs have significant advantages and distinct limitations in free-space optics and spectroscopy:

Coupling efficiency and turbulence/jitter tolerance: FMFs admit $_{lm}$ 0– $_{lm}$ 1 dB link gain and lower BER compared to SMFs in FSO links under turbulence/jitter, leveraging multi-mode energy capture and scale-adapted LG mode overlaps (Fan et al., 2020).
AO-assisted spectroscopy: Instruments such as NIRPS use FMFs to balance coupling efficiency ( $_{lm}$ 250%) under poor seeing, compact spectrograph size, and relaxed AO requirements, with modal noise mitigated by continuous tip–tilt scanning and spectrograph scrambling (Blind et al., 2017).
Modal noise mechanisms and mitigation: Modal phase drift and excitation in FMFs can lead to substantial speckle-induced RV instability; active modal mixing and AO-driven strategies are necessary for sub-m/s precision.

FMFs thus mediate the trade-off between throughput, instrument stability, and practical fiber interface constraints in advanced optical measurement systems.

FMFs—via precise refractive-index engineering, tailored amplification/coupling methods, and sophisticated modal control—enable SDM networks, distributed photonic processing, nonlinear-optical control, and quantum-classical multiplexing well beyond the limitations of conventional fiber architectures. Their current deployment spans high-throughput telecom, field-deployed OAM fibers with 400 Tbit/s aggregate capacity (Liu et al., 2024), and emerging quantum optical networks. Ongoing research is focused on minimizing nonlinear cross-talk, maximizing modal equalization and isolation, and developing integrated photonic multiplexers scalable to kilo-channel regimes.