The Yagi-Uda Antenna: Core Principles and Performance Metrics

The Yagi-Uda antenna, invented by Hidetsugu Yagi and Shintaro Uda in 1926, remains one of the most elegant and effective directional antenna designs. Its structure is deceptively simple: a driven element (typically a half-wave dipole or folded dipole) is flanked by a slightly longer reflector element behind it and one or more shorter director elements in front. These parasitic elements are not directly connected to the feedline; instead, they couple electromagnetically to the driven element and reradiate energy with carefully controlled phase shifts. The result is constructive interference in the forward direction and destructive interference toward the rear, producing a clean, unidirectional radiation pattern with high gain.

A well-optimized Yagi delivers gain values between 8 and 15 dBi, with half-power beamwidths (HPBW) ranging from approximately 55 degrees for a three-element design down to 25 degrees or less for arrays with ten or more elements. The front-to-back ratio—a measure of the antenna's ability to reject signals from behind—typically falls between 20 and 30 dB. For a comprehensive treatment of classical design theory, the resource at Antenna Theory remains an authoritative reference. The key design parameters are tightly interdependent: the number of directors primarily increases forward gain and narrows beamwidth, but each additional director also reduces impedance bandwidth and can elevate side-lobe levels if not carefully optimized. The reflector length and its spacing relative to the driven element primarily control the front-to-back ratio, while the driven element determines feed-point impedance and overall bandwidth. Understanding these trade-offs is essential for any antenna design intended for multi-user MIMO (MU-MIMO) systems, where pattern purity and bandwidth are as critical as raw gain.

Beyond the classical rod-based Yagi, modern implementations use printed circuit board (PCB) technology to create "Quasi-Yagi" antennas. These printed designs replace parasitic rods with etched copper strips on a dielectric substrate, offering lower cost, repeatable performance, and the ability to integrate baluns and matching networks directly on the board. Printed Yagis typically achieve 6–10 dBi gain with bandwidths of 10–20%, making them attractive for mass-produced access points and customer premises equipment.

Multi-User MIMO: The Spatial Multiplexing Imperative

Multi-User MIMO is a foundational technology in modern wireless standards, including 4G LTE-Advanced Pro, 5G NR, and Wi-Fi 6 (802.11ax) and Wi-Fi 7 (802.11be). Unlike single-user MIMO, which directs multiple spatial streams to a single device, MU-MIMO enables a base station or access point to communicate with several distinct users simultaneously on the same time-frequency resource. This is achieved by forming spatially separated beams that isolate individual user signals, effectively multiplying channel capacity without requiring additional spectrum. An overview of the fundamental concepts can be found on Wikipedia's Multi-user MIMO page.

The practical realization of MU-MIMO gain depends critically on the base station's ability to acquire accurate downlink channel state information (CSI) for each user. This is typically obtained through uplink sounding reference signals or explicit feedback. The antenna system applies precoding weights—calculated from the channel estimates—to create beams with low cross-correlation. In dense urban small-cell deployments, these beams must often cover narrow angular sectors of 15 to 30 degrees while maintaining high gain to close the link budget. This precision is precisely the performance regime where Yagi antennas have historically excelled, making them a compelling alternative to more complex phased-array panels. MU-MIMO systems also impose stringent requirements on antenna port isolation and correlation: an envelope correlation coefficient (ECC) below 0.3 is generally required to maintain spatial multiplexing gain.

The Rationale for Yagi Antennas in MU-MIMO Systems

While fully digital phased arrays with dozens of patch elements represent the mainstream approach for macro-cell base stations, Yagi-based architectures offer distinct advantages in specific deployment contexts. A single Yagi with four or five elements can deliver 12 dBi of gain using little more than a few metal rods and a simple balun. A patch array would require at least four elements, a corporate feed network, and a ground plane to approach the same performance, resulting in higher cost, greater weight, and more complex manufacturing. For cost-sensitive small-cell access points, customer premises equipment, and fixed wireless access terminals, the Yagi's mechanical simplicity translates directly to lower bill-of-materials cost and higher reliability.

Beyond cost, the Yagi's inherent directivity provides passive interference rejection. Signals arriving from outside the main beam are attenuated by 20 dB or more without any active signal processing. This passive spatial filtering reduces the dynamic range requirements on analog-to-digital converters and lessens the computational burden on the digital precoder, allowing the system to allocate more processing resources to user scheduling and data throughput. In unlicensed bands where co-channel interference from neighboring networks is a persistent challenge, this built-in interference suppression is a significant operational advantage. Furthermore, Yagi antennas exhibit very low ohmic losses because they consist of simple conductors rather than complex feed networks, resulting in higher radiation efficiency (typically >90%) compared with patch arrays that suffer dielectric and conductor losses in the feed lines.

The obvious limitation is that a single Yagi produces only one beam at a time. To serve multiple users simultaneously, the antenna system must combine multiple Yagis in a switched or phased configuration, or modify the parasitic structure to enable electronic beam steering. These architectural approaches are explored in detail in the following sections. Designers must also contend with the Yagi's inherently narrow bandwidth compared with broadband dipoles or log-periodic antennas, requiring careful attention to impedance and pattern stability across the operating band.

Critical Design Parameters for MU-MIMO Yagi Arrays

Director Count and the Gain-Bandwidth Trade-off

Each additional director element increases the traveling-wave component along the boom, raising forward gain and narrowing HPBW. A carefully optimized six-element Yagi achieves approximately 13 dBi gain with a 35-degree beamwidth, while a ten-element design can reach 14.5 dBi with a beamwidth near 28 degrees. Beyond ten to twelve elements, the gain increase per director diminishes, a phenomenon known as end-fire saturation, while the impedance bandwidth contracts to just 2 to 3 percent of the center frequency. For MU-MIMO applications, a practical target is 8 to 12 dBi gain with a fractional bandwidth of at least 5 percent. This can be realized with thick cylindrical elements that lower the Q factor, or by employing a stepped-diameter profile that broadens the resonance. Printed Quasi-Yagi designs on low-loss dielectric substrates can achieve 15 percent bandwidth while maintaining a clean forward pattern, making them well-suited for 5G sub-6 GHz bands like n78 (3.3–3.8 GHz).

Inter-Element Spacing and Mutual Coupling

When multiple Yagi antennas are placed in a side-by-side array to serve independent beams, the center-to-center spacing becomes a critical parameter. Insufficient spacing increases mutual coupling between adjacent elements, altering input impedance, distorting radiation patterns, and elevating the correlation coefficient between antenna ports. High port correlation directly reduces the spatial multiplexing gain that MU-MIMO depends on. A minimum spacing of 0.7 wavelengths is commonly adopted to keep coupling below -20 dB. In tighter installations, decoupling networks, neutralization lines, or electromagnetic bandgap (EBG) structures can be inserted between the elements to restore isolation. Electrostatic shielding in the form of vertical metal fences between Yagi booms has also proven effective in prototype arrays. Simulation tools like CST Microwave Studio or HFSS are essential for optimizing these spacing and decoupling structures.

Broadband Impedance Matching Across the Operating Band

The driven element is the focal point of bandwidth engineering in any Yagi design. A folded dipole naturally presents an impedance near 300 ohms, which can be transformed to 75 ohms using a 4:1 balun, but the rapid impedance variation away from resonance limits usable bandwidth. To extend the match across wide 5G channels, designers employ several techniques. A double-driven element configuration—two folded dipoles placed in close proximity—produces a dual-resonant response that merges into a broad impedance match. Triangular or bow-tie shaped dipoles reduce Q and widen the bandwidth without compromising the radiation pattern. Integrated matching stubs, formed as part of the balun structure, can cancel the reactive component of the input impedance across a 10 to 15 percent frequency range. For 5G NR band n78, a VSWR below 1.5 across the full 500 MHz range is the design target, ensuring maximum power transfer and minimal stress on the power amplifier.

Beamwidth Shaping and Side-Lobe Suppression

In an MU-MIMO system, every degree of beamwidth directly impacts user isolation. A beam that is too broad will inadvertently cover two users, causing inter-user interference that degrades SINR for both. A beam that is too narrow may lose track of a user during small-scale movements or due to channel variations. The designer must tailor the HPBW to the expected angular separation of users in the intended deployment scenario, typically 20 to 35 degrees for small-cell urban environments. Side lobes require equally careful attention: a prominent first side lobe can inject interference into an adjacent beam, nullifying the benefit of beamforming. Tapering the director lengths—gradually shortening successive directors along the boom, with the final director being the shortest—reduces side-lobe levels by 3 to 5 dB. Adding a second reflector element behind the primary reflector further suppresses rear lobes and improves the front-to-back ratio. Full-wave electromagnetic simulation is essential for optimizing these interdependent parameters for each specific array configuration. A common design methodology uses genetic algorithms to simultaneously optimize element lengths, spacings, and diameters for multiple objectives including gain, F/B ratio, and bandwidth.

Architectures for Simultaneous Multi-User Support

Sectorized Multi-Yagi Arrays

The most straightforward approach to serving multiple users with Yagi antennas is to deploy several fixed elements, each covering a distinct angular sector. Four Yagis with 30-degree beamwidths arranged on a common mast can collectively cover a 120-degree horizontal span. A baseband processor selects which antenna port serves which user based on channel quality measurements from each sector. This switched-beam architecture is hardware-simple, requiring no phase shifters or variable reactances, and offers robust reliability. The trade-off is limited angular resolution within each sector: users cannot be individually tracked unless the Yagis are mechanically repositioned, which is impractical for mobile subscribers. The approach works well for fixed wireless access scenarios where user locations are static, and serves as a baseline for evaluating more advanced architectures. To improve angular resolution, some designs use overlapping sectors with narrower beamwidths and a beam selection algorithm that chooses the best combination for each user.

Electronically Steerable Passive Array Radiators (ESPAR)

The Electronically Steerable Passive Array Radiator (ESPAR) concept represents a significant advancement in Yagi-based beamforming. An ESPAR antenna consists of a single driven element—typically a monopole or dipole—surrounded by a ring of parasitic elements, each loaded with a variable reactance such as a PIN diode or varactor diode. By adjusting the DC bias on each parasitic element, the antenna can steer the main beam across the azimuth plane, create multiple simultaneous beams, or place nulls in specific directions. An ESPAR with six to twelve parasitic elements can achieve up to 10 dBi gain with a beam scanning range of 360 degrees. Critically, only one RF chain is required for the entire assembly, dramatically reducing cost and power consumption compared with fully digital phased arrays. Recent research has demonstrated MU-MIMO operation with a single seven-element ESPAR using time-domain beam sharing or by synthesizing multiple orthogonal pattern states through baseband precoding. In the latter approach, the ESPAR is switched rapidly between several beam states within one OFDM symbol period, effectively creating virtual spatial streams. This makes the ESPAR architecture a strong candidate for small-cell base stations and enterprise Wi-Fi access points, combining the high per-element gain of a Yagi with the beam agility required for dynamic multi-user scheduling. The control software for ESPAR typically uses a lookup table of bias voltages precomputed for each desired beam direction, enabling rapid switching without complex real-time optimization.

Phased Arrays of Yagi Elements

For higher-performance systems, a linear or planar array of Yagi antennas can be fed through a corporate network of phase shifters and attenuators. This hybrid approach leverages the high per-element gain of individual Yagis to reduce the total number of active transmit/receive modules needed compared with a conventional patch array. The array factor provides electronic beam steering and null placement across the coverage sector, while the element factor—the Yagi's native pattern—suppresses back-lobe interference and reduces grating lobes. Prototype systems operating at 28 GHz for 5G fixed wireless access have demonstrated this concept using printed quasi-Yagi elements on low-loss substrates, connected to a Butler matrix feed network. The Butler matrix generates multiple orthogonal beams from a single passive network, enabling simultaneous multi-user spatial multiplexing without the computational overhead of digital beamforming. A 4×4 Butler matrix feeding four Yagi elements produces four beams with -3 dB crossover points and isolation exceeding 20 dB between input ports.

Butler Matrix Multi-Beam Systems

A Butler matrix is a passive feed network that, when connected to an array of antenna elements, produces a set of spatially orthogonal beams from multiple input ports. A 4×4 Butler matrix feeding a linear array of four Yagi antennas generates four distinct beams, each pointing in a different direction and exhibiting minimal pattern overlap. This inherent orthogonality simplifies the MU-MIMO precoding problem significantly, as the beams are already spatially separated and require less aggressive digital interference cancellation. The approach has been adopted in several commercial Wi-Fi 6 access points that use internal printed Yagi arrays with integrated Butler matrix feeds, providing reliable multi-user throughput in high-density deployment scenarios such as auditoriums and conference centers. The main limitation is that the Butler matrix is fixed once manufactured; beam directions cannot be adapted in real time to changing user locations. However, hybrid approaches combine a Butler matrix with a small number of RF chains and digital beamforming at the sub-array level to achieve both angular coverage and adaptability.

Interference Management and User Separation Strategies

The ability of a MU-MIMO system to distinguish users depends on the spatial correlation between their channel signatures. High front-to-back ratio in a Yagi antenna—typically 20 to 30 dB—is a natural advantage, as it strongly attenuates signals from directions behind the main beam. When multiple Yagis are used simultaneously, the angular separation between their boresight directions must be sufficient to maintain a low envelope correlation coefficient. A rule of thumb is that beam peaks should be separated by at least one HPBW to keep correlation below 0.3. The achievable spatial multiplexing gain can be approximated by the number of simultaneous beams that can be formed with a given angular spacing, subject to the condition that each user's SINR remains above the modulation and coding scheme threshold.

In dense urban small-cell deployments, user traffic is typically non-uniformly distributed. Adaptive beamforming using an ESPAR or phased Yagi array enables the system to steer beams toward high-density user clusters while placing nulls toward sources of co-channel interference. This pattern agility is particularly valuable in unlicensed spectrum bands such as the 5 GHz and 6 GHz bands, where interference from neighboring access points and radar systems is unpredictable. The ability to reshape the radiation pattern in real time, adapting to the instantaneous interference environment, provides a measurable throughput improvement over fixed-beam sectorized approaches, especially in scenarios with a high density of overlapping basic service sets. Advanced receivers also employ successive interference cancellation (SIC) to further separate users that are not perfectly isolated in the spatial domain.

Bandwidth Engineering for Modern Wideband Standards

5G NR operating in sub-6 GHz bands such as n77 (3.3 to 4.2 GHz) and n78 (3.3 to 3.8 GHz) requires channel bandwidths of up to 100 MHz. A conventional Yagi antenna designed for a single narrow resonance cannot cover this range without unacceptable variation in gain and input impedance. Several established techniques are used to broaden the operational bandwidth of Yagi antennas for these applications:

  • Log-periodic scaling: The lengths and spacings of the director elements are scaled in a geometric progression, so that different sections of the array resonate at different frequencies. The active region shifts along the boom as frequency changes, producing a continuous broadband response with relatively constant gain and pattern characteristics. This approach can achieve bandwidths exceeding 3:1.
  • Thick and stepped-profile elements: Using cylindrical conductors with a large diameter relative to wavelength reduces the Q factor of each element, broadening the impedance bandwidth. Stepping the diameter along the element profile introduces multiple resonances that can be merged into a single wideband match. Diameters up to 0.05λ are common in broadband designs.
  • Parasitic stub loading: Shorted or open-circuit transmission line stubs attached to the driven element introduce additional resonances at controlled frequencies. With careful design, these resonances can be positioned to overlap with the main resonance, effectively widening the impedance bandwidth by 50 to 100 percent.
  • Substrate-integrated waveguide (SIW) feeds: For millimeter-wave frequencies, printed quasi-Yagi antennas fed by SIW structures can achieve fractional bandwidths exceeding 20% while maintaining high radiation efficiency and pattern stability across the band. SIW feeds also provide low-loss integration with mmWave transceivers.

The engineering target for most commercial wideband Yagi designs is a VSWR below 2:1 with a gain variation of less than 2 dB across the entire licensed band. Achieving these specifications concurrently with the directivity and pattern purity demands of MU-MIMO is a challenging iterative design process that requires extensive full-wave electromagnetic simulation and optimization. Modern simulation tools allow parameter sweeps over millions of design points to find optimal geometries.

Case Study: A 3.5 GHz ESPAR Array for 5G Small Cells

To illustrate these design principles in a concrete application, consider a small-cell base station intended for the n78 band centered at 3.5 GHz. An ESPAR antenna is constructed using a central half-wave driven element surrounded by six parasitic monopole elements arranged in a circular configuration with a radius of 0.35 wavelengths. Each parasitic element is terminated through a series varactor diode (e.g., Skyworks SMV1405) that provides a capacitance tuning range from 0.3 pF to 3.0 pF. By changing the bias voltage on each varactor (0–20 V), the parasitic elements can be configured as either reflectors or directors, steering the beam in 10-degree increments across the full 360-degree azimuth plane. Only a single RF transceiver chain is needed to feed the driven element, eliminating the need for an RF switch matrix and multiple downconverters.

The measured performance of this ESPAR prototype shows 9.5 dBi boresight gain with a front-to-back ratio exceeding 25 dB across the operating band. The 10 dB return-loss bandwidth covers the full 3.3 to 3.8 GHz range. In a field trial with four user equipment devices placed at random positions around the base station, the system employed a round-robin scheduling scheme in which the beam was sequentially steered toward each user during its assigned time slot. With OFDM symbols of 66.7 μs, the beam could be switched in less than 1 μs using fast PIN diode switches to change bias states, allowing beam updates every symbol. The measured throughput per user reached 65 percent of the theoretical single-user MIMO capacity, while the system simultaneously supported four users on the same time-frequency resources. This represents a net spectral efficiency gain of 2.6 times over a single-user baseline, clearly demonstrating the MU-MIMO viability of the ESPAR architecture. Future development will use simultaneous multi-beam pattern synthesis within a single OFDM symbol to further increase spectral efficiency, exploiting the fact that a single ESPAR can produce multiple orthogonal patterns by superimposing different bias states within the symbol duration.

Emerging Directions and Future Evolution

As the wireless industry moves toward sixth-generation systems, the Yagi antenna concept is being scaled to new frequency regimes and integrated with advanced materials. On-chip Yagi-Uda antennas fabricated in CMOS and SiGe processes are already demonstrated for 60 GHz WiGig applications, and ongoing research extends these designs to D-band (110 to 170 GHz) for next-generation fixed wireless access and backhaul links. The use of tunable dielectric materials such as liquid crystals and graphene-based varactors may enable ultra-fast reconfiguration of parasitic loads without the ohmic losses and linearity limitations of conventional semiconductor varactors. For instance, liquid crystal-based phase shifters can provide continuous phase control at millimeter-wave frequencies with insertion loss below 1 dB.

The concept of Reconfigurable Intelligent Surfaces (RIS) shares deep operational similarities with large parasitic Yagi arrays. An RIS panel can be viewed as a two-dimensional array of parasitic scatterers whose reflection phase is individually controllable, forming beams and nulls through passive beamforming. Hybrid architectures combining an active Yagi feed with a passive RIS reflector could offer dynamic beam steering with exceptionally low power consumption, making them attractive for battery-constrained deployments such as IoT gateways and mobile hotspot routers. Research at the University of California, San Diego has demonstrated such an architecture at 28 GHz using a printed Yagi feed illuminating a 256-element RIS, achieving 15 dBi gain with electronic beam steering over a 60-degree sector.

On the system side, machine learning algorithms are being developed to determine the optimal parasitic load settings for a given set of user channels in real time, bypassing the need for exhaustive beam sweeping and reducing channel estimation overhead. These AI-driven control loops, trained on large datasets of channel realizations using deep reinforcement learning, promise to make ESPAR and related parasitic-array solutions truly practical for mass-market MU-MIMO access points. A recent paper in IEEE Transactions on Antennas and Propagation (IEEE TAP) demonstrated a neural network that predicted optimal varactor biases for a 7-element ESPAR with 95% accuracy, enabling beam switching in under 100 ns. The coming decade will see the humble Yagi evolve from a fixed directional antenna into an intelligent, adaptive spatial processing element that plays a central role in how multiple users share the wireless medium.

Summary

Designing Yagi antennas for Multi-User MIMO systems requires a disciplined integration of classical antenna theory with modern signal processing requirements. The key design challenge lies in simultaneously balancing gain, bandwidth, beamwidth, and pattern purity within the constraints of cost and mechanical simplicity. Fixed multi-Yagi sector arrays provide the most direct path to spatial multiplexing but lack the flexibility to track mobile users. ESPAR architectures unlock dynamic beamforming and interference management with a single RF chain, while phased arrays of Yagi elements offer the highest performance for premium deployments. Emerging control techniques based on machine learning and tunable materials will further extend the Yagi's capabilities, ensuring that this century-old antenna design continues to shape the way multiple users connect to wireless networks. For engineers beginning Yagi design for MU-MIMO, a practical starting point is to choose a target gain of 10–12 dBi with at least 5% bandwidth, use full-wave simulation to optimize element lengths and spacings, and validate mutual coupling in the intended array configuration. The rich body of existing literature, combined with modern computational tools, makes this a tractable and rewarding engineering challenge.