Movable Antenna-Enhanced Wireless Communications: General Architectures and Implementation Methods

• Massive and extremely large-scale antenna arrays are expected to expand, increasing the importance of practical antenna architectures.Source: §I · p. 1
• Existing MA studies primarily optimize antenna positions and characterize performance limits while assuming individually and globally adjustable elements.Source: §I · p. 1
• Individual global position adjustment can cause high energy consumption, tuning delays, and infrastructure changes, especially for massive arrays.Source: §I · p. 1
• The proposed architectures vary the moving unit, flexibility, scale, range, and functionality through sliding, turning, and folding movements.Source: §I · p. 2
• The article surveys candidate direct mechanical and equivalent electronic implementation methods, compares their advantages and disadvantages, and presents numerical examples.Source: §I · p. 2

• Element-level MAs move individual antennas within one-dimensional arrays, two-dimensional regions, three-dimensional spaces, or specially shaped regions.Source: A. Element-Level MAs · p. 2
• Three-dimensional orientation control changes yaw, roll, and pitch, affecting radiation patterns, polarization, beam tracking, and time-varying coverage.Source: A. Element-Level MAs · p. 2
• Combining three-dimensional movement and orientation gives each MA six degrees-of-freedom.Source: A. Element-Level MAs · p. 3
• Rapid simultaneous adjustment of multiple elements can require high implementation complexity and power consumption, while increasing channel-estimation and optimization overhead.Source: A. Element-Level MAs · p. 3
• Restricting movement and orientation to local regions reduces complexity, creating a trade-off between MA-assisted performance and antenna-movement complexity.Source: A. Element-Level MAs · p. 3

• Array-level MAs divide an antenna array into sub-arrays whose elements move or rotate together, with different sub-arrays controlled independently.Source: B. Array-Level MAs · p. 3
• Compared with element-level MAs, array-level architectures are more amenable to hardware implementation because they reduce independent control complexity and power consumption.Source: B. Array-Level MAs · p. 3
• Array-level designs can use existing compact antenna arrays, whereas element-level MAs require unusually large element spacing to accommodate movement.Source: B. Array-Level MAs · p. 3
• A sliding array uses movable sub-arrays that change location within a given region, providing less channel-reconfiguration flexibility but more efficient implementation.Source: 1) Sliding array · p. 3
• Sliding sub-arrays can avoid deep-fading or high-interference positions and adjust inter-sub-array distances for beam coverage and array aperture.Source: 1) Sliding array · p. 3

2) Rotatable array:

• A rotatable array rolls to adapt coverage, with its aspect ratio determining horizontal and vertical spatial resolution.Source: 2) Rotatable array · p. 4
• Mode H or Mode V is selected according to whether users are distributed primarily along the horizontal or vertical dimension.Source: 2) Rotatable array · p. 4

3) Turnable array:

• A turnable array rotates sub-arrays in yaw and pitch to serve distributed user clusters, with sub-array allocation optimized for users’ rate requirements.Source: 3) Turnable array · p. 4

4) Foldable array:

• A foldable array adaptively folds or unfolds sub-arrays into geometric shapes, including folding during windy conditions and unfolding afterward.Source: 4) Foldable array · p. 4

Hybrid architectures:

• Array-level MAs can be hybridized with conventional FPAs at cellular base stations, allowing movable sub-arrays to offload traffic or address coverage gaps.Source: 4) Foldable array · pp. 4–5

• Dual-scale MAs mount antennas on UAVs or terrestrial vehicles to combine platform mobility with local antenna position and orientation adjustment.Source: C. Dual-Scale MAs · p. 5
• Platform locations can first be optimized for strong path gains with user clusters, followed by antenna locations and orientations for coverage performance.Source: C. Dual-Scale MAs · p. 5
• Dual-scale MAs require joint location-and-orientation optimization, while wind can cause inaccurate positions and rotational angles for UAV-mounted MAs.Source: C. Dual-Scale MAs · p. 5

• External mechanical structures use actuators to convert control signals and energy into antenna torque or displacement.Candidate actuators include electric motors and MEMS; motors can provide one-dimensional rotation or two-dimensional movement.Source: 1) External Machinery · p. 5
• Efficiently moving or rotating multiple antennas simultaneously with low energy use and latency remains an unresolved implementation issue.Source: 1) External Machinery · p. 5
• Continuous electrowetting moves fluid-metal antennas through voltage-induced surface-tension changes and Marangoni forces.Source: 2) Liquid Fluidity · p. 5
• Deployable antenna arrays use internal transformable structures, often based on origami folding, to alter geometry for storage and performance reconfiguration.Source: 3) Internal Structure Transformation · p. 5

4) Application to Array-Level MAs:

• Worm-and-wheel gear sets driven by motors can produce movement or rotation of antenna arrays.Large antenna arrays may require large conveyors for movement.Source: 4) Application to Array-Level MAs · p. 5
• Equivalent antenna movement and rotation can be implemented electronically rather than through physical movement.Source: B. Electronically Driven Equivalent MAs · pp. 5–6
• Dual-mode patch antennas emulate movement by displacing the phase center through concurrent excitation of multiple modes.Exciting TM11 and TM21 modes can transform a uniformly spaced array into a nonuniform configuration without physical adjustment.Source: 1) Dual-Mode Antennas · p. 6

2) Dense Array Antennas:

• Dense array antennas approximate position movement by activating different elements within a densely deployed transmit/receive region.Reconfigurable devices such as PIN diodes can be controlled by FPGA coding sequences to switch the active antenna.Source: 2) Dense Array Antennas · p. 6
• Orientation-reconfigurable antennas emulate rotation by changing beam pointing direction without physically rotating the antenna.Source: 3) Orientation-Reconfigurable Antennas · p. 6
• Controlling four PIN-diode-loaded parasitic antennas generates beams in different directions, providing quick-response, high-accuracy equivalent rotation.Source: 3) Orientation-Reconfigurable Antennas · p. 7

Response speed

• Mechanically driven MAs typically respond more slowly because physical movement and collision constraints can limit element or array motion.The constraint is especially relevant to global antenna element or array movement.Source: 1) Speed of Reaction and Movement · p. 7
• Electrically driven MAs achieve faster response by replacing physical displacement with electronic control, but their limited tunable states can restrict broad movement or rotation.The tradeoff is between response speed and the breadth of achievable equivalent movement.Source: 1) Speed of Reaction and Movement · p. 7, 2) Range of Movement · p. 7

Movement range

• Mechanical systems can provide full-coverage movement through adjustable control ranges, while electrically driven systems may offer less coverage because they rely on multiple reconfigurable radiation states.Mechanical systems benefit from mature designs and straightforward operating principles.Source: 2) Range of Movement · p. 7

Accuracy and reliability

• Mechanical precision can degrade under temperature, humidity, vibration, and wear, whereas electrically driven MAs avoid physical wear and can improve long-term reliability and stability.Electrical accuracy remains sensitive to electronic component performance and temperature changes.Source: 3) Movement Resolution · p. 7

Cost and Maintenance

• Mechanically driven MAs generally have lower implementation cost but require maintenance and calibration because wear can increase operational costs and downtime.Their mature manufacturing processes and simpler installation reduce initial cost, while degradation creates recurring burdens.Source: 4) Cost and Maintenance · p. 7
• Electrically driven MAs are preferable for delay-sensitive applications, whereas mechanically driven MAs suit delay-tolerant applications such as IoT and smart homes.The choice also depends on budgetary constraints and technical requirements.Source: 4) Cost and Maintenance · p. 7

Performance Evaluation

• 220% higher sum rate is achieved by omnidirectional MA schemes over fixed-position antennas in the evaluated four-MA, four-user setup.The result uses zero-forcing precoding and jointly optimizes each MA’s 3D position and orientation through Bayesian optimization.Source: §IV · pp. 7–8
• Local antenna movement yields comparable sum-rate performance to global movement, supporting a more cost-effective MA implementation.For omnidirectional MAs, adding 3D rotation provides the same sum rate as 3D movement alone because rotation does not affect the radiation pattern.Source: §IV · p. 8
• A dual-mode two-element antenna array closely matches a conventional array with physical spacing d = dpc for dpc = 0.8λ and dpc = 1.2λ.Different excitation schemes alter the distance between the physical centers of the dual-mode antennas, enabling equivalent antenna movement without physical adjustment.Source: §IV · p. 8

V. CONCLUSION

• The proposed MA architectures exploit additional degrees of freedom, including 3D rotation and large-scale movement, for different communication requirements and scenarios.The architectures are designed to balance overall cost and communication performance.Source: §V · p. 8
• The article surveys mechanical and electronic controls as implementation methods for MA-aided communication systems.These methods correspond to direct physical movement and equivalent electronic control.Source: §V · p. 8
• Numerical results demonstrate that electronic control can achieve equivalent antenna movements without physically displacing the antennas.This result supports electronic implementation as an alternative to physical adjustment in the evaluated architectures.Source: §V · p. 8
• Other mechanical and electrical approaches may also be available because of continuing advances in reconfigurable antenna technologies.The article gives shape-memory polymers whose properties can change under external stimuli as one example.Source: §V · p. 8