Source-linked AI summary
Increasing Indoor Spectrum Sharing Capacity using Smart Reflect-Array
Xin Tan, Zhi Sun, Josep M. Jornet, Dimitris Pados
TL;DR
Indoor wireless networks face severe spectrum crowding from dense users and bandwidth demands. The paper proposes a passive, reconfigurable smart reflect-array that controls reflected-signal phases to enhance desired links and cancel interference without modifying user devices. Experiments, theory, and simulations report higher spectrum-spatial efficiency and transport capacity, while the experimental validation is limited to two simultaneous transmissions.
Problem
Dense indoor users, high bandwidth demands, and many coexisting services create a need for more efficient spectrum sharing.
Method
The paper uses a passive smart reflect-array in the wireless channel to control reflector phase shifts, enhance desired signals, and cancel interference.
Results
Experiments, theoretical deduction, and simulations demonstrate higher spectrum-spatial efficiency, while numerical analysis shows improved transport capacity with reflect-arrays.
Takeaways & Limitations
Smart reflect-arrays can support same-band simultaneous indoor transmissions using existing users’ hardware and software.
Takeaways & Limitations
The experimental validation considers two simultaneous transmissions; multi-user testbed implementation and real-time control remain future work.
Abstract
from arXiv · showhide
The radio frequency (RF) spectrum becomes overly crowded in some indoor environments due to the high density of users and bandwidth demands. To accommodate the tremendous wireless data demands, efficient spectrum-sharing approaches are highly desired. To this end, this paper introduces a new spectrum sharing solution for indoor environments based on the usage of a reconfigurable reflect-array in the middle of the wireless channel. By optimally controlling the phase shift of each element on the reflect-array, the useful signals for each transmission pair can be enhanced while the interferences can be canceled. As a result, multiple wireless users in the same room can access the same spectrum band at the same time without interfering each other. Hence, the network capacity can be dramatically increased. To prove the feasibility of the proposed solution, an experimental testbed is first developed and evaluated. Then, the effects of the reflect-array on transport capacity of the indoor wireless networks are investigated. Through experiments, theoretical deduction, and simulations, this paper demonstrates that significantly higher spectrum-spatial efficiency can be achieved by using the smart reflect-array without any modification of the hardware and software in the users' devices.
I. Introduction
Indoor wireless environments face severe spectrum crowding from dense users, high bandwidth demands, and many coexisting services. The paper proposes smart reflect-arrays to improve spectrum sharing without modifying user devices, and evaluates their effects experimentally, theoretically, and numerically.
- Indoor RF spectrum is especially crowded because conference halls and shopping malls combine tremendous user density, bandwidth demands, and coexisting wireless services.
- Cognitive radio can improve spectrum efficiency when licensed users are inactive but is ineffective when dense indoor users are highly active.
- Large antenna arrays required by adaptive beamforming for high spatial resolution are impractical for portable, wearable, and smaller devices.
- The proposed smart reflect-array enhances useful signals and cancels interference by optimally controlling reflector phase shifts, enabling simultaneous same-band transmissions.
- Experiments, theoretical deduction, and simulations demonstrate significantly higher spectrum-spatial efficiency without modifying users’ hardware or software.
II. System Architecture and Proof-of-Concept Experiment
The paper presents a smart reflect-array spectrum-sharing architecture and develops an experimental testbed to verify its feasibility.
- The system architecture and proof experiment are introduced to evaluate the proposed reflect-array-based spectrum-sharing solution.
A. System Architecture
The proposed architecture uses passive smart reflect-arrays to reshape wireless signal propagation between users. This spatial modulation enhances intended links, suppresses interference, and preserves compatibility with existing wireless devices and services.
- A. System Architecture: Smart reflect-arrays tune each reflector’s electromagnetic phase response to spatially modulate transmissions toward selected regions while limiting interference elsewhere.
- A. System Architecture: The resulting high-resolution spatial pattern creates private regions for individual wireless transmissions, allowing multiple users and services to share one frequency band simultaneously.
- A. System Architecture: Unlike MIMO, beamforming, and active relays, the reflect-array performs passive spatial control in the middle of the wireless transmissions.
- A. System Architecture: Because spatial diversity is created away from transmitters and receivers, users can retain their wireless devices and services without hardware or software changes.
B. Experimental Testbed Designed and Implemented
The proof-of-concept testbed uses electronically controlled reflectors and peripheral circuits to tune the smart reflect-array’s electromagnetic response. Its design targets indoor WiFi operation and independently controls 48 reflector units.
- B. Experimental Testbed Designed and Implemented: The testbed uses electronically controlled capacitors on microstrip patches to vary reflector resonant frequencies and expand the usable operating range.
- B. Experimental Testbed Designed and Implemented: The reflect-array operates at 2.4 GHz with 48 rectangular reflector units, each controlled by a bias voltage.
- B. Experimental Testbed Designed and Implemented: Four MEGA2560 microcontrollers generate 48 independent PWM controls, while RC low-pass filters convert them into 0–5 V bias voltages.
C. Experiment Results
The proof-of-concept setup uses a reflect-array between two transmitters and a receiver to control reflected signals. With optimized tuning, interference is reduced and the desired communication link achieves higher SINR.
- Experiment setup: The experiment places a receiver 0.6 m in front of the reflect-array and uses Tx1 as the signal source.Tx2 is deployed to interfere with communication between Tx1 and the receiver.
- Experiment setup: The proof-of-concept testbed includes a 6×8 = 48-reflector array with independently controlled phase shifts.Micro-controllers provide control voltages that tune the reflector varactors.
- Without reflect-array: Without the reflect-array, Tx1 and Tx2 produce nearly equal received signal strengths around -45 dBm because both transmission distances are 0.6 m.The comparison is made using the received signal strength from each transmitter.
- With reflect-array: With optimized reflector tuning, interference is canceled to -73 dBm and SINR increases to about 30 dB.This enables communication between Tx1 and the receiver while preventing interference from Tx2.
III. Increasing Transport Capacity by Smart Reflect-Array
The paper extends the reflect-array concept from two simultaneous transmissions to indoor networks with multiple user pairs. It presents the array as a means to serve more users within limited spectrum bands.
- Multiple-user capacity: The proposed system is extended to serve a number of 2L users forming L communication pairs.The reflect-array handles signals from sources as well as interference from other transmitting nodes.
- Multiple-user capacity: The paper motivates exploring spectrum-sharing capacity for many indoor users operating in limited spectrum bands.The stated goal is to improve RF spectrum utilization efficiency for wireless systems and services.
A. Analysis of the Effect of Reflect-array
The reflect-array changes the multipath channel by controlling the phase of reflected paths. The analysis models desired signals, interference, and noise for multiple communication pairs and relates performance to phase control.
- Channel influence: The received signal is the superposition of the direct path and signals reflected by the reflect-array patches.Consequently, the received signal depends strongly on the phases of multipath propagation.
- Single-link model: The single-link model represents a bandpass signal with BPSK symbols, carrier frequency f_c, multipath attenuation and phase, reflector-induced phase shifts, and noise.The 0-th path denotes the line-of-sight path, while φ_i is induced by reflector i.
- Multiple-user model: For L communication pairs, the received signal at receiver l includes the intended transmission and interference from the other L−1 transmitting nodes.The model accounts for reflected effects on both the desired signal and interference.
- Phase control: The phases of both desired signals and interference are controlled through the induced reflector phases φ_i.The first term in equation (2) represents the desired transmission, and the second represents interference sources.
- Performance criterion: Communication feasibility is evaluated using the signal-to-interference-plus-noise ratio, whose network effects are determined by the phase-control vector v_φ.The SINR is derived from the received signal containing the source, interference, and noise.
B. An Upper Bound of Transport Capacity
This section derives information-theoretic upper bounds for transport capacity by constraining simultaneous reliable communication through SINR and propagation effects. The analysis identifies interference-related terms and phase control as determinants of the bound.
- The upper-bound derivation starts from the transport capacity definition for the network in Fig. 9.
- The SINR restriction bounds the feasible rate R for simultaneous reliable communication across all communication pairs.R is the feasible data rate when all communication pairs can communicate reliably at rate R.
- The denominator accounts for noise, source power, and interference after attenuation and phase shifts over direct and reflected propagation paths.For reflected paths, dl,k,i is the propagation length through reflector i, while i = 0 denotes the direct path.
- The interference term I becomes the dominating factor because the upper bound depends on minimizing the denominator of the capacity fraction.
- The derivation assumes phase control φi ∈ [−π, π] and propagation distances dl,k,i ∈ [dmin, dmax], with bounds determined by network geometry.The resulting capacity expression considers L communication pairs, each operating at rate R.
C. Achievable Bound of Transport Capacity for Arbitrary Networks
This section develops an achievable transport-capacity bound for arbitrary network geometries because the theoretical upper bound may be unattainable when propagation lengths are geometrically coupled. The method searches node deployments and optimizes reflect-array phases for each deployment status.
- C. Achievable Bound of Transport Capacity for Arbitrary Networks: The achievable-bound algorithm accounts for network geometry because propagation lengths cannot vary independently across their theoretical ranges.Consequently, the theoretical upper bound can become unachievable in practical deployments.
- C. Achievable Bound of Transport Capacity for Arbitrary Networks: A D × D m^2 square is divided into M × M grid pixels, and a reflect-array is placed at (D/2, 0).The grid determines the spacing between adjacent intersection points.
- C. Achievable Bound of Transport Capacity for Arbitrary Networks: Each deployment status is represented by a vector containing the x and y positions of all 2L network nodes.
- C. Achievable Bound of Transport Capacity for Arbitrary Networks: For each fixed node placement, the phase controls φi vary from −π to π to search for the maximum transport capacity.The resulting maximum for status n is denoted C(n) under optimal phase control φ(n)∗.
- C. Achievable Bound of Transport Capacity for Arbitrary Networks: The procedure traverses node-deployment combinations, moving one node to generate successive statuses and evaluating each status's maximum capacity.The achievable bound is obtained by considering all deployment statuses.
IV. Numerical analysis
The numerical analysis evaluates upper and achievable transport-capacity bounds as reflect-array deployment, patch count, communication-pair count, and room size vary. Reflect-arrays improve transport capacity, while practical achievable bounds remain below ideal theoretical upper bounds.
- Reflect-arrays improve the upper bound of transport capacity compared with deployment without a reflect-array.
- 0.2 × 10^6 bits · m/s capacity increase is obtained by increasing the number of patches from 24 to 48.
- Transport capacity increases as the number of communication pairs increases because network transport capacity sums the capacities of all communications.
- Transport capacity increases with edge length from 5 m to 10 m because interference nodes can be deployed farther from receivers in larger indoor spaces.
- Achievable transport capacity improves with more reflect-array patches, while phase shifts are searched from −π to π in steps of π/180.
- 0.2 × 10^6 bits · m/s further capacity improvement is reported when two reflect-arrays are used, and successive deployment of additional arrays increases transport capacity by about 0.5×10^6 bits·m/s.
- The achievable and theoretical upper-bound curves differ because the upper bounds assume an ideal node deployment not present in practical situations.
V. Conclusion
The paper proposes smart reflect-arrays for indoor spectrum sharing and supports the approach with experiments, theoretical derivation, algorithms, and numerical analysis. The authors report significantly improved transport capacity, while noting that the experiments validate only two simultaneous users and that multi-user real-time control remains future work.
- The proposed smart reflect-array approach improves indoor spectrum-sharing capacity, with feasibility verified experimentally and effects evaluated theoretically and numerically.
- The two-user experiments validate feasibility, but multi-user testbed implementation and real-time optimal reflect-array control remain future work.