Random Settings Boost 5G Data Speeds with Intelligent Reflective Surfaces

Researchers are tackling the challenge of effectively deploying reconfigurable intelligent surfaces (RIS) in fifth-generation new radio (5G NR) systems without the computational complexity of precise optimisation. L. Yashvanth, Raju Malleboina, and Venkatareddy Akumalla, all from the Dept. of ECE at the Indian Institute of Science (IISc), Bengaluru, India, alongside Debdeep Sarkar, Nekkanti Guna Sai Kiran, and Chandra R. Murthy, demonstrate a practical approach to harnessing RIS gains. Their work, conducted entirely within the Dept. of ECE, Indian Institute of Science (IISc), Bengaluru, India, reveals that random RIS phase configurations, combined with the proportional fair scheduling already inherent in 5G NR, can achieve near-optimal throughput. This is significant because it bypasses the need for complex control mechanisms, potentially enabling widespread and cost-effective implementation of RIS technology to enhance wireless communication performance.

These surfaces, which can be digitally controlled to reflect radio signals, have emerged as a promising technology for improving signal coverage and data rates in next-generation networks.

The research team successfully integrated a custom-built RIS with a real-time 5G NR system implemented using the OpenAirInterface (OAI) framework, allowing for end-to-end testing and validation under realistic wireless conditions. By randomly adjusting the phase shifts of the RIS elements, and leveraging the proportional fair (PF) scheduling algorithm within the 5G system, the researchers observed significant gains in throughput.

The PF scheduler, designed to prioritize users with the best channel conditions, naturally aligns with the randomly configured RIS, effectively directing signal strength towards the most favourable receiver. Results confirm that even a randomly configured RIS, with minimal overhead, can deliver performance on par with more sophisticated, optimised RIS designs in a real-world 5G NR environment.

This finding suggests a viable strategy for deploying RIS technology without the substantial computational burden and signalling overhead typically associated with precise control and optimisation. Throughput gains of up to 45% were achieved when comparing the system with and without the RIS deployed, demonstrating a substantial enhancement in data transmission capabilities.

Evaluation of key performance metrics under random RIS configurations revealed a consistent enhancement in signal quality. Reference signal received power (RSRP) demonstrated a marked increase, indicating a stronger signal at the user equipment, while the block error rate (BLER) remained low, signifying reliable data transmission. The modulation and coding scheme (MCS) index also increased, reflecting the system’s ability to utilise higher-order modulation schemes for improved data rates.

Further analysis showed that the effectiveness of this approach is contingent on the averaging window of the PF scheduler. Judicious selection of this parameter allows the scheduler to prioritize transmission to the user equipment aligned with the randomly configured RIS, effectively exploiting multi-user diversity and capturing much of the benefit of beamforming without the computational burden of optimisation.

This combination delivers performance comparable to optimised RIS designs in practical 5G NR wireless communication systems, offering a low-complexity pathway to enhanced network performance. A 1024-element RIS prototype, operating at a centre frequency of 5GHz with a 300MHz bandwidth, forms the core of this work. The RIS employs a single-layer architecture comprising a 32 × 32 uniform planar array (UPA), where each unit cell measures λ0/4, equivalent to 60mm at 5GHz.

Each unit cell integrates an SMP1302, 040LF PIN diode to enable switching between two states, achieved through distinct electrical responses modelled in simulation. In the ON state, the diode presents a series connection of a 3Ω resistor and a 0.7 nH inductor, while the OFF state is represented by a series combination of a 0.3 pF capacitor and the same 0.7 nH inductor.

These configurations yield reflection phases of −28◦ and 150◦, providing an approximate 180◦ phase difference crucial for one-bit digital coding. To establish the desired beamforming, the research team analytically computed the required reflection-phase gradient based on the illumination and beam-steering angles, then realised this gradient using the one-bit quantized phase levels of 0◦ and 180◦ across the RIS array.

Full-wave electromagnetic simulations, conducted using CST Microwave Studio, validated scan-loss-free beam-steering across a 0◦ to 60◦ range, achieving a half-power beamwidth (HPBW) of 6◦ and maintaining sidelobe levels below −9 dB. A field-programmable gate array (FPGA) based controller module manages the phase shifts across the RIS cells, with beam steering functionality experimentally verified within an anechoic chamber.

The study deliberately bypassed complex optimisation of RIS phases, instead leveraging the proportional fair (PF) scheduling mechanism inherent in 5G NR to prioritize data transmission to user equipment (UE) aligned with the randomly configured RIS. The persistent challenge of extending reliable wireless coverage, particularly indoors and in dense urban environments, may have found an unlikely ally in reconfigurable intelligent surfaces (RIS).

This work demonstrates that RIS, essentially smart mirrors for radio waves, can significantly boost 5G performance even when their configuration is not meticulously optimised. This is a potentially transformative finding because it sidesteps a major practical hurdle in RIS deployment; the complexity of optimal RIS control has long threatened to outweigh the benefits, particularly in dynamic environments.

A system that performs well with random configurations dramatically lowers the barrier to entry, opening the door to widespread adoption, such as cost-effective, passively powered surfaces embedded in buildings to enhance connectivity without requiring intricate control infrastructure. However, the reliance on proportional fair scheduling within the 5G network is a critical caveat, as its effectiveness will depend on network load and specific traffic patterns.