Adapting pretrained artificial intelligence and machine learning (AI/ML) models for on-device inference across diverse environments is challenging due to the extensive offline data collection and training required. This highlights the necessity for online model training or adaptation, where devices locally process training data and perform adaptation via on-device engines. Such capabilities and technologies are essential for various wireless AI/ML features requiring on-device inference to be considered in the upcoming Third Generation Partnership Project (3GPP) 6G standardization works. For battery-powered devices, model adaptation must be efficiently executed in real-time with well-designed timelines, as environmental conditions can change rapidly and vary significantly across locations. In this paper, we present results from over-the-air (OTA) prototyping experiments using handset devices to demonstrate the practical effectiveness of real-time on-device adaptation. We use AI/ML based beam prediction, as outlined by 3GPP 5G-Advanced standards, as an example. Our OTA results show that on-device model adaptation can be completed within 30 seconds in an unfamiliar environment to achieve benchmark performance, strongly supporting the feasibility of real-time on-device model adaptation in 6G communications. The findings and methodologies discussed in this paper can be extended to benefit various other wireless and non-wireless real-world scenarios.

Artificial intelligence (AI) or machine learning (ML) based beam prediction is currently studied in the 3rd Generation Partnership Project (3GPP) fifth generation (5G)-Advanced new ratio (NR) standardization for future commercialization and standard evolution towards sixth generation (6G) communications, wherein time domain (TD) beam prediction is an important use case. The targets for such 3GPP studies and standardization are to lower power consumed at user equipment (UE) and reference signal (RS) overhead that are currently needed by frequent beam measurements due to UE rotation and mobility. To meet such targets, in this paper, we investigate AI/ML based algorithms facilitating TD beam prediction suitable for 5G-Advanced beam management (BM), including RS receive power (RSRP) prediction and beam change prediction. The proposed AI/ML algorithms are first evaluated through computer simulations with new UE mobility models based on recent standard evolutions in 3GPP. Then we further present over-the-air test results achieved by such AI/ML algorithms, using based station (BS) and UE compliant with 3GPP standards. Evaluation results show that the proposed schemes can accurately predict future beams and reduce large amount of the power consumed at the UE for BM, which also demonstrate feasibility of AI/ML based BM for 5G-Advanced and future 6G communications.

Analog beamforming is the predominant approach for millimeter wave (mmWave) communication given its favor-able characteristics for limited-resource devices. In this work, we aim at reducing the spectral efficiency gap between analog and digital beamforming methods. We propose a method for refined beam selection based on the estimated raw channel. The channel estimation, an underdetermined problem, is solved using compressed sensing (CS) methods leveraging angular domain sparsity of the channel. To reduce the complexity of CS methods, we propose dictionary learning iterative soft-thresholding algorithm, which jointly learns the sparsifying dictionary and signal reconstruction. We evaluate the proposed method on a realistic mm Wave setup and show considerable performance improvement with respect to code-book based analog beamforming approaches.

In this paper, a doubly iterative receiver for orthogonal frequency-division multiplexing (OFDM) systems is designed to mitigate impulsive interference based on pulse blanking. Although being a very efficient countermeasure, pulse blanking introduces intercarrier interference in OFDM systems leading to system performance degradation. The doubly iterative receiver is designed to overcome this drawback. The inner loop combines iterative demodulation and decoding to cancel intercarrier interference. To make soft information available to the outer loop, the equivalent noise power is derived to employ soft demodulation. The outer loop improves the performance of pulse blanking by performing a hypothesis test, which is used to decide if peaks in the received signal are due to impulsive interference or the large peak-to-average-power ratio common to OFDM signals. This allows to optimize the blanking threshold through maximizing the signal-to-interference-and-noise ratio. To evaluate performance and complexity of the proposed receiver, extrinsic information transfer (EXIT) chart analysis is carried out, where EXIT functions for pulse blanker, demodulator, and decoder are derived. An EXIT chart-based trellis search approach is introduced to calculate the loop schedule achieving target bit-error-rate performance with minimized complexity. Numerical results show a considerable bit-error-rate performance improvement as well as a remarkable complexity reduction.

Iterative channel estimation and detection (ICED) can provide a better performance as the channel estimation can be improved through iterations. For a multiple‐input multiple‐output channel, owing to a large size of the signal alphabet, ICED becomes less practical unless a low‐complexity detector such as a lattice reduction‐based detector is employed. However, since the lattice basis reduction has to be carried out for each iteration, the resulting complexity can be still high. In this study, a computationally efficient technique was proposed to perform the lattice basis reduction within ICED over both static and slowly time‐varying block‐fading channels, where orthogonal defect of basis, as well as error probability are considered to decide whether or not basis reduction is needed for each iteration.

For iterative detection and decoding (IDD) in multiple‐input–multiple‐output (MIMO) systems, although the maximum a posteriori (MAP) detector can achieve an optimal performance, because of its prohibitively high computational complexity, various low‐complexity approximate MAP detectors are studied. Among the existing MIMO detectors for the non‐IDD receivers, lattice reduction (LR)‐aided detectors can provide a near maximum‐likelihood (ML) detector's performance with reasonably low complexity, and they could be modified to be used in the IDD receivers. In this study, the authors propose a bit‐level LR‐aided MIMO detector whose performance can approach that of the MAP detector, where a priori information is taken into account for soft‐decisions. Furthermore, the proposed method can be extended to large dimensional MIMO systems by channel matrix decomposition and successive interference cancellation, by which a significant complexity reduction can be achieved. Through simulations and complexity analysis, it is shown that a near‐optimal performance is obtained by the authors proposed low‐complexity bit‐level LR‐aided detector for the IDD in the MIMO systems.

Hierarchical modulation (HM) enables unequal priority transmissions using a signal constellation of non-uniformly spaced constellation points, which is an important property for broadcast systems. Using bit interleaved coded modulation (BICM) and iterative decoding (ID), the performance of HM systems can be further improved. In this paper, we study HM system with BICM-ID or HM-BICM-ID. Since the performance of BICM-ID depends on signal mapping rules, we derive a mapping rule based on distance properties to minimize the bit error rate (BER). Furthermore, we perform the BER performance analysis for HM-BICM-ID based on binary erasure channel (BEC) modeling, which provides BER prediction without time-consuming simulations. Based on the derived BER prediction scheme, we can also perform system design and optimization for applications of HM-BICM-ID. For example, we are able to optimize the constellation priority parameter in a broadcast system to maximize the throughput.

For iterative detection and decoding (IDD) in multiple-input multiple-output (MIMO) systems, the log-likelihood ratio (LLR) of each coded bit can be found by an optimal bit-wise maximum a posteriori probability (MAP) detector. However, since this MAP detector requires a prohibitively high computational complexity, low-complexity suboptimal detectors are desirable. In this paper, lattice reduction (LR)-based MIMO detection is investigated to derive a low-complexity detector that can achieve near MAP performance for IDD. In order to approximate LLR values incorporating the extrinsic information provided by a soft-input soft-output (SISO) decoder, bit-wise LR-based minimum mean square error (MMSE) filters are derived. Furthermore, in order to minimize the performance degradation due to quantization (or rounding) errors in the LR-based detection, a low-complexity integer perturbed list generation method is proposed, where no tree search is used by taking advantage of a near orthogonal channel basis obtained by LR. Through a complexity analysis and simulations, it is shown that the proposed approach achieves near optimal performance, while the complexity is comparable with that of the MMSE soft cancellation method, which is known to be computationally efficient. As a bit-wise detector, a parallel implementation of the proposed method would be straightforward, which lowers the detection delay.