Recently, methods have been explored to construct radio wave propagation models using machine learning, enhancing generalizability via transfer learning and fine-tuning. The conventional applications of machine learning are intended to estimate propagation environments within the learned area. Therefore, estimating propagation environments in unknown areas that are not represented in the training data is not expected and is challenging. This is presumably because, with limited data, the machine learning model does not adequately learn the physical propagation mechanisms. In this paper, we propose a path loss estimation method that classifies the dominant paths at each receiver based on physical propagation mechanisms. Specifically, we categorize the propagation paths into simple direct waves and single-reflection paths, as well as complex non-line-of-sight (NLOS) and estimate each accordingly. For line-of-sight (LOS) path loss, we use a statistical model fitted to measurement data to achieve easy and high-precision estimation. Additionally, for NLOS, we consider reflection loss based on a practical environment for single-reflection paths and apply a radar equation method that uses first Fresnel zone to estimate path loss accurately from close to distant ranges. For other NLOS, we propose a machine learning model that utilizes information from first Fresnel zone centered around the primary propagation path, estimating path loss at specific locations. The evaluation of the root mean squared error (RMSE) between actual measurements and estimates indicates that the proposed method improves the maximum RMSE by up to 11.5 dB compared with traditional machine learning approaches.

This paper presents a full-stack testbed for evaluating inter-site central processing unit (CPU) cooperation in large-scale cell-free massive multiple-input multiple-output (CF-mMIMO) systems. In such systems, multiple CPUs are distributed across different sites to address computational and fronthaul capacity limitations. However, this distributed architecture introduces inter-site interference that degrades radio quality at site edges. To mitigate this issue, two inter-site CPU cooperation schemes, tight CPU cooperation (TCC) and loose CPU cooperation (LCC), have been proposed. Despite their theoretical promise, their practical feasibility in an end-to-end system and their effectiveness under real outdoor conditions remain unclear. To address these gaps, we have developed a full-stack 5G testbed to experimentally validate the feasibility and performance of TCC and LCC. Measurements conducted over a 5400 m2outdoor area demonstrate that both methods significantly improve uplink throughput by suppressing inter-site interference. We further analyze the relationship between throughput, signal power, and interference power on a per-user basis. Finally, we discuss key challenges for scaling CF-mMIMO deployments.

Toward 6G in around 2030, cell-free massive MIMO (CF-mMIMO) is expected to provide high radio quality everywhere. To deploy CF-mMIMO over a wide area, a distributed site architecture has been proposed. Central processing units (CPUs), which process radio signals, are deployed to multiple sites, which are physical bases for aggregating radio signals. In a distributed site architecture, the radio quality of user equipments (UEs) located at the border of the site degrades because access points (APs) cannot be used in signal processing between sites. We consider the application of a multi-homing technique to the fronthaul (FH) between the AP and sites. By multi-homing, an AP connects to multiple sites via the fronthaul in order to cooperate with APs in different sites. Since the radio signals of the multi-homing AP can be received by the CPUs of multiple sites, radio quality is improved by the cooperation of APs in different sites. However, multi-homing has an issue: the FH cost increases due to the installation of additional FH links. In order to balance the suppression of FH cost increase and improvement in radio quality, we propose an AP connection method for minimizing the number of total FH links. In the proposed AP connection method, APs in different sites are connected directly. The number of FH links can be reduced because connected APs can share one FH link. The results of computer simulation and FH cost estimation show that the proposed multi-homing methods improve the 5%-tile throughput 1.7-fold compared to the distributed site architecture while suppressing the FH cost increase compared to existing connection methods.

We analyze the impact on the wireless quality of time division duplex Cell-Free massive multiple-input multiple-output (TDD CF-mMIMO) caused by phase variation in reference signals that modulate RF signals due to clock distribution over a radio access network (RAN). Phase variation in reference signals causes temporal phase variation due to temporal difference in transmission and reception in TDD. In addition, since APs are distributed among a RAN, phase variation also occurs due to clock distribution through fronthaul depending on the placement of reference clocks (RCs), which is a challenge for practical wide-area deployment. We propose RC distribution for practical user-centric RAN, considering use of IEEE 1588 or radio over fiber (RoF) technology. We analyze phase variation of reference signals and evaluate SINR of TDD CF-mMIMO by numerical simulations, revealing that clock distribution from distributed RCs with RoF secures high SINR.

The next-generation mobile networks will be devel- oped based on virtualized and distributed radio access network (RAN), necessitating cost-effective and flexible deployment solutions to meet latency requirements without incurring excessive costs. The O-RAN alliance has defined the architecture for these future networks, where the RAN controller is split into the Non-Real-Time RAN Intelligent Controller (Non-RT RIC) and the Near-Real-Time RAN Intelligent Controller (Near-RT RIC). These controllers, virtualized and deployable on general- purpose processors at edge nodes, can be distributed across the entire geographical area covered by a mobile network. This is particularly vital for user-centric RAN with cell-free massive MIMO (CF-mMIMO). The inherent high control traffic overload, due to not only continuous access point (AP) cluster updates but also from extensive controls for the user-centric RAN, runs the risk of overloading the transport network. Therefore, con- troller deployment optimization becomes essential. Additionally, in scenarios involving critical applications, latency is a key factor to consider, adding to the optimization challenges. This paper addresses the Near-RT RIC deployment problem, considering latency requirements and transport network limitations. Mixed integer linear programming (MILP) is utilized to solve the deployment issue. The solution's effectiveness is validated against different network topologies and design parameters.

This paper proposes a distributed deep reinforcement learning (DRL) method with multiple learners for AP clustering in large-scale Cell-Free massive MIMO (CF-mMIMO). In the deployment of large-scale CF-mMIMO with many user equipments (UEs) and access points (APs), it is necessary to perform AP clustering according to the demand and movements of each UE in a lightweight manner and with high inference accuracy. However, existing DRL-based methods have struggled to learn diverse and site-specific radio environments and provide high inference accuracy with small amounts of data and small neural network (NN) models for lightweight. To address this problem, the proposed method classifies learners for each radio environment with the Reference Signal Received Power (RSRP) between surrounding APs of the UE to perform learning and inference with these multiple learners. Furthermore, the proposed method dynamically adjusts the association between UE and learners based on the fluctuation in RSRP due to the UE's movements, thereby ensuring sufficient agility for user mobility. This dynamic association of UEs and learners for each radio environment enables efficient learning and improved inference accuracy by focusing on UEs in similar radio environments, even with small amounts of data and small NN models. Simulation evaluations based on actual urban structures demonstrated that the proposed method realizes AP clustering with higher inference accuracy than existing methods, even with small amounts of learning data and small NN models.

We consider fairness scheduling in a user-centric cell-free massive MIMO network, where L remote radio units, each with M antennas, serve $K \approx LM$ user equipments (UEs). Recent results show that the maximum network sum throughput is achieved where ${K_{{\text{act}}}} \approx \frac{{LM}}{2}$ UEs are simultaneously active in any given time-frequency slots. However, the number of users K in the network is usually much larger. This requires that users are scheduled over the time-frequency resource and achieve a certain throughput rate as an average over the slots. We impose throughput fairness among UEs with a scheduling approach aiming to maximize a concave component-wise non-decreasing network utility function of the per-user throughput rates. In cell-free user-centric networks, the pilot and cluster assignment is usually done for a given set of active users. Combined with fairness scheduling, this requires pilot and cluster reassignment at each scheduling slot, involving an enormous overhead of control signaling exchange between network entities. We propose a fixed pilot and cluster assignment scheme (independent of the scheduling decisions), which outperforms the baseline method in terms of UE throughput, while requiring much less control information exchange between network entities.

We consider scalable RAN management for large-scale deployment in a distributed central processing units (CPUs) environment with cell-free massive MIMO (CF-mMIMO). In distributed CPUs among multiple sites, there is a problem of low radio quality for users at the site edge areas, which are the boundaries between sites. This is due to inter-site interference between UEs processed by different CPUs and a reduction in received signal power owing to the inability to associate with the surrounding APs connecting to other sites. To address this problem, our approach is to deploy the vCPUs of users at the site edges to higher-level sites based on the physical hierarchical structure of the RAN. This hierarchical deployment allows the formation of broad AP clusters associated with group APs across the sites and improves radio quality. However, deploying vCPUs at higher-level sites causes the radio signal to flow into the backhaul (BH), significantly increasing the transmission load depending on the size of the AP cluster. Thus, we formulate an optimization problem to improve user throughput everywhere under the constraints of the RAN physical resources by managing the deployment of vCPUs and AP clustering. This optimization problem is non-linear and non-convex and requires inverse matrix calculations, resulting in computational complexity. Therefore, we propose a lightweight list-processing algorithm with reference signals of APs around users that does not use inverse matrices calculation and metaheuristic search. Simulation results show that the proposed method improves user throughput and provides lightweight calculation with a large number of UEs compared to existing methods.

This paper suggests and analyzes adaptive bit allocation for the quantization of uplink signals of a cell-free massive MIMO (CF-mMIMO) system under limited fronthaul capacity. Specifically, we consider a CF-mMIMO system with hybrid processing, where at each access point (AP) a singular-value decomposition (SVD) reduces the number of streams that need to hauled, each stream is quantized with an adaptive number of bits, and a central processing unit (CPU) decodes the uplink signals. The hybrid processing, which the authors previously proposed, had been shown its potential to reduce fronthaul load without severe degradation of the spectral efficiency. However, as the bandwidths of the wireless system increases, the fronthaul capacity becomes comparatively tight, and the quantization noise would degrade the spectral efficiency severely. In order to improve the performance under such a scenario, this paper proposes algorithms for adaptive bit allocation of the output streams, based on the optimization of the average SNR, or the sum capacity. In addition, appropriate selection of the number of streams of the hybrid processing in each AP is also discussed. Computer simulations verify the effectiveness of these proposed methods.

In this paper, we propose low-complexity digital predistortion (DPD) schemes based on FastGRNN to compensate for the nonlinearity of RF power amplifiers. Conventionally, high-precision recurrent neural network (RNN) models, such as long short-term memory (LSTM) and gated recurrent unit (GRU), have been used to model the behavior of amplifiers, and their excellent compensation performance has been shown in terms of error vector magnitude (EVM) and adjacent channel power ratio (ACPR) has been demonstrated. However, their complex structures result in high computational complexity. To solve this issue, the proposed method is designed to significantly reduce the complexity without significant performance degradation by appropriately applying the FastGRNN models to the DPD. Complexity analysis and experiments using a power amplifier in the 2.0 GHz frequency band showed that the proposed method achieved comparable EVM performance to LSTM with 29.2% floating point operations (FLOPs) and 27.1% trainable parameters.

In Japan, the 5th generation mobile communication system (5G) became commercially available in 2020. The millimeter wave bands such as 28GHz is being used to achieve the peak rate of 10 [Gbps] or higher targeted for 5G. In the late 2020s, low latency and high-capacity data transmission over both the up and down links will become important. This is because 5G will be utilized in the late 2020s, and telemedicine and teleoperation using 4K/8K and other high-definition video transmission will become widespread. In this study, we propose a relaying method that converts frequency multiplexing into spatial multiplexing during relaying, with the goal of achieving low latency and high capacity relaying communications. The user equipment, base stations, and relay stations have different conditions in terms of transmission power and number of antennas. Therefore, the proposed method achieves high capacity by frequency multiplexing in the link where the number of antennas is limited. In addition, the proposed method uses spatial multiplexing to achieve high capacity while suppressing the increase in resource usage in the link where multiple antennas are available. The 39 GHz band, which has more frequency resources than the existing 5G bands, is used for the evaluation in the link of frequency multiplexing. Then, the 28 GHz band, which is used commercially in 5G, is used for the evaluation in the link of spatial multiplexing. For low latency relaying, analog circuits are used during the relaying process to convert between frequency-multiplexed and space-multiplexed signals without modulation and demodulation, while maintaining the number of multiplexes. In this paper, simulation evaluations show that the proposed method improves the communication distance where the throughput exceeds 4 [Gbps] to 6.5 times that of 39 GHz band 5G communications without relaying, and to 1.3 times that of RF repeaters in conventional relaying methods that use the 39 GHz band both before and after relaying, indicating that the uplink communication distance can be extended.

The pilot contamination in cell-free massive multiple-input-multiple-output (CF-mMIMO) must be addressed for accommodating a large number of users. In previous works, we have investigated a decontamination method called subspace projection (SP). The SP separates interference from co-pilot users by using the orthogonality of the principal components of the users’ channel subspaces. For CF-mMIMO system with SP, non-overloaded pilot assignment (PA) and overloaded PA can be considered. Non-overloaded PA, where each radio unit (RU) does not assign the same pilot to different users, limits the number of associated RUs per each UE and this reduces the potential spectral efficiency (SE) of the system. On the other hand, non-overloaded PA reduces channel estimation error induced by contamination. This paper compares non-overloaded PA and overloaded PA, and introduces overloaded PA methods adjusted for the decontamination in order to improve the sum SE of CF systems. Numerical simulations show that the overloaded PA methods give higher SE than that of non-overloaded PA at a high user density scenario.

Digital predistortion (DPD) using neural network(NN) has attracted attention as a promising technique for compensating complex nonlinear distortions caused by wideband radio frequency power amplifiers. However, NN-DPD is difficult to implement in hardware due to the large number of multiplications. In this paper, we propose an NN-DPD using DeepShift that replaces neural network multiplications with bitwise shift and sign flipping during both training and inference. First, when DeepShift was applied, we showed that applying residual learning to DPD can reduce performance degradation. Next, we examined pre-trained models using floating-point baseline models and scratch-trained models and found that the pre-trained models perform better in exchange for requiring pretraining. The scratch model, on the other hand, has the advantage of improving the training computational complexity. In an actual experiment of passing a signal with DPD applied through a power amplifier at 2.14 GHz, the error vector magnitude was slightly degraded from 1.31 % trained by the floating-point model to 1.57 % with the pre-trained model and 1.90 % with the scratch model.

Reconfigurable meta-surface reflectors, which are key components of intelligent radio environment (IRE) and intelligent reflecting surfaces (IRSs), have recently been gathering attention as a solution for coverage hole problems of mobile communication areas. In recent researches, many proposals related to a reconfigurable meta-surface reflector using diodes and new materials such as liquid crystal (LC) and graphene have been reported. Although reconfigurable metasurface reflectors using new materials have the capability to control reflection waves, it remains a concept and a simulation in high-frequency bands, which are not used in current mobile communication service. In this paper, we proposed and developed the world's first direction-variable liquid crystal meta-surface reflector. Through electromagnetic field simulation based on FDTD and measurement in an anechoic chamber, it was verified that the proposed meta-surface reflector achieves 260-degree reflection phase control in each unit cell and ±60-degree reflection direction control.

Cell-free massive multiple-input multiple-output systems are expected to provide faster and more robust connections to user equipments (UEs) by cooperation of a massive number of distributed access points (APs). Energy efficiency (EE) is becoming an important indicator to design and operate networks; to improve EE, use of sleep-mode of APs (SMA), also called AP switch on/off, for selected APs has been investigated. Although previous works analyze the performance of SMA in the presence of user-centric clustering (UCC), these two techniques are assumed to not affect each other. In this paper, we propose a new greedy combining algorithm (GCA), where SMA and UCC work alternately to obtain better performance, and show its superiority over a conventional algorithm. Example simulations show that GCA can achieve 44% higher total EE for 8 UEs and 59% for 16 UEs with 64 APs. Additionally, GCA also provides higher minimum spectral efficiency thanks to its structure of the algorithm.

In B5G mobile communication, ultra-high-speed UL transmission of over 100Gbps is expected to lead to the realization of a cyber-physical system [1]. Although the UL channel capacity can be improved by increasing the number of UE transmitting antennas and adopting the UL MIMO scheme, the advantage conferred by having more antennas cannot be maximized due to the fact that the maximum transmission power of a UE is limited. Furthermore, although THz band radio waves are expected to be used for high-speed data transmission, it is difficult to obtain a communication distance similar to that of conventional mobile communication systems due to the large propagation loss. In order to realize multi-antenna uplink MIMO for UEs, we have proposed a novel UE configuration referred to as the “virtualized terminal”, in which UE antennas are distributed to the relay devices, a UE transmits UL signals in the THz wave band, and relay devices convert the signals to mmW band radio signals and relay them to a BS. In this paper, the communication distance of the virtualized terminal is evaluated by the link budget calculation using a practical configuration and the parameters of the RF components. Furthermore, we show the THz antenna gain of each THz link which satisfies the required C/N and the maximum data transmission speed of the virtualized terminal with uplink MIMO.

The virtualized terminal using a frequency-converting radio frequency (RF) relay device has been proposed as a new terminal configuration to achieve the 100+ Gbps uplink (UL) speeds required for Beyond 5G (B5G) mobile communication [1]. In B5G, where time division duplex (TDD) is considered to be applied as a duplexing method, the RF relay device should be able to detect UL and downlink (DL) timing and switch the relay circuit by the power fluctuation of the received signal for compact size and low power consumption. We propose a new timing detection method for RF relay devices that uses not only the received DL signal, but also the received UL signal. The lower bounds of the required carrier to noise ratio (C/N) and received power at the receiver to enable detection of TDD timing were obtained from the received signals for both UL and DL by simulation. Based on the experimental value of the conventional type of user equipment (UE), the receiver sensitivity of the UE communicating with the RF relay device using the 4.8 GHz frequency bandwidth was estimated. As a result, we found that even if the received DL signal of the RF relay device is lower than the lower limit of timing detection, it can still be received by the UE, and the RF relay device can switch the relay circuit if the UL signal from the UE has sufficient received power for timing detection, thus increasing the distance that can be communicated between the base station (BS) and UE.

Reconfigurable meta-surface reflectors are a promising technology to overcome the problem of coverage holes in millimeter wave communications. This reflector is composed of many tunable passive elements, whose reflection amplitude changes depending on the reflection phase. Thus, to fill coverage holes with higher reflected power, it is necessary to design a scattering pattern considering both reflection phase and amplitude. However, previous studies, which focus on phase adjustment to design the scattering pattern of the reconfigurable meta-surface reflector, assume that the reflection amplitude is constant regardless of the reflection phase. In this paper, a phase adjustment method based on both reflection phase and phase-dependent reflection amplitude is proposed to increase the received signal power in the desired reflected direction. The effectiveness of the proposed method is confirmed via numerical simulations and experimental measurements by using the developed liquid crystal metasurface reflector. The simulation and measurement results show that the proposed method enables an increase in received signal power in the desired reflected direction.

In higher frequency bands such as the millimeterwave band, which is expected to be utilized in the future, the received power fluctuates significantly due to changes in the propagation environment. Therefore, the spatial and temporal spectrum sharing between different wireless systems will be more practical. Some methods have been proposed to predict future spectrum usage, but prediction errors are unavoidable. There is a trade-off between the detection rate of future available spectrum resources and the false detection rate caused by prediction errors. Still, there is no established method for balancing these two factors. One conservative prediction to reduce false detections is to set a certain margin. However, the fluctuation tendency depends on the location, which results in an excessive margin in some sites, resulting in a lower detection rate. In this paper, we propose a method to set a location-dependent margin by considering the error characteristics of the constructed prediction model of received power. The proposed method enables us to increase the detection rate while reducing the false detection rate. The evaluation results by simulation in an urban area show that the proposed method can improve the detection rate by 10.6% while keeping the false detection rate, compared to when using a constant margin.

For the efficient use of limited frequency resources, the spectrum sharing allocation between mobile networks and different existing radio systems is promising. In this scheme, we focus on an allocation method that involves sharing spectrum bands with a mobile network operator (MNO). Recently, to improve the spectrum efficiency, dynamic spectrum allocation methods according to the MNO demand were proposed. However, in demand-based dynamic allocation, channel discontinuity in the frequency and time directions in the same MNO occurs when mapping the channel to MNOs. This channel discontinuity reduces the utility of shared frequencies. To address this problem, we propose a novel spectrum allocation method that maximizes frequency and time continuity for demand-based dynamic spectrum sharing. Our performance evaluation shows that compared with the existing methods, the proposed algorithm can improve the spectrum efficiency by 6%, channel continuity by 23%, and fairness by 5%.

This paper presents an adaptive impedance control of antennas on user equipment (UE) for a Long Term Evolution (LTE) system that supports MIMO reception in downlink and single antenna transmission in uplink. The proposed control method balances the performances of both links, duplexing with a frequency division duplex (FDD), without additional information from eNode B (eNB). In addition, computer simulations verify the effectiveness of the proposed method in the usage environment, where a user holds the UE under multipath propagation channels.

In next generation broadband wireless access (BWA) systems, cell shrinkage and blind zones might occur at cell edges due to constraints on transmission power and the increase in propagation loss with broadband and higher frequency bands, respectively. Thus, hierarchical BWA systems consisting of micro and macro base transceiver stations (BSs) installed around the macro BS are considered as one of the candidate radio access network architectures for the next generation BWA systems. In the hierarchical BWA system, a wireless backhaul between micro and macro BSs needs high capacity transmission to serve all subscriber stations (SSs) in each micro BSs. In this paper, we developed a prototype of a time division duplex (TDD) / orthogonal frequency division multiplexing (OFDM) wireless broadband backhaul, with a space division multiple access (SDMA) technique applied, for the next generation hierarchical BWA systems and carried out laboratory experiments in an anechoic chamber room to evaluate the transmit beamforming performances. The laboratory experiments showed that the TDD/SDMA OFDM wireless backhaul was capable of realizing accurate transmit beamforming with a calibration function in each sub-carriers, and the throughput performance with a 4-element array antenna was achieved more three times more than that without SDMA.

We carried out field trials of the developed FDD-SDMA packet cellular testbed system and demonstrated the effectiveness of the proposed forward link beamforming algorithm. Our forward link beamforming algorithm use not only unidirectional access terminal estimation but also angular spread estimation that is the proposed method and available to extract angular spread characteristics from a correlation matrix accumulated with reverse link signals through eigen decomposition. Moreover, the proposed beamforming algorithm is equipped with forward link array antenna weighting vector normalization to prevent from demodulation and channel estimation errors with omni-beamformed pilot channel. During the field trials, we confirmed that the throughput performance per access point of the proposed forward link beamforming algorithm exceeded that of the conventional beam-null forming algorithm with only unidirectional access terminal estimation and were convinced of the effectiveness of the proposed algorithm in a real field environment with angular spread. space division multiple access (SDMA); transmit beamforming; forward link beamforming; downlink beamforming; smart antenna; array antenna; frequency division duplexing (FDD); packet cellular system; field trial.

The authors have studied the packet cellular system utilizing SDM (space division multiplexing) technology. This paper evaluates downlink beamforming performance of SDM based FDD packet cellular system by computer simulation under the condition that access terminals are independently and identically distributed in each cell following uniform distribution utilizing premeasured radiation patterns of the 12-element circular array antenna and makes explicit the relation between the angular spread and the number of spatial scheduled users corresponding to the maximal cell throughput. The simulation results show that the average cell throughput can reach up to 9Mbps. The results also show that as angular spread increases, average cell throughput becomes smaller, but can still reach up to 4.44Mbps even when angular spread equals to 10 degree. This indicates that the realistic approaches we adopted are quite promising for next generation mobile communication systems.