Advanced 5G around 2025 is expected to support various services with a combination of different quality requirements including data rate, latency, and error rate. To satisfy the quality requirements, we proposed adaptive placement of base station functions in a radio access network (RAN). When traffic congestion occurs on the transport link, the quality requirements cannot be satisfied. To avoid congestion on the transport link, one approach is to formulate a constraint on the transport link capacity based on predicted user traffic. Then, function placement is determined under the constraint. However, the predicted value of user traffic has errors, because user traffic changes in a random manner. In addition, traffic on the transport link is changed by function placement even if the same user traffic occurs. In this paper, we propose a control method for the constraint on the transport link by feedback loop. Our method calculates the constraint by considering both feedback of prediction error of the user traffic and the next base station function placement. We evaluated the proposed method by performing simulations. The result shows that our method can determine base station function placement without congestion on the transport link caused by the constraint when the user traffic changes dynamically.

We started a new research project in the “advanced 5G” era that aims at accommodating various types of communications involving current and emerging services with different data flow-level quality requirements. In this paper, the objectives and the technical aspects of the research project are introduced. We propose an architecture based on a virtualized radio access network (vRAN) that enables adaptive control of equipment resources and location of functions in the vRAN environment in accordance with spatially and temporally changing communication demands. The seven planned research items that are essential for realizing the advanced 5G network are listed as follows: blockage prediction, new radio access technologies (RATs) and their implementations with software-defined radio (SDR), adaptive interference and resource control, integration of radio and fiber resource control, highly efficient access transmission control, adaptive placement of BS functions, and quality aware traffic pattern prediction.

In the context of 5G mobile networks, various types of services need to be supported. Network slicing is key to enabling a diverse range of services to be accommodated in the same physical network by providing a customized logical network. In a radio access network (RAN), the placement of RAN functions can be customized for each slice in order to satisfy service requirements. In this paper, we propose a RAN slicing architecture which has multiple sets of functional splits and function placement in one cell. For better resource efficiency with the proposed architecture, resource allocation should be updated by RAN controller at a specific interval. We propose a resource allocation method for the architecture which maintains isolation between slices while taking the fulfillment of the Service Level Agreement (SLA) into consideration. The results show that the proposed method guarantees isolation between slices and keeps the probability of SLA fulfillment close to that of the conventional method.

This paper presents a novel clustering and centralized resource scheduling algorithm for 3- dimensional (3D) in-building femtocells (FCs) for an intra MAC functional split heterogeneous CRAN with control-/user-plane decoupled architecture to reuse resources in FCs. We consider two levels of FC clustering where level 1 clustering defines an initial 3D FC cluster size such that each FC is allocated statically with an equal number of orthogonal resource blocks subject to interference constraints. Whereas, level 2 clustering defines a cooperating set of FCs per floor per level 1 cluster to enable joint transmission Coordinated Multipoint (CoMP) during off-states of their transmit power. We find the probability of CoMP- state of a femto UE by modeling its traffic activity in non-CoMP-and CoMP-state using the Birth-Death process, derive the CoMP capacity gain, and update the initial level 1 cluster size subject to fronthaul's bandwidth and latency constraints. With numerical and simulation results, we show that the proposed clustering and centralized scheduling algorithm, by applying CoMP in level 1 clusters each consists of 2 floors with 4 FCs per floor in a building, can achieve the maximum capacity gain of 44% per building and 32.15% for the overall system.

In this paper, an explicit interpretation on how different enabling technologies over generations are brought up and coordinated for migration from a distributed, to a centralized, and then to a virtualized RAN for 5G and beyond cellular is given. In doing so, by pointing out major enabling technologies, we present a framework for RAN evolution and discuss its architectural migration over time and generations, namely from the current 4G to the upcoming 5G and beyond mobile networks, based on both functional split and physical placement in access networks.

In traditional heterogeneous cellular networks (Het-Net), base stations (BS) basically associate users based on received signal power. However, in the case of multiband HetNet, in which macro BSs and smallcell BSs use different frequency bands, this conventional cell association method is not effective because there is no interference between macro BSs and smallcell BSs. Additionally, there are huge differences in coverage and available bandwidth. These differences cause inefficient association which causes the imbalance between achievable rate and traffic demand. In order to overcome this problem, we propose a novel cell association method based on the combinatorial optimization. The proposed method considers achievable rates, traffic demands and the number of users belonging to each BSs simultaneously. Numerical simulation results show that the proposed association method achieves system rate gain twice as high as the conventional one.

Due to the nature of mobile communications, traffic load is not equal throughout the cellular network. While much traffic is concentrated on a specific cell (i.e., called a hotspot cell), the traffic load on adjacent cells is often low. To efficiently accommodate the imbalance in traffic, load balancing (LB) between base stations has attracted attention in recent years. This paper proposes an advanced LB method for LTE/LTE-A cellular networks which utilizes exiting mechanisms such as handover parameter control (HPC) and cell coverage control (CCC). Since the two mechanisms have advantages and disadvantages, the proposed method selects the appropriate LB method according to the situation. The paper presents overall procedures and a detailed selection algorithm to maximize the offloading effect on the hotspot cell with avoidance of interference on surrounding cells. In addition, through a simulation, we confirmed that 1) the proposed method could achieve higher offloading effect than the native HPC by 20% of cell capacity and that 2) it could avoid interfering with the surrounding cells compared to CCC when the traffic load on the hotspot cell is relatively low.

People use wireless communication systems regardless of the location, outdoor or indoor. It is important to predict the coverage area of wireless communication systems in all places. The goal of this study is the accurate estimation of coverage area in buildings including multi-story buildings. If the received power at the outside buildings of each floor is estimated correctly, the coverage area in indoor buildings for wireless communication can be predicted by higher accuracy. Therefore, the height pattern that shows the characteristic of received power above ground considering surrounding buildings is necessary. To design the novel height pattern model, this paper investigates the characteristics of the received power corresponding to the variation in receiver height near the building. As an approach to analyzing the effect of the varying receiver height and buildings, this paper compared simulated and measured values to traditional propagation loss models. A ray-tracing simulation and measurement in an actual environment were also performed. As a result, the received power near the building was observed to be close to that of the Walfisch-Ikegami model at a lower height than that of the surrounding buildings. In contrast, when the receiver height exceeded that of the surrounding buildings, the characteristic of the received power resembled that of the ground-reflection model by two rays. In addition, the level of average received power was close to that in the free space propagation loss. However, Authors found that there was greater standard deviation of the received power near the building than far away. Moreover, the received power of the measurement is lower than that of the ray-tracing simulation, making it difficult to estimate the received power at each height with the traditional model. To design the novel height pattern, the receiver height and its distance from the closest building must therefore be considered.

Multiple Input Multiple Output (MIMO) wireless system is the most promising technology for next generation cellular mobile systems due to its high-speed data transmission capability. This paper focuses on the MIMO propagation characteristics in 2GHz bands. From the viewpoint of cell coverage planning of OFDMA-MIMO systems, it is important to know the key parameters required to estimate MIMO capacity. This paper shows the difference in the propagation characteristics of dual and single polarized MIMO by conducting experimental measurements. It emerged that the ratios of the eigenvalues of the channel matrix, which represent the channel capacity, depend on cross-polarization discrimination (XPD). It was confirmed that there is a positive correlation between XPD and the eigenvalue ratio for dual polarization and a negative correlation for single polarization respectively.

This paper presents throughput performance of 2 × 2 multiple-input multiple-output (MIMO)-orthogonal frequency division multiplexing (OFDM) using a testbed in line with the Long Term Evolution (LTE) numerology. Then, this paper clarifies the relation among the throughput, modulation schemes and antenna polarization. The field experiments clearly demonstrate that a measured throughput exceeding 200 Mbps can be obtained in a coverage area more than 50% and a throughput of 456.0 Mbps is achieved for a limited number of users even when a maximum Doppler frequency is higher than 125 Hz. In addition, it is found that the single polarization antenna configuration is suitable for 16QAM while the dual polarization antenna configuration is suitable for 64QAM. It is concluded from analysis of the eigenvalue ratio and signal to noise power ratio that the aforementioned trade-off is caused by sensitivity to the eigenvalue ratio depending on the modulation scheme and distribution of the eigenvalue ratio depending on the MIMO antenna configuration.

Base Station (BS) Cooperation technology is expected for the future radio access technology in cellular architecture to improve user throughput at cell edge regions. The effectiveness of the technology highly depends on the radio transmission quality from the BSs in cooperation. The visibility state between a BS and a mobile station (MS) dominates the radio transmission quality at higher frequency bands expected for next generation cellular systems such as IMT-Advanced. In order to evaluate the performance of the BS cooperation technology, this paper establishes a visibility state model between BSs and an MS in typical urban areas with high-rise buildings. The visibility state is simulated using three- dimensional building maps of several areas in Tokyo metropolitan regions. This paper first analyzes the changes in the visibility state between BSs and the MS to establish the visibility state transition model in terms of the number of visible BSs from the MS. The visible state hold distance, which is the metrics of the dynamics in the visibility state transition, is then evaluated. From the results of these analyses, the paper clarifies that the visibility state holding distance could approximately be expressed as an exponential function of the visibility probability between a BS and an MS.

Gravity model has recently proposed for use to predict the movement of cellular mobile subscribers. In this paper, additional data are collected and similar validation process is performed to further our investigation. The results also verify that Gravity model can be used to predict user movements accurately, with the exception that the computed number of trips needs to be scaled by a factor ranging from 0.98 to 0.99. The impact of user mobility on cell planning is then discussed. The optimal design can be obtained through performing multi-period joint optimization.

Gravity model used to model transport flow has proven invaluable in urban planning. Some recent work has proposed to use gravity model to predict user movements in cellular networks, however, there is generally a lack of practical evidence to justify the use of gravity model for such application. In this paper, we briefly summarize the theory of mobility gravity model which is derived based on the maximization of entropy. A simple algorithm which can be used to fit a set of observed trips as close to gravity model as possible is presented. We apply the algorithm to the census data collected in year 2000 from the train stations in the Tokyo metropolitan area. Statistical test is finally performed to validate the suitability in using gravity model to predict user movements for this set of data. Finally, how the model can be used to predict user behavioral change when there is a change in design parameters is discussed.

There are many active mobile users today in trains, especially during the rush hour, hence the potential for mobile communication traffic in trains to rocket in the near future; as applications become diversified and communication tariffs becomes cheaper due to flat rate pricing. The increase in such traffic within a train will mean a mass of accessing users, moving along the railway tracks. In CDMA systems, the signal-to-noise ratio (SNR) of users in the train is greatly influenced by the fluctuation in the number of users accessing the base station. With this in mind, we propose a cell planning scheme using integer linear programming and taking account of the movement of a train containing many accessing users. The scheme minimizes the number of base stations, ensuring the SNR of a user within a target train remains equivalent to or exceeds the required SNR. Finally, simulations using the proposed scheme verify that the SNR of a user in a train for all areas calculated based on optimally-sited base stations reaches a constant value when the interference from other accessing users in a train is much larger than that of those outside the train

A number of wireless systems have recently adopted an adaptive modulation scheme to improve frequency efficiency. Unlike mobile systems, FWA (fixed wireless access) systems including IEEE 802.16 systems used for MAN (metropolitan area network) are allowed to use quasi-millimeter and millimeter bands, but those are vulnerable to rain attenuation. Adaptive modulation scheme is effective in those systems to improve the availability of wireless links. A 26 GHz-band broadband FWA system with a five-level (1024 QAM to BPSK) adaptive modulation scheme was proposed and developed by the authors for field experiments. This paper introduces field experimental results concerning switching thresholds for adaptive modulation controls using the developed FWA system. The numerical results reveal that conventional ideas, in which identical switching thresholds in terms of BER (bit error rate) apply to all modulation methods and switching thresholds maximizing Layer-2 throughput, can neither obtain the maximum TCP (transmission control protocol) throughput nor satisfy the upper limit of the UDP (user datagram protocol) packet loss rate.

In satellite communications systems, in addition to frequency resources, total satellite transmitting power, (i.e., satellite power) is a crucial radio resource, due to the limited capacity of onboard electric generators. To increase system capacity, satellite power constraints have to be taken into account, as well as the frequency constraints. As one approach to this issue, we propose a traffic distribution algorithm using linear programming, which maximizes the accommodated traffic in a multiple-satellite system under satellite power constraints. The algorithms are applicable to all types of satellite systems with multiple satellites (geosynchronous Earth orbit, medium Earth orbit, low Earth orbit, or a combination of any) to increase system capacity with respect to the total amount of traffic accommodated by the systems. Finally, this paper evaluates the performance of the proposed algorithm by distributing traffic demand on the earth's surface to cells irradiated by spot-beams of a satellite and allocating frequency resources to the cells. By using the algorithm that permits power constraints, the ratio of accommodated traffic in a system was improved 12% compared with the case when no power constraints were used.

Broadband fixed wireless access (BFWA) networks with Multipoint-to-Multipoint (MP-MP) systems of mesh topology have recently been considered as a promising approach to realize high quality and high performance networks. Needless to say, it is of prime importance to adopt an effective frequency assignment method suitable for MP-MP systems in order to make maximum use of frequency resources over the system. From the viewpoint of the scalability to network growth, a distributed architecture is preferable compared to a centralized architecture for assigning frequency. This paper firstly identifies constraints on frequency assignment for MP-MP systems and then, proposes an autonomous frequency assignment method. The proposed method dynamically determines an optimum frequency slot for every radio channel in every wireless link, where the number of radio channels is adaptively changed to deal with traffic and link quality changes. In the method, the optimum frequency slot is selected by using an evaluation function. This paper introduces three kinds of evaluation functions and investigates their performances from simulation results.

In satellite communications systems, the total satellite transmitting power (in short, 'satellite power') is one of the important radio resources, as well as spectrum resources, because of the limited capability of onboard electric generators. Therefore, for attaining optimum satellite radio resource utilization, it is not enough to just consider the spectrum resources but also to take account of the satellite power. This paper proposes algorithms using the linear programming for optimally distributing the traffic demand to multiple satellites in a system. The algorithms are applicable to any kind of satellite system with multiple satellites (GEO, MEO, LEO or combination of any) to increase the system capacity in terms of the total amount of traffic accommodated by the system. Numerical examples demonstrate that the proposed algorithms contribute to maximizing the system capacity.