This work studies learning-aided indoor fingerprint positioning (FP) for multiple-input multiple-output (MIMO) orthogonal frequency-division multiplexing (OFDM) systems, by exploiting their channel response matrices (CRMs). The existing learning methods of CRM-based FP mostly suffer from two main limitations: i) they require retraining when CRM dimensions change, which severely limits their practicality since the number of antennas and subcarriers may vary dynamically during practical implementation; ii) they overlook the temporal correlations in CRMs along the user trajectory, failing to exploit this information to improve positioning accuracy. Motivated by these limitations, we propose a novel deep learning-assisted FP method, named adaptive spatio-temporal neural network (A-STNN). This method consists of two key components: i) an adaptive mechanism that transforms spatial-frequency domain CRMs (SFCRMs) of arbitrary dimensions into truncated angle-delay domain CRMs (T-ADCRMs) with a unified dimension, enabling the neural network to handle varying antenna and subcarrier configurations; and ii) an STNN that exploits attention mechanisms to jointly extract spatial and temporal features embedded in T-ADCRM sequences along the user’s trajectory, thus improving positioning accuracy. Upon examination of simulation and measurement datasets, A-STNN achieves a prominent improvement in positioning accuracy over existing FP methods, in addition to its unique generalization capability across different antenna and subcarrier settings.

Conventional wireless channel measurements and modeling typically study channels via discrete spatial sampling. As the sixth-generation (6G) wireless communication places higher demands on the accuracy of channel state information, this discrete approximation becomes insufficient and motivates research on continuous-space channels. In this work, 28 GHz indoor continuous-space channel measurements are conducted, and key channel characteristics are analyzed. Based on the analysis, the necessity of continuous-space channel research is validated, and the role of channel maps is demonstrated. Furthermore, a continuous-space channel map construction method using the Graph SAmple and aggreGatE (GraphSAGE) algorithm is proposed. By the learning on spatial aggregation function rather than performing a passive weighted sum, the proposed GraphSAGE-based map construction method can reduce the performance bias caused by discrete spatial sampling and recover the continuous-space channels accurately. Continuous-space measurements are used as benchmarks to compare the performance of the GraphSAGE-based channel map. Extensive experiments confirm the superiority of the proposed GraphSAGE-based method over existing artificial intelligence (AI) algorithms, providing a robust approach for the 6G continuous-space channel map construction.

The rapid increase in the number of multiple-input multiple-output (MIMO) antennas towards 6G cause a massive burden on channel state information (CSI) feedback, and thus necessitates the development of CSI compression techniques. Recent research investigated autoencoder (AE) to reduce compression loss, which can eliminate the need for model assumptions in the classic compression approaches. However, the existing AE-based methods are mostly trained and used for a single specific scenario, lacking the generalization capability across different scenarios. In this paper, we propose a novel CSI compression method named attention-infused autoencoder network (AiANet), which can cope with both indoor and outdoor scenarios with the same trained model. Specifically, AiANet employs an adaptive mechanism to learn the channel-wise and spatial features of CSI, in order to address the distinct channel characteristics in different scenarios. Also, a locally-aware self-attention (LASA) module is developed to capture the global and local features of CSI, so as to enhance compression capability for local details. For a fair comparison, AiANet and the benchmark methods are trained with the same COST2100 Indoor and Outdoor datasets. Results show that when tested in the same scenarios, the proposed approach can always outperform state-of-the-art learning methods such as ACRNet, with a normalized mean squared error (NMSE) improvement of up to 3.4 dB. When evaluated in two unseen DeepMIMO scenarios (urban canyon and office floor), AiANet demonstrates superior cross-scenario generalization capability over the benchmark methods, with a NMSE increase of up to 3.18 dB. In addition, AiANet requires a model complexity comparable to ACRNet, with a median end-to-end inference latency below 4 ms on edge devices.

This work investigates deep learning (DL)-enabled fingerprint positioning (FP) in a single access point (AP) indoor scenario. Compared to existing FP methods that mostly rely on multiple APs, single AP-based FP faces the tricky issue of substantially reduced spatial information, making those methods less effective or even invalid. To address this issue, we propose a novel DL-enabled FP method to exploit the spatio-temporal features of the angular domain channel matrices (ADCMs) along the user trajectory in single-AP multiple-input multiple-output (MIMO) systems. Specifically, multi-layer convolutional neural networks (CNNs) are designed to extract the spatial features of each ADCM, which reflects position-specific angular distribution of signal propagation. Then a long short-term memory (LSTM) with channel attention (CA) is utilized to capture the temporal dependencies of the sequential ADCMs. Furthermore, a data augmentation method by flipping the ADCM sequence is proposed to reduce the reliance on labor-intensive large-scale data collection for model training. The simulation results show that, in the same single-AP indoor scenario, the proposed approach can significantly increase the positioning accuracy over existing CNNbased and LSTM-based methods, such as ABPN and SIABR, with an improvement of up to 41.7% and 70.8%, respectively.

The explosive growth of the sixth-generation ( 6 G ) wireless system necessitates efficient base station (BS) deployment to balance coverage, data rates, and economic costs. In this paper, we propose a novel BS deployment scheme that jointly optimizes the number and locations of the deployed BSs, considering limited BS throughputs and practical non-uniform user distributions. A variant multi-dimensional knapsack problem is formulated and then solved by proposing a novel depth-limited backtracking dynamic programming (DLB-DP) algorithm with a BS-user association (BSUA) algorithm. We compare the proposed DLB-DP algorithm with three state-of-the-art benchmark algorithms in a hotspot scenario. The results demonstrate that the proposed algorithm outperforms the considered alternatives in terms of coverage, data rates, and robustness under varying user densities.

The ever growing number of antennas in massive multiple-input multiple-output (MIMO) systems have significantly burdened the demand for the feedback of channel state information (CSI). This drives the research on CSI compression for massive MIMO. Recent development of autoencoder-based methods has proven to break the limit of conventional data compression methods in terms of both accuracy and computational complexity. However, like conventional methods, those autoencoder-based methods are trained for certain channel scenarios (such as indoor and outdoor) and would require a dedicated model for each scenario, limiting the practicability. It is challenging to develop a unified model to compress CSI across different channel scenarios, as their properties are diverse. In this paper, such a model is proposed for the first time, which is named attention-infused autoencoder network (AiANet). A dual attention mechanism is developed to capture the spatial and channel features of distinctive CSI. Multi-resolution convolutions are also employed to enhance the ability of feature extraction. Simulation results demonstrate that AiANet can substantially outperform ACRNet, which is a state-of-the-art auto encoder model. In terms of normalized mean squared error (NMSE), the proposed method can achieve an improvement of up to 3.69 dB in indoor and 1.26 dB in outdoor.

Minimizing energy consumption is critical for the long-range (LoRa) Internet of Things (IoT) networks, to extend the battery lifetime of end devices (EDs) while reducing the maintenance cost. To overcome the excessive computational complexity required by the traditional optimization methods, learning approaches such as deep neural network (DNN) and reinforcement learning (RL) have been researched in wireless networks, where star topologies are widely employed. In contrast, LoRa usually deploys relays to assist the connection between the EDs and the gateway (GW), leading to a much more complex network topology. Consequently, DNN and RL would become less effective in LoRa, since those methods are difficult to learn the complex network topology constructed by LoRa. In this paper, we propose a learning method based on graph neural network (GNN), which is known for its prominent ability to capture and represent intricate graph-structural dependencies, to tackle the energy optimization problem for relay-assisted LoRa. Specifically, the multi-hop LoRa network is modeled as a directed graph, with channel state information (CSI) defined as node features, while the spreading factor and transmission power are deemed labels. A hierarchical message aggregation mechanism is proposed to effectively capture the multi-hop structural dependencies, followed by the process of inductive learning. Results show that against conventional optimization algorithms, the proposed method can achieve near-optimal energy consumption with a gap of 10% to 16%, while reducing the runtime by about six orders of magnitude. Compared to DNN, the GNN-based model can provide an energy saving of up to 32%, at a similar level of inference time.

Energy optimization is crucial for extending battery life and reducing maintenance costs in long-range (LoRa) Internet of Things (IoT) networks. Traditional optimization methods usually need excessive computational complexity, limiting the practicability. This drives the recent development of machine learning (ML)-based optimization, such as deep neural networks (DNNs) and reinforcement learning (RL), in wireless local area networks (WLANs). However, compared to WLANs, LoRa owns a more complicated network topology due to the engagement of multi-hop, which is difficult for the existing ML methods to learn. Motivated by this, we propose a topology-aware graph neural network (GNN) learning method, which is specially tailored to tackle the energy optimization problem in multi-hop LoRa networks. By leveraging each node’s topological position to adaptively determine the optimal message-passing depth, the model better integrates the neighborhood information with node feature representations, enhancing the prediction of transmission parameter and overall energy efficiency. Also, a closed-form model of collision probability is derived for the nodes in LoRa, to measure the energy consumption due to retransmissions. Simulation results show that against traditional optimization methods such as game theory, the proposed method can reduce the runtime by five orders of magnitude, with an energy optimization gap below 13%. Compared to existing GNN-based methods, it achieves up to 50% lower energy consumption 20% fewer outage probability, at a similar level of inference time.

Hybrid light fidelity (LiFi) and wireless fidelity (WiFi) networks (HLWNets) are an emerging indoor wireless communication paradigm, which combines the complementary advantages of LiFi and WiFi. Meanwhile, load balancing (LB) becomes an essential and critical challenge, due to the nature of hybrid networks. The existing LB methods are mostly network-centric, relying on a central unit to make a solution for the users all at once. Consequently, the solution needs to be updated for all users at the same pace, regardless of their moving status. This would affect the network performance in two aspects: 1) a lower update frequency would compromise the connectivity of fast-moving users; 2) a higher update frequency would cause unnecessary handovers as well as hefty feedback costs for slow-moving users. Motivated by this, we investigate user-centric LB so that users can update their solutions at different paces. The research is developed upon our previous work on adaptive target-condition neural network (ATCNN), which carries out LB for individual users in quasi-static channels. In this paper, a deep neural network (DNN) model is designed to enable an adaptive update interval for each individual user. This new model is termed as mobility-supporting neural network (MSNN). Associating MSNN with ATCNN, a user-centric LB framework named mobility-supporting ATCNN (MS-ATCNN) is proposed to handle resource management and mobility management simultaneously. Results show that at the same level of average update interval, MS-ATCNN can achieve a network throughput up to 215% higher than conventional LB methods such as game theory (GT), especially for a larger number of users. In addition, MS-ATCNN costs an ultra-low inference time in sub-milliseconds, which is two to three orders of magnitude lower than the GT baseline.

This study aims to address key technical issues in the segmentation of Cerebral MicroBleeds (CMBs) based on Low-Resolution (LR) Magnetic Resonance Imaging (MRI) data. There are two challenges in this task. First, the CMB lesions are typically small in size and easily confused with various mimics. Second, anisotropy becomes more prominent and adverse in LR MRI sequences than HR sequences. To address these issues, we propose a Progressive Learning based Knowledge Distillation method. This method progressively transfers knowledge from HR models to their LR counterparts, thereby minimizing the occurrence of false positives attributable to noise from Super-Resolution. To further eliminate the influence of anisotropy, an encoding-enhanced network, called E2U-Net, is proposed in this paper. It can effectively capture anisotropic information and mitigates potential feature loss. The experimental results on multiple publicly accessible CMBs datasets demonstrated the superiority of our proposed approach over existing deep-learning methods.

Load balancing (LB) is a key challenge in hybrid light fidelity (LiFi) and wireless fidelity (WiFi) networks (HLWNets), due to the nature of heterogeneous access points (APs). Machine learning has the potential to provide a complexity-friendly LB solution with near-optimal network performance, at the cost of a non-trivial training process. The state-of-the-art learning-aided LB methods require retraining when the network environment (particularly the user number) changes, significantly limiting their practicability. In this paper a novel deep neural network (DNN) structure, named adaptive target-condition neural network (A-TCNN), is proposed to tackle the LB issue for a varying number of users, without the need for retraining. Unlike the existing LB methods conducting AP selection for all users together, the new method performs AP selection for a single target user, upon the condition of other users. Also, A-TCNN involves an adaptive mechanism which maps any smaller number of users to a preset number by splitting the users’ data rate requirements, without affecting the AP selection result for the target user. Once trained, A-TCNN can be used for any user numbers not exceeding the maximum user number that the network can support. Results show that apart from the adaptiveness to a varying user number, A-TCNN provides a higher network throughput (up to 45%) than the conventional DNN in most cases, especially for a larger scale of network. In terms of computational complexity, A-TCNN can achieve a sub-millisecond level runtime, which is 2 orders of magnitude lower than fuzzy logic and 3 orders of magnitude lower than game theory.

Non-reciprocal antenna beam patterns are promising to be utilized in asymmetric massive multiple-input multiple-output (MIMO) systems for future sixth-generation communications. The inconsistency of uplink (UL) and downlink (DL) channels makes channel modeling in this scenario challenging. In this paper, a novel geometry-based stochastic model (GBSM) is proposed for non-reciprocal terahertz (THz) channels. A directional effective scatterer generation algorithm is designed to depict the inconsistency of bidirectional propagation conditions. The correlation function between UL and DL is derived and analyzed, which validates the ability to characterize the non-reciprocal channels. To mimic THz propagation features, molecular absorption and diffuse scattering are introduced to the model, which is verified by measured data. In addition, the non-stationarities in space, time, and frequency domains are characterized, respectively. Statistical properties are compared between analytical and simulation results, and good agreements are shown. Finally, the accuracy of the model is verified by comparing with the ray tracing data.

This work studies the impact of channel blockage caused by human bodies in visible light communications (VLC), which is one of the key wireless technologies in 6G. As operating on the optical spectra, VLC channels can be blocked by opaque objects, especially by human users themselves. This human blockage has a nonnegligible impact on the performance of VLC. The current literature mostly considers human bodies as shaped objects to evaluate the numerical performance of VLC. In this paper, the coverage probability of line-of-sight (LoS) signal-to-noise ratio (SNR) under the impact of human blockage is derived in a closed-form expression and validated against simulations. The impact of human blockage on non line-of-sight (NLoS) is also analysed in specific room environments. Results show that increasing the distance between body and device from 10cm to 30cm can nearly halve the blockage probability, from 22% to 12%. Also, a shorter separation between VLC APs can effectively mitigate the LoS SNR degradation caused by human blockage, while the NLoS SNR degradation is irrelevant to the APs' separation.

This work studies the quality of service (QoS) performance of wireless networks that integrate light fidelity (LiFi) and wireless fidelity (WiFi). While the hybrid network is potential for improving network capacity, load balancing becomes essential and challenging due to the nature of heterogeneous access points (APs). A number of studies have been conducted to address this issue, focusing on maximising the network capacity with user fairness constraints. However, in practice, QoS metrics including packet loss ratio and latency are important to network services. In this paper, QoS-driven load balancing is studied for hybrid LiFi and WiFi networks (HLWNets) in two scenarios: single-AP association (SA) and multi-AP association (MA). In each case, an optimisation problem is formulated to minimise the packet loss ratio and latency, and a low-complexity iterative algorithm is proposed to solve the problem. Results show that the novel methods, especially MA, can effectively balance the traffic loads among the APs and improve the QoS performance. In addition, the more subflows the better performance MA provides. Targeting the same level of QoS, MA can achieve a system throughput up to 160% higher than the signal strength strategy and 130% higher than the proportional fairness load balancing.

Light-fidelity (LiFi) is recognised as a promising technology for next generation wireless access networks. However, limited research efforts have been spent on the uplink (UL) transmission system in LiFi networks. In this article, a wireless infrared (IR)-based LiFi UL system is investigated. In particular, we focus on the performance of a single static user under the influence of random device orientation and link blockage. Simulations and mathematical analysis have been used to evaluate the UL system performance. The analytical expressions for the UL optical wireless channel and signal-to-noise ratio (SNR) statistics under the effects of random device orientation and link blockage are derived. The results show that the effects of random orientation and link blockage may lead to a decrease in coverage probability by 10% - 40% with various SNR thresholds.

Combining the high transmission speed of light fidelity (LiFi) and the ubiquitous coverage of wireless fidelity (WiFi), hybrid LiFi and WiFi network (HLWNet) has attracted intensive research interest. While such a network can boost the system capacity, it faces a challenging issue in handover since the coverage areas of LiFi and WiFi completely overlap each other. Also, LiFi has a relatively short coverage range from a single access point (AP). As a result, HLWNets are susceptible to frequent handovers, in terms of both handovers between LiFi and WiFi and handovers between two LiFi APs. Therefore handover skipping was introduced, which enables handovers between two non-adjacent APs. Conventional handover skipping schemes rely on knowledge about the user's trajectory, and are difficult to implement in practice. Alternatively, the rate of change in reference signal received power (RSRP) can be used to reflect the user's moving direction. Based on this, an adaptive handover scheme is proposed in this paper, which varies the user's network preference according to the user's speed. Results show that in comparison with the standard and trajectory-based schemes, the proposed approach can improve user throughput by up to about 120% and 30%, respectively.

Automatic Speech Recognition (ASR) is greatly developed in recent years, which expedites many applications on other fields. For the ASR research, speech corpus is always an essential foundation, especially for the vertical industry, such as Air Traffic Control (ATC). There are some speech corpora for common applications, public or paid. However, for the ATC, it is difficult to collect raw speeches from real systems due to safety issues. More importantly, for a supervised learning task like ASR, annotating the transcription is a more laborious work, which hugely restricts the prospect of ASR application. In this paper, a multilingual speech corpus (ATCSpeech) from real ATC systems, including accented Mandarin Chinese and English, is built and released to encourage the non-commercial ASR research in ATC domain. The corpus is detailly introduced from the perspective of data amount, speaker gender and role, speech quality and other attributions. In addition, the performance of our baseline ASR models is also reported. A community edition for our speech database can be applied and used under a special contrast. To our best knowledge, this is the first work that aims at building a real and multilingual ASR corpus for the air traffic related research.

Combining the high-speed data transmission of light fidelity (LiFi) and the ubiquitous coverage of wireless fidelity (WiFi), hybrid LiFi and WiFi networks (HLWNets) are recently proposed to improve the system capacity of indoor wireless communications. Meanwhile, load balancing becomes a challenging issue due to a complete overlap between the coverage areas of LiFi and WiFi. User mobility and light-path blockages further complicate the process of load balancing, since the decision for a horizontal or a vertical handover in a mobile environment with ultra-small cells is non-trivial. These issues are managed separately in most conventional methods, which might cause frequent handovers and compromise throughput. A few studies address these issues jointly for selecting access points at each time instant but require excessive computational complexity. In this paper, a joint optimisation problem is formulated to determine a network-level selection for each user over a period of time. A novel algorithm based on fuzzy logic is also proposed to reduce the computational complexity that is required to solve the optimisation problem. Results show that compared to the conventional method, the proposed approach can improve system throughput by up to 68%, while achieving very low computational complexity.

The increasing number of mobile devices challenges the current radio frequency (RF) networks, e.g. wireless fidelity (WiFi) networks. Light Fidelity (LiFi) is considered as a promising complementary technology, which operates within the visible light spectrum and infrared spectrum. In an indoor scenario, a hybrid LiFi/WiFi network (HLWN) provides a potential solution to future wireless communications where LiFi augments WiFi in providing ultra-high speed and low latency wireless connectivity. In this paper, dynamic load balancing (LB) with handover in HLWNs is studied. The orientation-based random waypoint (ORWP) mobility model is considered to provide a more realistic framework to evaluate the performance of HLWNs. Based on the low-pass filtering effect of the LiFi channel, we firstly propose an orthogonal frequency division multiplexing access (OFDMA)-based resource allocation (RA) method in LiFi systems. Also, an enhanced evolutionary game theory (EGT)-based LB scheme with handover in HLWNs is proposed. In the EGT scheme, each user adapts their strategy to improve the payoff until LB is achieved across LiFi and WiFi. Then, the LiFi system uses the proposed OFDMA-based RA method while the WiFi system applies the carrier sense multiple access with collision detection (CSMA/CA). Simulation results show that in the LiFi system the OFDMA-based RA scheme outperforms the time division multiple access (TDMA) scheme in terms of both user data rate and fairness. Regarding LB in HLWNs, the proposed EGT scheme can achieve a remarkable enhancement in throughput compared to benchmark schemes, such as hard threshold (HT) scheme and random access point assignment (RAA) scheme.

A deep learning-based safety monitoring framework for air traffic control (ATC) systems is proposed in this paper to reduce human errors and relieve the controllers' workload by regulating the controlling procedure, eliminating communication misunderstanding, monitoring flight conformance, and detecting potential conflicts. The framework comprises automatic speech recognition (ASR), controlling intent inference (CII), and control safety monitoring (CSM) subsystems. The pipeline of the proposed framework can be described as follows: the ASR subsystem translates the pilot-controller voice communications (PCVCs) into texts, which are then converted to the predefined data structure by the CII subsystem. Three types of air traffic safety measures, including repetition check, flight conformance verification, and potential conflict detection, are finally validated by the CSM subsystem. An improved end-to-end ASR model with convolutional, bidirectional long short-term memory (BLSTM) and fully connected (FC) layers is trained using the connectionist temporal classification loss function. The BLSTM and FC combined CII model is designed to infer the controlling intent and slot filling. A language model is also trained in this subsystem to improve the overall performance of the framework. After converting the PCVCs to ATC data, the CSM subsystem checks the given safety monitoring tasks and sends warnings to the current system. The experimental results show that the proposed ASR model obtains a better performance than that of other approaches, and the tasks in the CII subsystem are fulfilled with a high classification precision. The CSM subsystem is also tested to confirm its safety monitoring function by playing back the data and several simulated instructions. To the best of our knowledge, this is pioneering work in the safety monitoring of flight control by recognizing the PCVCs with deep learning-based methods.

Light fidelity (LiFi) is a relatively new wireless communication technology that exploits the optical spectrum. Compared to wireless fidelity (WiFi), LiFi is densely deployed with each access point (AP) covering an area only a few meters in diameter. Also, LiFi users are susceptible to intermittent light-path blockages. Meanwhile, the user is served by a single AP in conventional LiFi systems. This would cause frequent handovers for LiFi users, resulting in a degradation in quality of service. In this paper, parallel transmission is investigated for LiFi, which is named PT-LiFi. With a delicate design of the transmitter and the receiver, PT-LiFi enables multiple LiFi APs to serve the user simultaneously. Particularly, data transmission continues without interruption when the user is losing connectivity to some of the connected APs. Resource allocation is studied for the PT-LiFi system, and a novel load balancing method is proposed to jointly allocate resource across the APs. Results show PT-LiFi can make efficient use of the densely deployed LiFi APs and provide a flexible way of load balancing. Compared with a conventional LiFi system, the proposed method can increase user throughput by up to 150% and improve user fairness by up to 15%.

This work investigates handover in hybrid light fidelity (LiFi) and wireless fidelity (WiFi) networks (HLWNets). In such a network, the handover process becomes challenging due to two main factors: i) the relatively short coverage range of a single access point (AP), and ii) the largely overlapping coverage areas of different networks. As a result, HLWNets are susceptible to frequent handovers. To reduce the handover rate, the concept of handover skipping (HS) was introduced, which enables handovers between non-adjacent APs. However, conventional HS methods rely on knowledge about the user's trajectory, which is not readily available at the AP. In this paper, a novel HS scheme is proposed on the basis of reference signal received power (RSRP) and its rate of change, with an adaptive network preference adopted. Since RSRP is commonly used in the existing handover schemes, the proposed method requires no additional signalling between the user and the AP. Simulation results show that the new method can effectively reduce unnecessary handovers, especially those between LiFi and WiFi. Compared to the standard and trajectory-based handover schemes, the proposed method improves network throughput by up to about 120% and 30%, respectively.

This paper compares an intensity modulation/direct detection (IM/DD) light fidelity (LiFi) system using direct current biased optical orthogonal frequency-division multiplexing (DCO-OFDM) and a coherent LiFi system using homodyne detection. The performance metrics of bit error ratio (BER) and the power efficiency are compared between the two systems. In addition, the channel direct current (DC) gain characteristics of the two systems are investigated. Simulation results show that the coherent system achieves a better BER performance than the IM/DD system, but the transmitter and the receiver of the coherent system need to be perfectly matched. Also, within the same coverage area, the coherent system is more power efficient than the IM/DD system.

Light fidelity (LiFi) is a recently proposed wireless communication technology that employs the light wave as a medium to transmit data. With the coverage range of a few meters, LiFi access points (APs) comprise an ultra-dense attocell network. As a result, users may quickly traverse the coverage area of one LiFi AP, even with a moderate speed. This would cause frequent handovers and a subsequent decrease in the user's throughput. Handover skipping (HS), which enables handovers between two non-adjacent APs, is able to reduce the handover rate. Conventional HS methods rely on the user's trajectory, which is difficult for APs to acquire in practice. In this paper, we propose a novel HS scheme on the basis of received signal strength (RSS). Specifically, the rate of change in RSS is exploited to indicate whether the user is moving towards a certain AP. Since RSS is already used in standard handover schemes, the proposed method does not require extra feedback. Results show that compared to the signal strength strategy, the proposed HS approach can effectively reduce the handover rate, especially for fast- moving users. With respect to the user's throughput, the new method can achieve an increase of up to 67% against the signal strength strategy.

Light Fidelity (LiFi) is an emerging technology for future high-speed indoor wireless communications. Co-channel interference (CCI) caused by the dense deployment of LiFi access points (APs) can be effectively mitigated by using angle diversity receivers (ADRs). ADRs require signal combining where the combining weights depend on the selection of serving APs. In this paper, the AP selection (APS) strategy considering handover is studied. A novel APS scheme based on evolutionary game theory (EGT) is proposed for the LiFi network using ADRs. The performance of the proposed scheme is comprehensively analysed and compared with the APS scheme based on signal strength strategy (SSS). The result shows that, in terms of ADRs with SBC/MRC, the EGT-based APS scheme achieves more than 5% improvement in quality of service (QoS) compared with the SSS-based APS scheme. With the sub-optimum weights of maximum ratio combining (MRC) for ADRs, the EGT-MRC scheme can achieve more than 20 Mbps data rate improvement compared with LiFi systems using single photodiode (PD) receiver.

Hybrid light fidelity (LiFi) and wireless fidelity (WiFi) networks (HLWNets) have been recently proposed as a promising technique for indoor wireless communications. Such a network combines the high-speed transmission of LiFi and the ubiquitous coverage of WiFi. Meanwhile, in addition to user mobility, intermittent light-path blockages make the handover issue intricate in HLWNets. Also, load balancing (LB) is necessary because the coverage areas of LiFi and WiFi networks are completely overlapped. This further increases the difficulty in handover as the decision for a vertical or a horizontal handover in a mobile environment with ultra-small cells is non-trival. Aimed at providing an optimal solution to the access point selection (APS) with given channel state information (CSI), the conventional LB method might cause frequent and unnecessary handovers. In this paper, a joint optimisation problem is formulated to simultaneously consider LB and handover in HLWNets. Results show that compared with the conventional LB method, the proposed method can improve the system throughput by up to 60%.

Channel adaptive signaling, which is based on feedback, can result in almost any performance metric enhancement. Unlike the radio frequency channel, the optical wireless communication (OWC) channel is relatively deterministic. This feature of OWC channels enables a potential improvement of the bidirectional user throughput by reducing the amount of feedback. Light-Fidelity (LiFi) is a subset of OWCs, and it is a bidirectional, high-speed, and fully networked wireless communication technology where visible light and infrared are used in downlink and uplink, respectively. In this paper, two techniques for reducing the amount of feedback in LiFi cellular networks are proposed: 1) limited-content feedback scheme based on reducing the content of feedback information and 2) limited-frequency feedback scheme based on the update interval. Furthermore, based on the random waypoint mobility model, the optimum update interval, which provides maximum bidirectional user equipment throughput, has been derived. Results show that the proposed schemes can achieve better average overall throughput compared with the benchmark one-bit feedback and full-feedback mechanisms.

Spatial modulation (SM) is a unique single-stream, multiple-input multiple-output (MIMO) transmission technique. One key property of the SM is that a single transmit antenna is activated at any given time, which completely avoids inter-channel interference in a single-user scenario. In the context of multi-user cellular networks, inter-cell interference (ICI) mitigation becomes the main challenge for the SM. Network MIMO is a technique that employs precoding to enable ICI cancellation among neighbouring cells. However, the cell-edge users might still experience a high level of the ICI due to the significant channel attenuation. In this paper, we propose a novel cooperative scheme based on the SM, named cooperative SM (CoSM). The basic concept is to reschedule the transmit antennas of multiple base stations for multiple co-channel users, so as to maximize the antenna diversity gain. Unlike the traditional antenna selection for a single user, the involvement of multiple users might cause conflict due to the selection of the same antenna. An antenna rescheduling scheme (ARS) is thus proposed to address this issue. The performance of the ARS is theoretically analyzed, and closed-form expressions are derived for the probability distribution of the received SNR. Also, a novel three-tier cellular architecture is proposed to accommodate the CoSM within the context of the network MIMO. Results show that the CoSM can improve signal-to-interference-plus-noise ratio by up to 4 dB over the network MIMO. In addition, compared with spatial multiplexing using the same ARS, the CoSM can halve the energy consumption while achieving the same bit error rate.

Light Fidelity (LiFi) is a recently proposed technology that combines illumination and high speed wireless communication using light emitting diodes (LEDs). Unlike radio frequency (RF) channels, LiFi channels do not exhibit fading characteristics as the detector size is much larger than the wavelength. Another distinguishing feature is that LiFi channels are mainly affected by the low-pass filtering effect caused by LEDs, while the multipath effect is inconspicuous. Therefore, unlike in RF systems, LiFi systems can hardly achieve the conventional multi-user diversity gain when using orthogonal frequency division multiplexing access (OFDMA). Due to the LED characteristics, users with high direct current (DC) signal-to-noise ratio (SNR) may be able to use a large modulation bandwidth in LiFi systems. This means that users with good channel conditions can be more likely to utilise high-frequency resources. In this paper, resource allocation (RA) in OFDMAbased LiFi systems is investigated and an optimal RA scheme and a low- complexity RA scheme are proposed. Simulation results show that by efficiently using high- frequency bandwidth resources, the RA schemes in OFDMA systems outperform those in TDMA systems in terms of both data rate and user fairness. Also, the low-complexity RA scheme is able to achieve nearoptimal performance at a reduction of 90% in computational complexity.

Recently a promising concept of hybrid networks based on light fidelity (LiFi) and wireless fidelity (WiFi) emerged. The idea is to combine ultra-small cell LiFi networks with ubiquitous coverage radio frequency (RF) communication systems. In such a hybrid network, WiFi access points (APs) serve a relatively large coverage area with limited bandwidth, and are thus susceptible to traffic overload. This issue is magnified with an increasing number of users because of the inefficient medium access control (MAC) in WiFi systems. LiFi can alleviate this issue by providing additional capacity. LiFi cells, however, have a limited coverage and this could result in significant handover overhead. A conventional load balancing (LB) method optimises the network throughput when the signal-to-noise ratio (SNR) of each user is known and fixed. Although this method delivers maximum throughput at a given time instance, it fails to consider the throughput loss due to handover, especially in an indoor scenario where users may frequently switch between APs. Taking the handover overhead into account, in this paper we propose a novel LB method that focuses on optimising the network throughput over a period of time. Simulation results show that the proposed method can increase the system throughput by up to 70% compared to existing LB methods.

Light fidelity (LiFi) uses light emitting diodes (LEDs) for high-speed wireless communications. Since an LED lamp covers a small area, a LiFi system with multiple access points (APs) can offer a significantly high spatial throughput. However, the spatial distribution of data rates achieved by LiFi fluctuates because users experience inter-cell interference from neighboring LiFi APs. In order to guarantee a quality of service (QoS) for all users in the network, an RF network is considered as an additional wireless networking layer. This hybrid LiFi/RF network enables users with low levels of optical signals to achieve the desired QoS by migrating to the RF network. With regard to moving users, the hybrid LiFi/RF system dynamically allocates either a LiFi AP or an RF AP to users based on their channel state information. In this paper, a dynamic load balancing scheme is proposed, which considers the handover overhead in order to improve the overall system throughput. Joint optimization algorithm (JOA) and separate optimization algorithm (SOA), which jointly and separately optimize the AP assignment and resource allocation, respectively, are proposed. Simulation results show that SOA can offer a better performance/complexity tradeoff than JOA for system load balancing.

Hybrid light fidelity (Li-Fi) and wireless fidelity (Wi-Fi) networks are an emerging technology for future indoor wireless communications. This hybrid network combines the high-speed data transmission offered by visible light communication and the ubiquitous coverage of radio-frequency techniques. While a hybrid network can improve the system throughput and users' experience, it also challenges the process of access point selection (APS) due to the mixture of heterogeneous access points. In this paper, the differences between homogeneous and heterogeneous networks regarding APS are discussed, and a two-stage APS method is proposed for hybrid Li-Fi/Wi-Fi networks. In the first stage, a fuzzy logic system is developed to determine the users that should be connected to Wi-Fi. In the second stage, the remaining users are assigned in the environment of a homogeneous Li-Fi network. Compared with the optimisation method, the proposed method achieves a close-to-optimal throughput at significantly reduced complexity. Simulation results also show that our method greatly improves the system throughput over the conventional methods, such as the signal strength strategy and load balancing, at slightly increased complexity.

Light Fidelity (LiFi) is a recently proposed technology that uses 300 THz visible light spectrum for high speed wireless communications as well as providing illumination. Basically, a LiFi access point (AP) covers only a few square meters, enabling a dense deployment of LiFi APs to improve the network throughput. However, channel blockage and shadowing in conjunction with inter-cell interference compromise the connectivity and system throughput of LiFi networks. In this paper, a network structure that combines LiFi with the conventional radio frequency (RF) system is considered. Users experiencing strong blockages can be switched to the RF system to achieve a higher data rate. In this paper, blockages, random orientation of LiFi receivers, and the user data rate requirement are characterised to model a practical communication scenario. A novel load balancing (LB) scheme based on evolutionary game theory is proposed for hybrid LiFi/RF networks. The performance of the proposed scheme is comprehensively analyzed. Results show that compared to state-of-the-art LB algorithms, the proposed scheme greatly improves user satisfaction levels at reduced computational complexity. Also, an optimal orientation of LiFi receivers and blockage density in hybrid networks would maximize the users quality of service.

Light Fidelity (LiFi) is a recently proposed technology that combines illumination and high speed wireless communication using light emitting diodes (LEDs). Since the used electromagnetic spectrum does not overlap with the radio frequency (RF) spectrum, a small cell LiFi attocell network can be added to the conventional RF network as an additional networking layer in order to mitigate the data traffic bottlenecks in high density environments. In such a hybrid LiFi/RF network where the LiFi attocell covers a few square meters, user movement may prompt frequent handovers, and the handover overhead would degrade the system throughput. The goal is to reduce the handover overhead by appropriately assigning users to either the RF or the LiFi access point (AP). In this study, a fuzzy logic (FL) based dynamic handover scheme is proposed. This FL scheme uses not only the channel state information (CSI), but also the user speed and desired data rate to determine whether a handover needs to be prompted. Simulation shows that the proposed scheme outperforms the conventional handover algorithms, and the performance improvement is approximately 40% in terms of both data rate and user satisfaction level.

Light Fidelity (LiFi) uses light emitting diodes (LEDs) for high speed wireless communications. Since employing a different range of the electromagnetic spectrum from radio frequency (RF) communications, LiFi can significantly alleviate the traffic bottlenecks in high density RF scenarios, typically present in an indoor environment. Hence, a combination of LiFi and RF networks becomes a promising candidate for future indoor wireless communications. In a practical indoor scenario, the optical interference from neighbouring LiFi access points (APs) and the blockages of line-of-sight (LoS) optical channels induced by people and objects are the main factors that cause significant optical channel variations. In this study, the effect of these two factors on the system throughput of a hybrid LiFi/RF network is investigated. In order to offer a fair comparison, area data rate, which is defined as the system throughput in a unit area, is used for performance evaluation. The simulation shows that there is an optimal distance between two neighbouring LiFi APs to achieve the highest area data rate. In addition, the area data rate increases with the density of blockages when the blockage density is below a certain threshold.

The cellular Light-Fidelity (Li-Fi) network is considered as a promising approach for high speed indoor data access. The conventional metric, based on which an access point (AP) is selected for each user, is signal strength. This metric offers the best channel quality for each user but does not guarantee the achievable data rate, since the resource of an AP is limited. In this paper, we propose a new metric for AP selection to improve the load balancing among APs by considering both the received signal-to-interference-plus-noise-ratio (SINR) and the traffic of AP. The orientation of mobile stations (MS) is also taken into account and its effect on users' performance is evaluated. In reality, receivers have random angles with the coordinate axes. We consider three standard angles similar to those used in mobile devices to model the device orientation. Based on this model, the effect of arbitrary orientation on user's throughput and satisfaction is investigated. Simulation results show that when the orientation of users is considered, the proposed AP selection metric outperforms the conventional metric.

This work studies the issue of access point selection (APS) in a hybrid wireless network accommodating visible light communication (VLC) and radio-frequency (RF) technologies. A hybrid network constructs multiple layers of coverage, and thus a user possibly acquires a high level of receiving signal strength (RSS) from more than one access point (AP). This fact undermines the effectiveness of conventional RSS-based APS methods. Another challenging factor is the dissimilarity between heterogeneous APs in terms of coverage area and capacity. In general, RF offers larger coverage area but lower capacity than VLC, and therefore the RF system is susceptible to overload. Although the issue of APS can be formulated as an optimisation problem, this approach requires a prohibitive amount of processing power. In this paper a two-stage APS method is proposed on the basis of fuzzy logic, with very low computational complexity. The new method first determines the users that should be connected to the RF system, and then assigns the remaining users as if in a stand-alone VLC network. Results show that when achieving the same amount of throughput, the proposed method can support up to 25% and 56% more users than the load balancing (LB) and signal strength strategy (SSS) methods, respectively.

In this paper, resource allocation (RA) and load balancing (LB) are jointly investigated in a visible light communication (VLC) system. With RA, the resource of access points (APs) is intelligently allocated to their users, while LB enables a user to select an AP other than the one offering the highest signal strength. To date, most research on VLC has treated those two issues separately, leading to an insufficient usage of system resource. In this study, a novel concept termed bidirectional allocation game is proposed, where allocating the AP resource to users and assigning users to APs are carried out together. Also, a fuzzy logic system is developed to cope with the complexity challenge introduced by the relatively small coverage area of a single VLC AP. Results show that the proposed method greatly outperforms conventional schedulers in terms of both throughput and fairness, while also achieving a fast convergence.

In this paper, the performance of non-orthogonal multiple access (NOMA) is characterized in a downlink visible light communication system for two separate cases. In the case of guaranteed quality of service (QoS) provisioning, we derive an analytical expression of the system coverage probability and show the existence of optimal power allocation coefficients on two-user paired NOMA. In the case of opportunistic best-effort service provisioning, we formulate a closed-form expression of the ergodic sum rate, which is applicable for arbitrary power allocation strategies. The probability that NOMA achieves higher individual rates than OMA is derived. Also, we give an upper bound of the sum rate gain of NOMA over OMA in the high signal-to-noise ratio regime. Both the theoretical and simulation results prove that the performance gain of NOMA over OMA can be further enlarged by pairing users with distinctive channel conditions. We also find out that the choice of light emitting diodes (LEDs) have a significant impact on the system performance. In the case of guaranteed QoS provisioning, the LEDs with larger semi-angles have better performance; while in the case of opportunistic best-effort service provisioning, the LEDs with 35° semi-angle give nearly optimal performance.

A coordinated multi-point visible light communication (CoMP-VLC) downlink is investigated, where the time-frequency (TF) resources are shared by a group of users relying on space division multiple access (SDMA). Linear zero-forcing (ZF) transmit precoding (TPC) is tailored to the CoMP-VLC system in order to eliminate the inter-user interference while accommodating the linear operating region of the light emitting diodes (LEDs). A SDMA user-group sharing the same TF has to host users having sufficiently different user-signatures, i.e. channel gains in a non-dispersive VLC channel. Hence we develop efficient resource allocation (RA) and SDMA user-grouping algorithms in the context of VLC systems that use intensity modulation and direct detection (IM/DD). Finding the optimal SDMA user-group requires an exhaustive search (ES), which has exponentially increasing complexity as the number of users increases. Therefore, a low-complexity user-grouping method is conceived and compared to three existing algorithms in terms of the attainable area spectral efficiency (ASE) and throughput fairness. Our simulation results demonstrate that the proposed algorithm gives a better performance-fairness trade-off than existing benchmarks.

The Internet of Things (IoT) is a new generation of network that can remotely and intelligently control distributed objects. Due to the large number of objects in the IoT, a high data traffic for the object communications is required, which is mostly routed through wireless links. However, the available spectrum for radio frequency (RF) wireless communications is exhausted so that each user in the IoT can only achieve very low data rate. In order to offer a better service to users, a light fidelity (Li-Fi) and radio frequency (RF) hybrid network is considered, where Li-Fi uses the large spectrum of visible light to achieve a high data rate, and the RF system guarantees a seamless coverage. In this study, a load balancing (LB) algorithm for the Li-Fi/RF hybrid IoT network is proposed based on evolutionary game theory (EGT). A key feature of the proposed algorithm is that users autonomously select the APs and adapt their strategies. Thus, compared with the conventional centralised algorithm, the computation load of the central unit (CU) can be reduced by using the EGT algorithm. Moreover, simulation results show that the proposed algorithm outperforms the conventional centralised algorithms in terms of the user satisfaction.

This research addresses the issue of resource and power allocation in small cell networks, and focuses on two aspects: i) the interference coordination among cells; and ii) the resource allocation among the users served by the same cell. Due to the density of small cells, centralised interference coordination schemes require an enormous level of communication among cells. Fuzzy logic (FL) is a promising low-complexity approach to realise autonomous interference coordination that does not need communication between cells. In this paper, we propose a novel FL method and associated decision-making algorithm for tackling resource allocation in small cell networks. Unlike the traditional FL method using the values of two states `yes' and `no' to describe how much a resource block (RB) should or should not be allocated, the proposed method employs a 3-state criterion that distinguishes high-quality RBs from medium-quality RBs. Also, the issue of allocating the RBs of a single cell to multiple users is studied in the FL method. Simulation results show that the proposed method can notably improve the performance of the traditional FL method in terms of both throughput and user satisfaction, without requiring extra processing power.

In this paper, the performance of non-orthogonal multiple access (NOMA) is characterized in a downlink visible light communication (VLC) system. Analytical expressions of the system performance are derived for two separate scenarios. In the scenario of achieving guaranteed quality of service (QoS), the outage probability of each user is studied and the effect of power allocation coefficients on the system coverage probability is investigated. In the scenario of providing opportunistic best-effort service, system ergodic sum rate is formulated based on a fixed power allocation (FPA) strategy. Both simulation and analytical results demonstrate that, in the first scenario, the maximum coverage probability can be achieved through the exhaustive search (ES) method to find the optimum set of power allocation coefficients. In the second scenario, it is shown that unlike orthogonal multiple access (OMA) techniques, NOMA can achieve a higher system capacity for a larger number of users. Also, the performance of NOMA can be further enhanced by choosing LEDs with a suitable semi-angle. When compared with OMA, NOMA can increase the system capacity by 125% if LEDs with 30° semi-angle are used.

A new concept for indoor positioning using visible light communication (VLC) is presented in this paper. The feasibility of uplink localization is verified so that the proposed system can work in conjunction with downlink positioning systems to improve overall localization accuracy. We propose to use an angle diversity transmitter (ADT) associated with accelerometers for uplink three-dimensional localization. Unlike other techniques using VLC for positioning, the proposed system can achieve indoor localization without making assumptions about the height or orientation angle of a mobile user. The received signal strength (RSS) method is used to minimize implementation costs while achieving satisfactory positioning accuracy. Simulation results show that an average localization error of less than 0.15 m can be achieved even when the receiver is tilted by 45°. Also, compared with a single light emitting diode (LED), the ADT is found to be more robust to accelerometer measurement errors.

In this paper, we propose an optimum transmit structure for spatial modulation (SM), a unique single-stream multiple-input multiple-output (MIMO) transmission technique. As a three-dimensional modulation scheme, SM enables a trade-off between the size of the spatial constellation diagram and the size of the signal constellation diagram. Based on this fact, the novel method, named transmission optimized spatial modulation (TOSM), selects the best transmit structure that minimizes the average bit error probability (ABEP). Unlike the traditional antenna selection methods, the proposed method relies on statistical channel state information (CSI) instead of instant CSI, and feedback is only needed for the optimal number of transmit antennas. The overhead for this, however, is negligible. In addition, TOSM has low computational complexity as the optimization problem is solved through a simple closed-form objective function with a single variable. Simulation results show that TOSM significantly improves the performance of SM at various channel correlations. Assuming Rayleigh fading channels, TOSM outperforms the original SM by up to 9 dB. Moreover, we propose a single radio-frequency (RF) chain base station (BS) based on TOSM, which achieves low hardware complexity and high energy efficiency. In comparison with multi-stream MIMO schemes, TOSM offers an energy saving of at least 56% in the continuous transmission mode, and 62% in the discontinuous transmission mode.

This work investigates the impact of power allocation for channel estimation (CE) in spatial modulation (SM). SM is a unique single-stream multiple-input multiple-output (MIMO) transmission technique that achieves spatial multiplexing gains. While SM completely avoids inter-channel interference, a single active antenna leads to a challenge for CE, i.e., all transmit antennas have to be activated sequentially to send the pilot signal. More time as well as energy is therefore consumed for CE. Driven by this motivation, we propose a closed-form optimal power allocation (OPA) for SM in this paper. Regardless of the SM transceiver structure, the optimal solution only depends on the pilot sequence length and the ratio between the number of pilots and the number of total symbols. Simulations results show that the bit error ratio performance of OPA-based SM tightly approaches the case of perfect channel state information with a gap of less than 0.8 dB.

Spatial modulation (SM) is a unique single-stream, multiple-input multiple-output (MIMO) transmission technique. Unlike multi-steam MIMO schemes, a single transmit antenna is activated in SM at any given time. Therefore, inter-channel interference is completely avoided. In addition, SM requires only one radio-frequency (RF) chain, regardless of the number of transmit antennas used. This key property results in a significant saving in quiescent power of power amplifiers. However, a challenge occurs when channel estimation (CE) is implemented for SM: the transmit antennas have to send pilots sequentially. Thus, in order not to compromise the throughput, the pilot number of each transmit antenna is restricted. This means we focus on improving the CE performance without increasing the number of pilots. In this paper, we propose a novel CE method for SM, in which the channel is jointly estimated across receive antennas. Simulation results show that the proposed approach outperforms the conventional method for various channel correlations between the receive antennas.

In this paper, a novel channel estimation (CE) method is proposed for spatial modulation (SM), a unique single-stream multiple-input-multiple-output transmission technique. In SM, there is only one transmit antenna being active at any time instance. While this property completely avoids inter-channel interference, it results in a challenge to estimate the channel information. In conventional CE (CCE) methods for SM, all transmit antennas have to be sequentially activated for sending pilots. Therefore, the time consumed in CE is proportional to the number of transmit antennas, which significantly compromises the throughput. By exploiting channel correlation, the proposed method, named transmission cross CE (TCCE), has the following characteristics: i) the entire channel is estimated by sending pilots through one transmit antenna; ii) it requires no overhead or feedback; and iii) it achieves a low computational complexity at the receiver. In addition, we propose an analytical framework to compute the distribution of the CE errors over time-varying fading channels. The corresponding average bit error probability (ABEP) bound of SM is also derived for the proposed method. Results show that the proposed ABEP bound matches with the simulations very well. When compared with CCE, the new method obtains a signal-to-noise ratio gain of up to 7.5 dB for medium and high correlations between the transmit antennas. Moreover, an adaptive CE technique can be readily implemented for SM via switching between CCE and TCCE.

In single-stream multiple-input multiple-output (MIMO) schemes, such as spatial modulation (SM) and space shift keying (SSK), a single transmit antenna is activated at any given time. Therefore, unlike multi-stream MIMO transmitters, simultaneous pilot transmissions are prohibitive because of a single radio-frequency (RF) chain. In state-of-the-art literature, the channels of different transmit antennas are individually estimated. As a result, more time is required to transfer pilots and the effective data rate is compromised. In this paper, we propose a novel channel estimation (CE) technique for single-stream MIMO systems. Given a pilot ratio, the proposed scheme achieves the same estimation period as multi-stream MIMO systems without needing any additional information or feedback. Simulations are implemented under a practical base station environment. Results show that for various speeds of the mobile user, the proposed approach significantly improves the performance of SM in comparison to the conventional CE method.

To improve the performance of spatial modulation (SM) over correlated MIMO channels, transmit optimized spatial modulation (TOSM) has been proposed recently. It trades off traditional signal constellation diagrams with spatial constellation diagrams to minimize the average bit error probability (ABEP). After the optimum number of transmit antennas is determined, the specific antennas need to be carefully chosen from the entire array to provide a minimum ABEP. Like in conventional transmit antenna selection (TAS) schemes, the problem can be solved by an exhaustive search. However, this results in an unaffordable complexity especially when the spectral efficiency is high. In this paper, we propose a creative TAS approach for TOSM. Given a required number of antennas, the novel technique determines the selection solution based on circle packing. Simulation results show that for various channel correlations and spectral efficiencies, the proposed method achieves performance results close to exhaustive search with a gap of less than 0.3 dB. The complexity is reduced to an extremely low level.

A unique characteristic of spatial modulation (SM) is the three dimensional constellation diagram. This enables to trade-off traditional signal constellation diagrams with spatial constellation diagrams where the latter is defined by the physical antenna array. In this paper we investigate the optimum pairs of signal and spatial constellation sizes with respect to average bit error probability (ABEP) and energy efficiency. The analysis is performed for varying antenna correlations and channel conditions. Both numerical and closed-form results are presented which show that a significant performance gain can be obtained when the optimal constellation pairs are used.