Search

It is known that deep neural networks (DNNs) classify an input image by paying particular attention to certain specific pixels; a graphical representation of the magnitude of attention to each pixel is called a saliency-map. Saliency-maps are used to check the validity of the classification decision basis, e.g., it is not a valid basis for classification if a DNN pays more attention to the background rather than the subject of an image. Semantic perturbations can significantly change the saliency-map. In this work, we propose the first verification method for attention robustness, i.e., the local robustness of the changes in the saliency-map against combinations of semantic perturbations. Specifically, our method determines the range of the perturbation parameters (e.g., the brightness change) that maintains the difference between the actual saliency-map change and the expected saliency-map change below a given threshold value. Our method is based on activation region traversals, focusing on the outermost robust boundary for scalability on larger DNNs. Experimental results demonstrate that our method can show the extent to which DNNs can classify with the same basis regardless of semantic perturbations and report on performance and performance factors of activation region traversals., Comment: 25 pages, 12 figures

Recompositing channel state information (CSI) from the beamforming feedback matrix (BFM), which is a compressed version of CSI and can be captured because of its lack of encryption, is an alternative way of implementing firmware-agnostic WiFi sensing. In this study, we propose the use of camera images toward the accuracy enhancement of CSI recomposition from BFM. The key motivation for this vision-aided CSI recomposition is to draw a first-hand insight that the BFM does not fully involve spatial information to recomposite CSI and that this could be compensated by camera images. To leverage the camera images, we use multimodal deep learning, where the two modalities, i.e., images and BFMs, are integrated to recomposite the CSI. We conducted experiments using IEEE 802.11ac devices. The experimental results confirmed that the recomposition accuracy of the proposed multimodal framework is improved compared to the single-modal framework only using images or BFMs.

The distributed inference (DI) framework has gained traction as a technique for real-time applications empowered by cutting-edge deep machine learning (ML) on resource-constrained Internet of things (IoT) devices. In DI, computational tasks are offloaded from the IoT device to the edge server via lossy IoT networks. However, generally, there is a communication system-level trade-off between communication latency and reliability; thus, to provide accurate DI results, a reliable and high-latency communication system is required to be adapted, which results in non-negligible end-to-end latency of the DI. This motivated us to improve the trade-off between the communication latency and accuracy by efforts on ML techniques. Specifically, we have proposed a communication-oriented model tuning (COMtune), which aims to achieve highly accurate DI with low-latency but unreliable communication links. In COMtune, the key idea is to fine-tune the ML model by emulating the effect of unreliable communication links through the application of the dropout technique. This enables the DI system to obtain robustness against unreliable communication links. Our ML experiments revealed that COMtune enables accurate predictions with low latency and under lossy networks., Comment: Submitted to IEEE Access

With regard to the implementation of WiFi sensing agnostic according to the availability of channel state information (CSI), we investigate the possibility of estimating a CSI matrix based on its compressed version, which is known as beamforming feedback matrix (BFM). Being different from the CSI matrix that is processed and discarded in physical layer components, the BFM can be captured using a medium-access-layer frame-capturing technique because this is exchanged among an access point (AP) and stations (STAs) over the air. This indicates that WiFi sensing that leverages the BFM matrix is more practical to implement using the pre-installed APs. However, the ability of BFM-based sensing has been evaluated in a few tasks, and more general insights into its performance should be provided. To fill this gap, we propose a CSI estimation method based on BFM, approximating the estimation function with a machine learning model. In addition, to improve the estimation accuracy, we leverage the inter-subcarrier dependency using the BFMs at multiple subcarriers in orthogonal frequency division multiplexing transmissions. Our simulation evaluation reveals that the estimated CSI matches the ground-truth amplitude. Moreover, compared to CSI estimation at each individual subcarrier, the effect of the BFMs at multiple subcarriers on the CSI estimation accuracy is validated.

This study experimentally validated the possibility of angle of departure (AoD) estimation using multiple signal classification (MUSIC) with only WiFi control frames for beamforming feedback (BFF), defined in IEEE 802.11ac/ax. The examined BFF-based MUSIC is a model-driven algorithm, which does not require a pre-obtained database. This contrasts with most existing BFF-based sensing techniques, which are data-driven and require a pre-obtained database. Moreover, the BFF-based MUSIC affords an alternative AoD estimation method without access to channel state information (CSI). Specifically, the extensive experimental and numerical evaluations demonstrated that the BFF-based MUSIC successfully estimates the AoDs for multiple propagation paths. Moreover, the evaluations performed in this study revealed that the BFF-based MUSIC achieved a comparable error of AoD estimation to the CSI-based MUSIC, while BFF is a highly compressed version of CSI in IEEE 802.11ac/ax., Comment: Submitted to IEEE Access

This paper demonstrates the feasibility of single-antenna and single-RF (radio frequency)- anchor received power strength indicator (RSSI) localization (SARR-LOC) with the assistance of the computer vision (CV) technique. Generally, to perform radio frequency (RF)-based device localization, either 1) fine-grained channel state information or 2) RSSIs from multiple antenna elements or multiple RF anchors (e.g., access points) is required. Meanwhile, owing to deficiency of single-antenna and single-anchor RSSI, which only indicates a coarse-grained distance information between a receiver and a transmitter, realizing localization with single-antenna and single-anchor RSSI is challenging. Our key idea to address this challenge is to leverage CV technique and to estimate the most likely first Fresnel zone (FFZ) between the receiver and transmitter, where the role of the RSSI is to detect blockage timings. Specifically, historical positions of an obstacle that dynamically blocks the FFZ are detected by the CV technique, and we estimate positions at which a blockage starts and ends via a time series of RSSI. These estimated obstacle positions, in principle, coincide with points on the FFZ boundaries, enabling the estimation of the FFZ and localization of the transmitter. The experimental evaluation revealed that the proposed SARR-LOC achieved the localization error less than 1.0 m in an indoor environment, which is comparable to that of a conventional triangulation-based RSSI localization with multiple RF anchors., Comment: Submitted to IEEE Internet of Things journal

The distributed inference framework is an emerging technology for real-time applications empowered by cutting-edge deep machine learning (ML) on resource-constrained Internet of things (IoT) devices. In distributed inference, computational tasks are offloaded from the IoT device to other devices or the edge server via lossy IoT networks. However, narrow-band and lossy IoT networks cause non-negligible packet losses and retransmissions, resulting in non-negligible communication latency. This study solves the problem of the incremental retransmission latency caused by packet loss in a lossy IoT network. We propose a split inference with no retransmissions (SI-NR) method that achieves high accuracy without any retransmissions, even when packet loss occurs. In SI-NR, the key idea is to train the ML model by emulating the packet loss by a dropout method, which randomly drops the output of hidden units in a DNN layer. This enables the SI-NR system to obtain robustness against packet losses. Our ML experimental evaluation reveals that SI-NR obtains accurate predictions without packet retransmission at a packet loss rate of 60%.

This paper proposes a fully decentralized federated learning (FL) scheme for Internet of Everything (IoE) devices that are connected via multi-hop networks. Because FL algorithms hardly converge the parameters of machine learning (ML) models, this paper focuses on the convergence of ML models in function spaces. Considering that the representative loss functions of ML tasks e.g, mean squared error (MSE) and Kullback-Leibler (KL) divergence, are convex functionals, algorithms that directly update functions in function spaces could converge to the optimal solution. The key concept of this paper is to tailor a consensus-based optimization algorithm to work in the function space and achieve the global optimum in a distributed manner. This paper first analyzes the convergence of the proposed algorithm in a function space, which is referred to as a meta-algorithm, and shows that the spectral graph theory can be applied to the function space in a manner similar to that of numerical vectors. Then, consensus-based multi-hop federated distillation (CMFD) is developed for a neural network (NN) to implement the meta-algorithm. CMFD leverages knowledge distillation to realize function aggregation among adjacent devices without parameter averaging. An advantage of CMFD is that it works even with different NN models among the distributed learners. Although CMFD does not perfectly reflect the behavior of the meta-algorithm, the discussion of the meta-algorithm's convergence property promotes an intuitive understanding of CMFD, and simulation evaluations show that NN models converge using CMFD for several tasks. The simulation results also show that CMFD achieves higher accuracy than parameter aggregation for weakly connected networks, and CMFD is more stable than parameter aggregation methods.

Millimeter wave (mmWave) beam-tracking based on machine learning enables the development of accurate tracking policies while obviating the need to periodically solve beam-optimization problems. However, its applicability is still arguable when training-test gaps exist in terms of environmental parameters that affect the node dynamics. From this skeptical point of view, the contribution of this study is twofold. First, by considering an example scenario, we confirm that the training-test gap adversely affects the beam-tracking performance. More specifically, we consider nodes placed on overhead messenger wires, where the node dynamics are affected by several environmental parameters, e.g, the wire mass and tension. Although these are particular scenarios, they yield insight into the validation of the training-test gap problems. Second, we demonstrate the feasibility of \textit{zero-shot adaptation} as a solution, where a learning agent adapts to environmental parameters unseen during training. This is achieved by leveraging a robust adversarial reinforcement learning (RARL) technique, where such training-and-test gaps are regarded as disturbances by adversaries that are jointly trained with a legitimate beam-tracking agent. Numerical evaluations demonstrate that the beam-tracking policy learned via RARL can be applied to a wide range of environmental parameters without severely degrading the received power., Comment: 13 pages, 13 figures, 3 tables, under submission for possible publication for IEEE

This paper discusses the feasibility of beam tracking against dynamics in millimeter wave (mmWave) nodes placed on overhead messenger wires, including wind-forced perturbations and disturbances caused by impulsive forces to wires. Our main contribution is to answer whether or not historical positions and velocities of a mmWave node is useful to track directional beams given the complicated on-wire dynamics. To this end, we implement beam-tracking based on deep reinforcement learning (DRL) to learn the complicated relationships between the historical positions/velocities and appropriate beam steering angles. Our numerical evaluations yielded the following key insights: Against wind perturbations, an appropriate beam-tracking policy can be learned from the historical positions and velocities of a node. Meanwhile, against impulsive forces to the wire, the use of the position and velocity of the node is not necessarily sufficient owing to the rapid displacement of the node. To solve this, we propose to take advantage of the positional interaction on the wire by leveraging the positions/velocities of several points on the wire as state information in DRL. The results confirmed that this results in the avoidance of beam misalignment, which would not be possible by using only the position/velocity of the node., Comment: 12 pages, 18 figures

Machine-learning-based prediction of future wireless link quality is an emerging technique that can potentially improve the reliability of wireless communications, especially at higher frequencies (e.g., millimeter-wave and terahertz technologies), through predictive handover and beamforming to solve line-of-sight (LOS) blockage problem. In this study, a real-time online trainable wireless link quality prediction system was proposed; the system was implemented with commercially available laptops. The proposed system collects datasets, updates a model, and infers the received power in real-time. The experimental evaluation was conducted using 5 GHz Wi-Fi, where received signal strength could be degraded by 10 dB when the LOS path was blocked by large obstacles. The experimental results demonstrate that the prediction model is updated in real-time, adapts to the change in environment, and predicts the time-varying Wi-Fi received power accurately.

This paper proposes a client selection method for federated learning (FL) when the computation and communication resource of clients cannot be estimated; the method trains a machine learning (ML) model using the rich data and computational resources of mobile clients without collecting their data in central systems. Conventional FL with client selection estimates the required time for an FL round from a given clients' computation power and throughput and determines a client set to reduce time consumption in FL rounds. However, it is difficult to obtain accurate resource information for all clients before the FL process is conducted because the available computation and communication resources change easily based on background computation tasks, background traffic, bottleneck links, etc. Consequently, the FL operator must select clients through exploration and exploitation processes. This paper proposes a multi-armed bandit (MAB)-based client selection method to solve the exploration and exploitation trade-off and reduce the time consumption for FL in mobile networks. The proposed method balances the selection of clients for which the amount of resources is uncertain and those known to have a large amount of resources. The simulation evaluation demonstrated that the proposed scheme requires less learning time than the conventional method in the resource fluctuating scenario.

This study develops a federated learning (FL) framework overcoming largely incremental communication costs due to model sizes in typical frameworks without compromising model performance. To this end, based on the idea of leveraging an unlabeled open dataset, we propose a distillation-based semi-supervised FL (DS-FL) algorithm that exchanges the outputs of local models among mobile devices, instead of model parameter exchange employed by the typical frameworks. In DS-FL, the communication cost depends only on the output dimensions of the models and does not scale up according to the model size. The exchanged model outputs are used to label each sample of the open dataset, which creates an additionally labeled dataset. Based on the new dataset, local models are further trained, and model performance is enhanced owing to the data augmentation effect. We further highlight that in DS-FL, the heterogeneity of the devices' dataset leads to ambiguous of each data sample and lowing of the training convergence. To prevent this, we propose entropy reduction averaging, where the aggregated model outputs are intentionally sharpened. Moreover, extensive experiments show that DS-FL reduces communication costs up to 99% relative to those of the FL benchmark while achieving similar or higher classification accuracy.

The goal of this work is the accurate prediction of millimeter-wave received power leveraging both radio frequency (RF) signals and heterogeneous visual data from multiple distributed cameras, in a communication and energy-efficient manner while preserving data privacy. To this end, firstly focusing on data privacy, we propose heteromodal split learning with feature aggregation (HetSLAgg) that splits neural network (NN) models into camera-side and base station (BS)-side segments. The BS-side NN segment fuses RF signals and uploaded image features without collecting raw images. However, the usage of multiple visual data leads to an increase in NN input dimensions, which gives rise to additional communication and energy costs. To overcome additional communication and energy costs due to image interpolation to blend different frame rates, we propose a novel BS-side manifold mixup technique that offloads the interpolation operations from cameras to a BS. Subsequently, we confront energy costs for operating a larger size of the BS- side NN segment due to concatenating image features across cameras and propose an energy-efficient aggregation method. This is done via a linear combination of image features instead of concatenating them, where the NN size is independent of the number of cameras. Comprehensive test-bed experiments with measured channels demonstrate that HetSLAgg reduces the prediction error by 44% compared to a baseline leveraging only RF received power. Moreover, the experiments show that the designed HetSLAgg achieves over 20% gains in terms of communication and energy cost reduction compared to several baseline designs within at most 1% of accuracy loss., Comment: 14 pages, 17 figures

This paper proposes a method to predict received power in urban area deterministically, which can learn a prediction model from small amount of measurement data by a simulation-aided transfer learning and data augmentation. Recent development in machine learning such as artificial neural network (ANN) enables us to predict radio propagation and path loss accurately. However, training a high-performance ANN model requires a significant number of data, which are difficult to obtain in real environments. The main motivation for this work was to facilitate accurate prediction using small amount of measurement data. To this end, we propose a transfer learning-based prediction method with data augmentation. The proposed method pre-trains a prediction model using data generated from ray-tracing simulations, increases the number of data using simulation-assisted data augmentation, and then fine-tunes a model using the augmented data to fit the target environment. Experiments using Wi-Fi devices were conducted, and the results demonstrate that the proposed method predicts received power with 50% (or less) of the RMS error of conventional methods.

Federated learning (FL) enables a neural network (NN) to be trained using privacy-sensitive data on mobile devices while retaining all the data on their local storages. However, FL asks the mobile devices to perform heavy communication and computation tasks, i.e., devices are requested to upload and download large-volume NN models and train them. This paper proposes a novel unsupervised pre-training method adapted for FL, which aims to reduce both the communication and computation costs through model compression. Since the communication and computation costs are highly dependent on the volume of NN models, reducing the volume without decreasing model performance can reduce these costs. The proposed pre-training method leverages unlabeled data, which is expected to be obtained from the Internet or data repository much more easily than labeled data. The key idea of the proposed method is to obtain a ``good'' subnetwork from the original NN using the unlabeled data based on the lottery hypothesis. The proposed method trains an original model using a denoising auto encoder with the unlabeled data and then prunes small-magnitude parameters of the original model to generate a small but good subnetwork. The proposed method is evaluated using an image classification task. The results show that the proposed method requires 35\% less traffic and computation time than previous methods when achieving a certain test accuracy.

Over-the-air computation (AirComp)-based federated learning (FL) enables low-latency uploads and the aggregation of machine learning models by exploiting simultaneous co-channel transmission and the resultant waveform superposition. This study aims at realizing secure AirComp-based FL against various privacy attacks where malicious central servers infer clients' private data from aggregated global models. To this end, a differentially private AirComp-based FL is designed in this study, where the key idea is to harness receiver noise perturbation injected to aggregated global models inherently, thereby preventing the inference of clients' private data. However, the variance of the inherent receiver noise is often uncontrollable, which renders the process of injecting an appropriate noise perturbation to achieve a desired privacy level quite challenging. Hence, this study designs transmit power control across clients, wherein the received signal level is adjusted intentionally to control the noise perturbation levels effectively, thereby achieving the desired privacy level. It is observed that a higher privacy level requires lower transmit power, which indicates the tradeoff between the privacy level and signal-to-noise ratio (SNR). To understand this tradeoff more fully, the closed-form expressions of SNR (with respect to the privacy level) are derived, and the tradeoff is analytically demonstrated. The analytical results also demonstrate that among the configurable parameters, the number of participating clients is a key parameter that enhances the received SNR under the aforementioned tradeoff. The analytical results are validated through numerical evaluations., Comment: 6 pages, 4 figures

This paper proposes a robust adversarial reinforcement learning (RARL)-based multi-access point (AP) coordination method that is robust even against unexpected decentralized operations of uncoordinated APs. Multi-AP coordination is a promising technique towards IEEE 802.11be, and there are studies that use RL for multi-AP coordination. Indeed, a simple RL-based multi-AP coordination method diminishes the collision probability among the APs; therefore, the method is a promising approach to improve time-resource efficiency. However, this method is vulnerable to frame transmissions of uncoordinated APs that are less aware of frame transmissions of other coordinated APs. To help the central agent experience even such unexpected frame transmissions, in addition to the central agent, the proposed method also competitively trains an adversarial AP that disturbs coordinated APs by causing frame collisions intensively. Besides, we propose to exploit a history of frame losses of a coordinated AP to promote reasonable competition between the central agent and adversarial AP. The simulation results indicate that the proposed method can avoid uncoordinated interference and thereby improve the minimum sum of the throughputs in the system compared to not considering the uncoordinated AP.

In this study, a contextual multi-armed bandit (CMAB)-based decentralized channel exploration framework disentangling a channel utility function (i.e., reward) with respect to contending neighboring access points (APs) is proposed. The proposed framework enables APs to evaluate observed rewards compositionally for contending APs, allowing both robustness against reward fluctuation due to neighboring APs' varying channels and assessment of even unexplored channels. To realize this framework, we propose contention-driven feature extraction (CDFE), which extracts the adjacency relation among APs under contention and forms the basis for expressing reward functions in the disentangled form, that is, a linear combination of parameters associated with neighboring APs under contention). This allows the CMAB to be leveraged with joint a linear upper confidence bound (JLinUCB) exploration and to delve into the effectiveness of the proposed framework. Moreover, we address the problem of non-convergence -- the channel exploration cycle -- by proposing a penalized JLinUCB (P-JLinUCB) based on the key idea of introducing a discount parameter to the reward for exploiting a different channel before and after the learning round. Numerical evaluations confirm that the proposed method allows APs to assess the channel quality robustly against reward fluctuations by CDFE and achieves better convergence properties by P-JLinUCB., Comment: 12 pages, 6 figures, 3 Tables

The goal of this study is to improve the accuracy of millimeter wave received power prediction by utilizing camera images and radio frequency (RF) signals, while gathering image inputs in a communication-efficient and privacy-preserving manner. To this end, we propose a distributed multimodal machine learning (ML) framework, coined multimodal split learning (MultSL), in which a large neural network (NN) is split into two wirelessly connected segments. The upper segment combines images and received powers for future received power prediction, whereas the lower segment extracts features from camera images and compresses its output to reduce communication costs and privacy leakage. Experimental evaluation corroborates that MultSL achieves higher accuracy than the baselines utilizing either images or RF signals. Remarkably, without compromising accuracy, compressing the lower segment output by 16x yields 16x lower communication latency and 2.8% less privacy leakage compared to the case without compression., Comment: 5 pages, 7 figures, to be published at IEEE Communications Letters

This letter proposes a novel random medium access control (MAC) based on a transmission opportunity prediction, which can be measured in a form of a conditional success probability given transmitter-side interference. A transmission probability depends on the opportunity prediction, preventing indiscriminate transmissions and reducing excessive interference causing collisions. Using stochastic geometry, we derive a fixed-point equation to provide the optimal transmission probability maximizing a proportionally fair throughput. Its approximated solution is given in closed form. The proposed MAC is applicable to full-duplex networks, leading to significant throughput improvement by allowing more nodes to transmit., Comment: 4 pages, 4 figures

Focusing on the received power prediction of millimeter-wave (mmWave) radio-frequency (RF) signals, we propose a multimodal split learning (SL) framework that integrates RF received signal powers and depth-images observed by physically separated entities. To improve its communication efficiency while preserving data privacy, we propose an SL neural network architecture that compresses the communication payload, i.e., images. Compared to a baseline solely utilizing RF signals, numerical results show that SL integrating only one pixel image with RF signals achieves higher prediction accuracy while maximizing both communication efficiency and privacy guarantees., Comment: 3 pages, Accepted in ACM CoNEXT 2019 Poster Session

For reliable millimeter-wave (mmWave) networks, this paper proposes cooperative sensing with multi-camera operation in an image-to-decision proactive handover framework that directly maps images to a handover decision. In the framework, camera images are utilized to allow for the prediction of blockage effects in a mmWave link, whereby a network controller triggers a handover in a proactive fashion. Furthermore, direct mapping allows for the scalability of the number of pedestrians. This paper experimentally investigates the feasibility of adopting cooperative sensing with multiple cameras that can compensate for one another's blind spots. The optimal mapping is learned via deep reinforcement learning to resolve the high dimensionality of images from multiple cameras. An evaluation based on experimentally obtained images and received powers verifies that a mapping that enhances channel capacity can be learned in a multi-camera operation. The results indicate that our proposed framework with multi-camera operation outperforms a conventional framework with single-camera operation in terms of the average capacity., Comment: arXiv admin note: text overlap with arXiv:1904.04585

Last year, IEEE 802.11 Extremely High Throughput Study Group (EHT Study Group) was established to initiate discussions on new IEEE 802.11 features. Coordinated control methods of the access points (APs) in the wireless local area networks (WLANs) are discussed in EHT Study Group. The present study proposes a deep reinforcement learning-based channel allocation scheme using graph convolutional networks (GCNs). As a deep reinforcement learning method, we use a well-known method double deep Q-network. In densely deployed WLANs, the number of the available topologies of APs is extremely high, and thus we extract the features of the topological structures based on GCNs. We apply GCNs to a contention graph where APs within their carrier sensing ranges are connected to extract the features of carrier sensing relationships. Additionally, to improve the learning speed especially in an early stage of learning, we employ a game theory-based method to collect the training data independently of the neural network model. The simulation results indicate that the proposed method can appropriately control the channels when compared to extant methods.

This paper proposes a cooperative mechanism for mitigating the performance degradation due to non-independent-and-identically-distributed (non-IID) data in collaborative machine learning (ML), namely federated learning (FL), which trains an ML model using the rich data and computational resources of mobile clients without gathering their data to central systems. The data of mobile clients is typically non-IID owing to diversity among mobile clients' interests and usage, and FL with non-IID data could degrade the model performance. Therefore, to mitigate the degradation induced by non-IID data, we assume that a limited number (e.g., less than 1%) of clients allow their data to be uploaded to a server, and we propose a hybrid learning mechanism referred to as Hybrid-FL, wherein the server updates the model using the data gathered from the clients and aggregates the model with the models trained by clients. The Hybrid-FL solves both client- and data-selection problems via heuristic algorithms, which try to select the optimal sets of clients who train models with their own data, clients who upload their data to the server, and data uploaded to the server. The algorithms increase the number of clients participating in FL and make more data gather in the server IID, thereby improving the prediction accuracy of the aggregated model. Evaluations, which consist of network simulations and ML experiments, demonstrate that the proposed scheme achieves a 13.5% higher classification accuracy than those of the previously proposed schemes for the non-IID case.

For millimeter-wave networks, this paper presents a paradigm shift for leveraging time-consecutive camera images in handover decision problems. While making handover decisions, it is important to predict future long-term performance---e.g., the cumulative sum of time-varying data rates---proactively to avoid making myopic decisions. However, this study experimentally notices that a time-variation in the received powers is not necessarily informative for proactively predicting the rapid degradation of data rates caused by moving obstacles. To overcome this challenge, this study proposes a proactive framework wherein handover timings are optimized while obstacle-caused data rate degradations are predicted before the degradations occur. The key idea is to expand a state space to involve time consecutive camera images, which comprises informative features for predicting such data rate degradations. To overcome the difficulty in handling the large dimensionality of the expanded state space, we use a deep reinforcement learning for deciding the handover timings. The evaluations performed based on the experimentally obtained camera images and received powers demonstrate that the expanded state space facilitates (i) the prediction of obstacle-caused data rate degradations from 500 ms before the degradations occur and (ii) superior performance to a handover framework without the state space expansion, Comment: 14 pages, 19 figures, Published at IEEE Transactions on Cognitive Communications and Networking

Sharing perceptual data with other vehicles enhances the traffic safety of autonomous vehicles because it helps vehicles locate other vehicles and pedestrians in their blind spots. Such safety applications require high throughput and short delay, which cannot be achieved by conventional microwave vehicular communication systems. Therefore, millimeter-wave (mmWave) communications are considered to be a key technology for sharing perceptual data because of their wide bandwidth. One of the challenges of data sharing in mmWave communications is broadcasting because narrow-beam directional antennas are used to obtain high gain. Because many vehicles should share their perceptual data to others within a short time frame in order to enlarge the areas that can be perceived based on shared perceptual data, an efficient scheduling for concurrent transmission that improves spatial reuse is required for perceptual data sharing. This paper proposes a data sharing algorithm that employs a graph-based concurrent transmission scheduling. The proposed algorithm realizes concurrent transmission to improve spatial reuse by designing a rule that is utilized to determine if the two pairs of transmitters and receivers interfere with each other by considering the radio propagation characteristics of narrow-beam antennas. A prioritization method that considers the geographical information in perceptual data is also designed to enlarge perceivable areas in situations where data sharing time is limited and not all data can be shared. Simulation results demonstrate that the proposed algorithm doubles the area of the cooperatively perceivable region compared with a conventional algorithm that does not consider mmWave communications because the proposed algorithm achieves high-throughput transmission by improving spatial reuse. The prioritization also enlarges the perceivable region by a maximum of 20%., Comment: IEICE TRANS. INF. & SYST

In millimeter wave (mmWave) vehicular communications, multi-hop relay disconnection by line-of-sight (LOS) blockage is a critical problem, especially in the early diffusion phase of mmWave-available vehicles, where not all the vehicles have mmWave communication devices. This paper proposes a distributed position control method for autonomous vehicles to make long relays connecting to road side units (RSUs) by avoiding blockages to communicate with each other via LOS paths. Even though vehicles with the proposed method do not use the whole information of the environments and cooperate with each other, they can decide their action (e.g., lane change and overtaking) to form long relays using only information of its surroundings (e.g., surrounding vehicle positions). The decision-making problem is formulated as a Markov decision process so that autonomous vehicles can learn a practical movement strategy of making long relays by a reinforcement learning (RL) algorithm. This paper designs a learning algorithm based on a sophisticated deep reinforcement learning algorithm, asynchronous advantage actor-critic (A3C), which enables vehicles to learn a complex movement strategy quickly by its deepneural-network architecture and multi-agent-learning mechanism. Once the strategy is well trained, vehicles can distributedly move to positions where the long relay to the RSU is established. Simulations results confirm that the proposed method can increase the relay length and coverage even if the traffic conditions and penetration ratio of mmWave communication devices in learning and operation phases are different., Comment: 16 pages, 11 figures, IEICE Transactions on Communications