Search

The intelligent radio is a novel emerging paradigm catching on in many practical applications, in order to provide context awareness techniques, able to learn by the environment conditions applying machine learning frameworks. The channel state information (CSI) prediction is one of the main challenges that an intelligent radio has to face with. This paper deals with the CSI prediction in the case of a buffer aided relaying communication scheme with adaptive link selection previously proposed in the literature where a direct source-destination link is not available and with both links affected by an independent fading. In particular, the CSI prediction has been reduced here to a time series prediction problem, performed by involving the usage of the recurrent neural networks. Finally, the validation of the effectiveness of the proposed intelligent radio buffer-aided relaying scheme with adaptive link selection approach has been pursued by presenting performance comparisons with the previously proposed scheme under the assumption of an ideal and actual CSI estimations, respectively. [ABSTRACT FROM AUTHOR]

The health status of the running gear in high-speed trains changes dynamically with time in a complete life cycle. Running gear systems composed of many coupled components are complex, and health statuses of which are difficult to predict in real time through a traditional health status prediction scheme. Lately, belief rule base (BRB), which is able to combine quantitative information and expert knowledge, has shown excellent expression in modeling complex systems. In the procedure of health status prediction, expert expertise can sufficiently enhance the accuracy and efficiency of this model. Therefore, this paper puts forwards a real-time health status prediction framework based on a multi-layer BRB with priority scheduling strategies for running gears. In the first-layer BRB, a time-series prediction model of multiple module BRB considering complete features is established. In the second-layer model, grey relation analysis (GRA) is employed in priority scheduling strategies of features. The third-layer BRB is used for assessing the health status of running gears by combining the features. In addition, the initial parameters of all module BRB given by experts may not be precise. Accordingly, the initial parameters in the BRB are updated by the recursive algorithm online. Finally, the proposed method is tested on the testing platform in running gears. The results make it clear the proposed method can predict the health status of running gears with much accuracy in real time. [ABSTRACT FROM AUTHOR]

Recently chaos theory has emerged as a powerful tool to address forecasting problems of nonlinear time series, since it is able to meet the dynamical and geometrical structures of very complex systems, reaching higher accuracy on the prediction values than the classical approaches. This paper aims at applying the chaos theory principles to different problems, in order to pursue high levels of accuracy on the predicted results. After the verification of the chaotic behavior of the datasets taken into analysis through the largest Lyapunov exponent research, the detection of the suitable embedding dimension and time delay has been carried out, in order to reconstruct the phase space of the underlying dynamical systems. Three different predictive methods have been proposed for different datasets. Finally, the performance comparison with the moving average model, a deep neural network based strategy, and a chaos theory based algorithm recently proposed in literature has been provided. [ABSTRACT FROM AUTHOR]

In vehicle trajectory prediction, the difficulty in modeling the interaction relationship between vehicles lies in constructing the interaction structure between the vehicles in the traffic scene. Majority of existing models only focus on the interaction between the historical trajectory of the vehicle and the surrounding vehicles in the spatial domain, and do not pay attention to the interaction between the vehicle and the non-Euclidean correlation structure (graph structure) that exists in the environment. In order to overcome the deficiencies in the existing models, this paper proposes the Environment-Attention Network model (EA-Net) to obtain the full interactive information between the vehicle and its driving environment. In the proposed model, a new type of parallel structure consisting of Graph Attention network (GAT) and Convolutional social pooling containing Squeeze-and-Extraction mechanism (SE-CS), is constructed as the environmental feature extraction module and embedded in LSTM encoder-decoder. This structure solves the limitation of the dimension and structure influence when constructing the interaction model between the vehicle and the surrounding environment, making the extracted feature information comprehensive and effective. The prediction accuracy of the model with the RMSE loss function is tested on two public datasets– NGSIM and highD, and compared with several state-of-the-art trajectory prediction algorithm models. The results show that the prediction accuracy of the proposed Environment-Attention Network in the two datasets is more than 20% higher than that of the single structure model, which indicates that the proposed model proposed has superior performance and better adaptability to different traffic environments compared with the existing models. [ABSTRACT FROM AUTHOR]

A wet road is slippery so vehicles often slow down their speed to increase the safety margin, thus usually reducing the average speed. This reduction in average speed may produce a chain reaction that shifts, extends, or amplifies a slowdown on downstream road segments. Conventional average-speed forecasting approaches are unable to respond to sudden chain reactions because these approaches do not consider the effect of weather factors and upstream road segments. Since accurate forecast of average speed can improve gas consumption, carbon dioxide emissions, and travel time, this paper proposes a short-term average-speed forecast and adjustment (ASFA) approach based on a study of prediction bias correlation among adjacent road segments and on weather factors, such as temperature, humidity, and rainfall. First, this approach applies an artificial neural network to predict the average speed of a road segment. Then, vehicles can monitor current traffic to calculate average speed via vehicular ad hoc networks (VANETs). Finally, the vehicles adjust the predicted average speed according to the observed average speed. Real traffic measurements and weather data are used for the evaluation of the proposed scheme and Civic Boulevard in Taipei City is selected as the prediction target. The results show that the proposed ASFA improves accuracy by 57.4% when compared with a hybrid approach on an urban street during rush hour. This paper estimates and simulates a case study by aggregating traffic data in 186 ms on Shi-Min Boulevard via the VANETs during rush hour. [ABSTRACT FROM PUBLISHER]

Long queues of vehicles are often found at signalized intersections, which increases the energy consumption of all the vehicles involved. This paper proposes an enhanced eco-approach control (EEAC) strategy with consideration of the queue ahead for connected electric vehicles (EVs) at a signalized intersection. The discharge movement of the vehicle queue is predicted by an improved queue discharge prediction method (IQDP), which takes both vehicle and driver dynamics into account. Based on the prediction of the queue, the EEAC strategy is designed with a hierarchical framework: the upper-stage uses dynamic programming to find the general trend of the energy-efficient speed profile, which is followed by the lower-stage model predictive controller to computes the explicit solution for a short horizon with guaranteed safe inter-vehicular distance. Finally, numerical simulations are conducted to demonstrate the energy efficiency improvement of the EEAC strategy. Besides, the effects of the queue prediction accuracy on the performance of the EEAC strategy are also investigated. [ABSTRACT FROM AUTHOR]

Electric vehicles (EVs) are considered to relieve energy crisis, and environmental problems due to their high efficiency, and low emissions, and energy management strategies (EMSs) have been extensively studied to improve the performance of hybrid energy storage systems (HESSs) for EVs. To effectively reduce HESS energy loss, and extend battery life, this paper proposes a predictive EMS (PEMS) for the battery/supercapacitor HESSs. First, the pattern sequence-based velocity predictor is presented to accurately predict the future short-term velocity profile. Second, the PEMS is proposed by formulating an HESS power split optimization problem, where the HESS energy loss, and the battery capacity loss are considered. Third, an improved chaotic particle swarm optimization algorithm is presented to solve the formulated optimization problem. Simulation results demonstrate that, compared with the benchmark, the proposed PEMS can effectively reduce the HESS energy loss, and extend the battery lifetime at the same time. [ABSTRACT FROM AUTHOR]

A hybrid method for the prediction of the remaining useful life (RUL) of Lithium-ion batteries considering error-correction is proposed in respect of capacity diving phenomenon in later capacity degradation. First, an improved empirical capacity degradation model is proposed based on our previous work, with the analytic expression further revised in this paper to enhance its fitting accuracy, parameter identifiability, and applicability to tracking algorithm. Unscented particle filter (UPF) algorithm is then implemented to obtain prognostic results with original error series. Next, to enhance the quality of original error data by reducing the local uncertainty, complete ensemble empirical mode decomposition (CEEMD) algorithm is utilized to reconstruct error series. The fundamental error evolution information is retained by selecting relatively highly correlated intrinsic mode functions (IMFs) of the decomposition results of original error series. Finally, after employing Gaussian process regression (GPR) algorithm, the prognostic error to correct the UPF-based prognostic result is obtained from the reconstructed error series. RUL prediction experiments for batteries with different working conditions have been conducted to verify the improved performance of the proposed model and the hybrid method, with mean absolute percentage errors of battery capacity degradation predicted less than 0.4%. [ABSTRACT FROM AUTHOR]

Rear-end collision prediction has gained an increasing attention for safety improvement in smart cities. It is urgent to design efficient warning strategies for rear-end collisions which is one of the main causes of traffic accidents. The existing researches have been conducted to predict the collisions. Learning-based methods are proposed to solve this complicated issue, which the traditional methods are difficult to solve. However, due to some limitations in terms of the feature extraction and prediction performance, back-propagation learning methods are facing challenge. In this paper, we proposed a novel Rear-end Collision Prediction Mechanism with deep learning method (RCPM), in which a convolutional neural network model is established. In RCPM, the dataset is smoothed and expanded based on genetic theory to alleviate the class imbalance problem. The preprocessed dataset is divided into training and testing sets as the input to train our convolutional neural network model. The experimental results show that compared with the Honda, Berkeley and multi-layer perception neural-network-based algorithms, RCPM effectively improves performance to predict rear-end collisions. [ABSTRACT FROM AUTHOR]

Ultra-dense networking (UDN) is a promising technology to improve the network capacity in the next-generation mobile communication system. The virtualization paradigm is tightly integrated into UDN to address the problem of interference management. However, mobility management based on virtual cells meets significant challenges in UDN due to the frequent handovers and massive signaling overhead. These problems become severe for vehicles owing to their high-speed movement. In this paper, driven by trajectory prediction using a real-world vehicle mobility dataset, we propose a proactive mobility management solution based on virtual cells. Four modules are designed in the centralized Software-Defined Networking controller to support the proposed solution. The proposed LSTM-DR framework predicts the next locations of vehicles by integrating Long Short-Term Memory (LSTM) networks and Dead Reckoning (DR) method. The active gNBs selection algorithm selects the serving gNBs to form virtual cells according to predicted locations and mobility preferences. The corresponding signaling procedure is then carefully designed to further reduce the signaling overhead. Simulation results show that the proposed prediction framework can achieve higher accuracy and robustness in trajectory prediction. The proposed proactive solution reduces the handover frequency and handover failure rate and thereby saves the handover signaling overhead significantly. [ABSTRACT FROM AUTHOR]

Video services have hold a surprising proportion of the whole network traffic in wireless communication networks. Accurate prediction of video traffic can endow networks with intelligence in resource management, especially for the forthcoming beyond the fifth-generation (B5G) networks. However, the existing approaches fail to accurately predict video traffic with all types of frames, due to the natures of strong long-range dependence, self-similarity and burstiness. Obviously, it is unable to meet the QoS and QoE requirements of dynamic bandwidth allocation. In this paper, we propose the feasibility of advanced machine learning methodology applied in nonstationary video traffic prediction, i.e., smoothing-aided support vector machine (SSVM) model. The model utilizes classical smoothing methods to preprocess video traffic by relieving the drastic fluctuation of video stream. It can provide an effective association for the subsequent support vector regression, as the preprocessed data becomes more smooth and continuous than the original unprocessed one. Experimental results show that our proposed model significantly outperforms the state of the art model, i.e., logistic smooth transition autoregressive, in prediction performance. The superior nonlinear approximation capacity is further demonstrated by visualized statistical analysis. [ABSTRACT FROM AUTHOR]

This paper demonstrates a real world case study of a new method for robust simultaneous estimation of two parameters using multiple data sources. The method is used to simultaneously estimate vehicle mass and road gradient. No additional sensors are required beyond those that would normally be found on a vehicle controller network. The estimation algorithm combines components driven by observer state error and also directly by the parameter error using a sliding-mode inspired regressor structure. The algorithm incorporates a novel information fusion method that is integral to the regressor structure and a supervised data-rejection system to limit estimation activity in periods of recognised error-promoting activity. The estimation method has been demonstrated in real time on a modified production passenger car platform. It has been shown to be effective at robustly predicting road gradient and offering more reliable and stable prediction of vehicle mass than existing estimation methods employed in the same multi-parameter estimation context. The estimator allows prediction of vehicle mass whose limiting factor is the bandwidth and accuracy of the available driveline torque information, not the algorithm itself, allowing the identification of a 150 kg change in mass on a 2000 kg vehicle in this case study. [ABSTRACT FROM AUTHOR]

Wireless fidelity (WiFi) and long term evolution (LTE) coexistence system has attracted significant attention recently. However, most studies on coexistence only consider the issues about downlink performance and fairness between WiFi network and LTE network. In this paper, we take a comprehensive consideration of both the performance of WiFi and LTE, to propose a downlink/uplink decoupling (DUDe) scheme for the fairness of the coexistence system. Through the process of modeling construction, problem formulation with reformulation, solution and simulations, we illustrate that DUDe can achieve proportional fairness in terms of sum throughput and load balancing between uplink and downlink. In our decoupling model for coexistence between WiFi and LTE, resource allocation is under adaptive spectrum partitioning and users can access each network of WiFi and LTE. Moreover, to make our proposal more realistic, we adopt support vector machine (SVM) to predict the total number of users in WiFi and LTE coexistence system, where the predictive simulation shows its high prediction accuracy with low complexity. We further apply the prediction results to the optimization problem of DUDe. Simulation results manifest that the proposed DUDe scheme has better performance in terms of sum throughput. The sum throughput of DUDe scheme can come up to almost twice more than that of coupled association scheme. The result shows that DUDe can also achieve load balance between uplink and downlink in the coexistence system. [ABSTRACT FROM AUTHOR]

For the 360-degree virtual reality (VR) videos, to achieve the same user-perceived playout quality with traditional video, the overall video bitrate has to be much higher, which significantly increases the bandwidth requirement. To overcome this issue, most of rate adaptation schemes are proposed based on the users’ Field of View (FoV), since users only view part of the entire videos when watching 360-degree videos. However, the FoV prediction is only valid in short terms which incurs low robustness against network dynamics and results in frozen screen events. Since the playback buffer can effectively avoid re-buffering by storing videos in advance, we propose a two-layered hierarchical buffer-based rate adaptation scheme for 360-degree VR videos in this paper. The tiles will be updated to enhance the video quality in the short buffer layer based on FoV prediction, meanwhile the long buffer layer takes charge of downloading new chunks to guarantee the buffer occupancy. To better improve the quality of experience (QoE), we jointly consider the users’ FoV, channel conditions and buffer status. Our method achieves a good trade-off between video quality and buffer occupancy. Besides, the maximum buffer level selection is optimized to further improve the performance. The simulation results show that our proposed scheme and algorithms outperform other methods by about 5%-30% of QoE improvement. [ABSTRACT FROM AUTHOR]

In time-varying fading channels, channel coefficients are estimated using pilot symbols that are transmitted every coherence interval. For channels with high Doppler spread, the rapid channel variations over time will require considerable bandwidth for pilot transmission, leading to poor throughput. In this paper, we propose a novel receiver architecture using deep recurrent neural networks (RNNs) that learns the channel variations and thereby reduces the number of pilot symbols required for channel estimation. Specifically, we design and train an RNN to learn the correlation in the time-varying channel and predict the channel coefficients into the future with good accuracy over a wide range of Dopplers and signal-to-noise ratios (SNR). The proposed training methodology enables accurate channel prediction through the use of techniques such as teacher-force training, early-stop, and reduction of learning rate on plateau. Also, the robustness of prediction for different Dopplers and SNRs is achieved by adapting the number of predictions into the future based on the Doppler and SNR. Numerical results show that good bit error performance is achieved by the proposed receiver in time-varying fading channels. We also propose a data decision driven receiver architecture using RNNs that further reduces the pilot overhead while maintaining good bit error performance. [ABSTRACT FROM AUTHOR]

With 5G already deployed, challenges related to handover exacerbate due to the dense base station deployment operating on a motley of frequencies. In this paper, we present and evaluate a novel data-driven solution, to reduce inter-frequency handover failures (HOFs), hereafter referred to as TORIS (Transmit Power Tuning-based Handover Success Rate Improvement Scheme). TORIS is designed by developing and integrating two sub-solutions. First sub-solution consists of an Artificial Intelligence (AI)-based model to predict inter-frequency HOFs. In this model, we achieve higher than the state-of-the-art accuracy by leveraging two approaches. First, we devise a novel feature set by exploiting domain knowledge gathered from extensive drive test data analysis. Second, we exploit an extensive set of data augmentation techniques to address the class imbalance in training the HOF prediction model. The data augmentation techniques include Chow-Liu Bayesian Network and Generative Adversarial Network further improved by focusing the sampling only on the borderline. We also compare the performance of state-of-the-art AI models for predicting HOFs with and without augmented data. Results show that AdaBoost yields best performance for predicting HOFs. The second sub-solution is a heuristic scheme to tune the transmit (Tx) power of serving and target cells. Unlike the state-of-the-art approaches for HOF reduction that tune cell individual offset, TORIS targets the main cause of HOFs i.e., poor signal quality and propagation condition, by proactively varying the Tx power of the cells whenever a HOF is anticipated. Results show that TORIS outperforms the state-of-the-art HOF reduction solution and yields 40%-75% reduction in HOFs. [ABSTRACT FROM AUTHOR]

In this paper, we introduce channel prediction to wireless communications aided by intelligent reflecting surface (IRS) in order to reduce the overhead brought by full channel estimation. Hopefully, fine-grained control of reflection coefficients on IRS can still be maintained with sufficiently accurate channel prediction. Innovatively, we consider channel prediction as an analogy to image super-resolution in computer vision since elements on an IRS are normally two-dimensionally arranged, which makes the popular deep learning techniques efficient solution. To this end, we first solve the element control problem via semi-definite relaxation (SDR) and then design an end-to-end deep-learning model for channel prediction. With estimation for only part of the entire channel state information (CSI), the total channel estimation overhead can be significantly reduced. We visually show that the purposed deep-learning model maintains considerably high accuracy for channel prediction, which makes the proposed channel-prediction method outperforms the existing group-based method in terms of achievable rate. Moreover, the tradeoff between channel estimation overhead and channel prediction accuracy is also investigated with numerical results. [ABSTRACT FROM AUTHOR]

Accurate state of energy (SOE) estimation of the battery pack is very important for the electric vehicle's driving range estimation, which is still a very challenging problem under actual vehicle application conditions at present. This paper proposes a method for estimating the SOE of the battery pack based on prediction, taking into account the future voltage and temperature changes of the battery pack. Firstly, a novel definition of the battery pack SOE is proposed, which considers the inconsistency of the battery pack. Secondly, the thermo-electric lumped parameter equivalent circuit model of a single cell in the battery pack is applied. The electrical parameters are identified using the forgetting factor recursive least-squares algorithm, and the extended Kalman filter is used to estimate the SOC of each cell. Finally, the future current, voltage and temperature response of the cell in the battery pack are accurately predicted and iterative calculations are performed to get SOE under the condition that the future output power of the battery pack is predicted by the Markov chain model method. The effectiveness of the proposed method is verified under dynamic conditions at different temperatures. The results show that the above method has high precision and good robustness, and can meet the needs of actual vehicle applications. [ABSTRACT FROM AUTHOR]

Wireless Virtual Reality (VR) users are able to enjoy immersive experience from anywhere at anytime. However, providing full spherical VR video with high quality under limited VR interaction latency is challenging. If the viewpoint of the VR user can be predicted in advance, only the required viewpoint is needed to be rendered and delivered, which can reduce the VR interaction latency. Therefore, in this paper, we use offline and online learning algorithms to predict viewpoint of the VR user using real VR dataset. For the offline learning algorithm, the trained learning model is directly used to predict the viewpoint of VR users in continuous time slots. While for the online learning algorithm, based on the VR user's actual viewpoint delivered through uplink transmission, we compare it with the predicted viewpoint and update the parameters and input viewpoints of the online learning algorithm to further improve the prediction accuracy. To guarantee the reliability of the uplink transmission, we integrate the Proactive retransmission scheme into our proposed online learning algorithm. Simulation results show that our proposed online learning algorithm for uplink wireless VR network with the proactive retransmission scheme only exhibits about 5 $\%$ prediction error. [ABSTRACT FROM AUTHOR]

As unmanned aerial vehicles (UAVs) have been widely employed, the spectrum resource for UAV communication becomes scarce. The UAVs may communicate based on spectrum sharing, requiring efficient spectrum prediction to improve the performance of spectrum sensing. However, since the existing prediction methods are generally designed to predict the spectrum states in the next moment at the current location, they can not be directly applied to UAVs which rapidly move to the next location when obtaining the prediction result. In another word, the UAVs require to predict the spectrum in the next moment at the next location. This is a method which can predict spectrum in the joint temporal and spatial dimensions. The main challenge is the historical data of the next location is usually not possible to be obtained in advance by the UAVs. This paper proposes a new prediction approach which firstly estimates the historical data of the next location based on the homotopy theory (HT), and then performs the prediction based on the hidden Markov model (HMM), referring to the HT-HMM based prediction. Experimental results show that the HT-HMM based approach is able to efficiently predict the spectrum states of the next location. [ABSTRACT FROM AUTHOR]

Remaining useful life (RUL) prediction of lithium-ion batteries plays an important role in intelligent battery management systems (BMSs). The current RUL prediction methods are mainly developed based on offline training, which are limited by sufficiency and reliability of available data. To address this problem, this paper presents a method for RUL prediction based on the capacity estimation and the Box-Cox transformation (BCT). Firstly, the effective aging features (AFs) are extracted from electrical and thermal characteristics of lithium-ion batteries and the variation in terms of the cyclic discharging voltage profiles. The random forest regression (RFR) is then employed to achieve dependable capacity estimation based on only one cell's degradation data for model training. Secondly, the BCT is exploited to transform the estimated capacity data and to construct a linear model between the transformed capacities and cycles. Next, the ridge regression algorithm (RRA) is adopted to identify the parameters of the linear model. Finally, the identified linear model based on the BCT is employed to predict the battery RUL, and the prediction uncertainties are investigated and the probability density function (PDF) is calculated through the Monte Carlo (MC) simulation. The experimental results demonstrate that the proposed method can not only estimate capacity with errors of less than 2%, but also accurately predict the battery RUL with the maximum error of 127 cycles and the maximum spans of 95% confidence of 37 cycles in the whole cycle life. [ABSTRACT FROM AUTHOR]

Channel prediction is a vital technique that can be used to support adaptive transmissions and mitigate the feedback delay of the channel state information (CSI) in the wireless communications, especially in the frequency division duplex (FDD) system. In this paper, we focus on the channel prediction issue in multiple input-multiple output orthogonal frequency division multiplexing (MIMO-OFDM) systems. First, we introduce the general channel prediction framework based on the temporal and spatial correlations in MIMO-OFDM systems. Then, we combine the broad learning idea and recurrent computation, and further introduce the broad echo state network (BESN) for channel prediction in MIMO-OFDM systems. Third, to estimate the output weight matrix, we offer two versions of the BESN, i.e., the basic BESN (B-BESN) and the group forward variable selection (GFVS)-based BESN (GFVS-BESN). In the latter, we develop the GFVS strategy to further extract useful information from those collected features in the BESN and make the BESN more able to process the CSI samples. Fourth, inspired by deep learning, we import the conjugate gradient descent backpropagation (CGDBP) technique to fine-tune the random weights and biases in the BESN and give the related derivations. Then, we prove the echo state property in the BESN and analyze the computational complexity. In the simulation section, we comprehensively evaluate the prediction performances for the standard Extended Vehicular A model (EVA) and Extended Typical Urban model (ETU) under different signal-to-noise ratios (SNRs), different antenna configurations, different spatial correlations, different maximum Doppler shifts and different channel prediction methods. The simulation results indicate that the BESN has excellent prediction performance in MIMO-OFDM systems. [ABSTRACT FROM AUTHOR]

Mobile traffic classification and prediction are key tasks for network optimization. Most of the works in this area present two main drawbacks. First, they treat the two tasks separately, thus requiring high computational capabilities. Second, they perform data mining on the information collected from the data plane, which is unsuitable for the mobile edge. To bridge this gap, this paper properly tailors a Multi-Task Learning model running directly at the edge of the network to anticipate information on the type of traffic to be served and the resource allocation pattern requested by each service during its execution. Our study exploits data mining from the control channel of an operative mobile network to also reduce storage and monitoring processing. Different configurations of neural networks, which adopt autoencoders (i.e. Undercomplete Autoencoder or Sequence to Sequence Autoencoder) as key building blocks of the proposed Multi-Task Learning methodology for common feature representations, are investigated to evaluate the impact of the observation window of traffic profiles on the classification accuracy, prediction loss, complexity, and convergence. The comparison with respect to conventional single-task learning approaches, that do not use autoencoders and tackle classification and prediction tasks separately, clearly demonstrates the effectiveness of the proposed Multi-Task Learning approach under different system configurations. [ABSTRACT FROM AUTHOR]

Predictive resource allocation (PRA) can leverage machine learning and historical data to boost the performance of mobile networks, say to support high throughput or improve the quality of service. While the large gain of PRA has been demonstrated by assuming perfect prediction or by optimizing with modeled prediction errors, how the prediction errors translate to the performance loss of PRA and how the future information should be predicted are not well understood. In this paper, we consider high throughput PRA policy for non-real-time (NRT) mobile users when the base stations also serve real-time (RT) users, where the future average rates of NRT users in a prediction window need to be predicted. The average rate can be predicted in various ways, where one typical way is to predict the rate directly from the historical average rates and another typical way is to predict the rate indirectly by first predicting traffic load and user trajectory. Given that the traffic load and user location can be gathered easily in cellular networks, we consider such an indirect prediction and analyze the impact of prediction errors of traffic load and trajectory on the prediction errors of average data rate. To this end, we first find the relation between the RT traffic load and the residual bandwidth after serving the RT users by resorting to effective capacity and effective bandwidth theory. To further justify why we analyze the indirect prediction, we compare the predictability of the time series used for the typical direct prediction and indirect prediction with entropy theory. Since different machine learning techniques lead to different prediction bias and prediction error variance, we then derive the relation between the prediction error statistics of the average data rate and those of traffic load and trajectory. We evaluate the performance loss of PRA using both modeled prediction errors and real predictions with deep learning. Our analyses show that for a network busy with RT service, the prediction errors of RT traffic load have larger impact on those of average data rate than the prediction errors of trajectory. The prediction bias has a larger impact on the throughput loss of PRA than the prediction error variance. This provides a guidance of designing learning techniques for predicting the required information. [ABSTRACT FROM AUTHOR]

Human mobility prediction techniques are instrumental for many important applications including service management and city planning. Previous work looks at the inherent patterns of a user's historical trajectories to predict his/her future location. Such method suffers when only a small number of historical locations are available, which is a main challenge for mobility prediction on sparse trajectory datasets. In this paper, we propose a multi-task learning based algorithm to predict users’ mobility by learning the mobility behaviors of different users at the same time and exploiting the similarities among them to overcome the sparse issue and improve the prediction performance. Specifically, we use a Bayesian mixture model to describe users’ spatio-temporal mobility patterns, where we introduce a novel von Mises distribution to model the temporal distribution of users’ mobility to better preserve its continuity across time. Then, we use the hierarchical Dirichlet process to perform joint prediction of all users’ mobility. This model allows us to leverage similar mobility patterns among users to improve the prediction performance for users with sparse trajectory data. We conduct extensive evaluations using data collected by a cellular network operator and mobile applications. Our evaluations show that our algorithms achieve prediction accuracy of 53.9% on average and 73.2% for users with high-quality mobility data, with the performance gap over 8.7% on average compared with the state-of-the-art algorithms, demonstrating the effectiveness of our proposed algorithm. [ABSTRACT FROM AUTHOR]

The quasi-static assumption of channels becomes invalid in a number of emerging applications of massive multiple-input multiple-output (MIMO) systems with high base station (BS)/user mobility, such as high speed train and unmanned aerial vehicle communications. In these situations, the time variation of channels shortens the channel coherence time and decreases the efficiency of traditional channel estimation schemes significantly. This paper focuses on the time-varying channel estimation problem in massive MIMO systems, where the mobility of BS/user is assumed to be high. The sparse property of the time-varying massive MIMO channels is analyzed, which reveals that the angular domain channels exhibit two kinds of spatial-temporal sparse (STS) structures, namely the temporal-common sparse structure and spatial-clustered sparse structure. By exploiting the STS structures, a novel structured variational Bayesian inference (VBI) framework is formulated to make a robust channel estimation during the training phase. A data-aided channel predication scheme is further proposed to combat the time variation of the channels during the data transmission phase without increasing the pilot consumption. The simulation results demonstrate the superiority of the proposed scheme in the time-varying scenarios with different BS/user mobilities. [ABSTRACT FROM AUTHOR]

The fifth-generation wireless network (5G) is expected to support vehicle to infrastructure (V2I) communication which requires an accurate understanding of its highly mobile and dynamic propagation environment for efficient physical layer design. Towards this end, this paper proposes a joint design of adaptive channel prediction, beamforming and scheduling for 5G V2I communications. The channel prediction algorithm is designed without using pilot signals and assuming the channel impulse response (CIR) model. In this regard, we first propose an adaptive recursive least squares (RLS) CIR prediction approach that uses few of the past estimated CIRs to predict one or more future orthogonal frequency division multiplexing (OFDM) block CIR coefficients. Then, we jointly design beamforming and vehicle equipment (VE) scheduling for each sub-carrier to maximize the uplink channel average sum spectrum efficiency (SE) by employing the predicted CIR. The beamforming problem is formulated as a Rayleigh quotient optimization where its global optimal solution can be obtained. Moreover, the VE scheduling design is formulated as an integer programming problem solved using a greedy search. To enhance the prediction performance, we propose a model-free Q-learning technique to learn a strategy for selecting the best CIR predictor at the current OFDM block. In addition, as the CIR coefficients might contain undesired taps with zero (negligible) average powers, we further examine a dominant CIR tap index identification problem (formulated as convex) from the past observed signals. We carry out extensive simulations to validate analytical expressions and examine the effects of different parameters including the VE speed on the achievable sum SE. Numerical simulations corroborate the superiority of the proposed channel prediction and scheduling algorithms over the existing ones and the relevance of identifying dominant CIR taps. [ABSTRACT FROM AUTHOR]

Load imbalance among small and macro cells is a major challenge that undermines the gains of emerging ultradense heterogeneous networks (HetNets). Existing load balancing (LB) schemes have one common caveat which is operating in reactive mode i.e., cell parameters are tweaked reactively in accordance with the dynamics of cell loads. The inherent reactiveness of these LB schemes hinder in achieving promising quality of experience (QoE) gains from 5G and beyond. To cope with this issue, in this paper we propose a novel proactive load balancing framework “OPERA” empowered by mobility prediction paradigm for future ultra dense networks (UDNs). The pro-activeness of OPERA stems from its novel capability that instead of passively waiting for congestion indicators to be observed and then reacting to them, OPERA predicts future cell loads and then proactively optimizes key antenna parameters and cell individual offsets (CIOs) to preempt congestion before it happens. OPERA also incorporates capacity and coverage constraints and load aware association strategy for ensuring conflict free operation of LB and coverage and capacity optimization (CCO) self-organizing network (SON) functions. Simulation results show that compared to real network deployments settings and published state-of-the-art reactive schemes, OPERA can yield significant gain in terms of fairness in load distribution and percentage of satisfied users. Superior performance of OPERA on several fronts compared to current schemes stems from its following features: 1) It preempts congestion instead of reacting to it; 2) it actuates more parameters than any current LB schemes thereby increasing system level capacity instead of just shifting it among cells; 3) while performing LB OPERA simultaneously maximizes residual capacity while incorporating throughput and coverage constraints; 4) it incorporates a load aware association strategy for ensuring conflict free operation of LB and CCO SON functions; 5) the ahead of time estimation of cell loads allows ample time for heuristics search algorithms to find LB solutions with high gain. [ABSTRACT FROM AUTHOR]

This paper presents an optimized situation-based power management strategy (SB-PMS) for multi-source electric vehicles. The presented SB-PMS is based on defining optimal control strategies offline for different vehicles states. In such control schemes, multiple characteristic variables, i.e. vehicle speed and power demand, are implemented to define vehicle states. The selection and discretization of characteristic variables have a significant impact on the solution optimality of SB-PMSs. Based on a comprehensive analysis of literature, the variables vehicle speed, power demand, speed dynamics, and on-board state-of-charge have been selected to investigate their impact solution optimality of SB-PMS. These variables are depicted as axes in a multi-dimensional space (grid-space), whereinto vehicle states can be mapped. Multiple structures of grid-space are generated based on different constellations of axes and different discretization levels for each axis. Comparative evaluation of grid-space structures is based on two aspects: first, the optimality of situated solutions related to vehicle states; second, the ability to develop predictive models using the statistics of state transitions in grid-space. Results analysis, compared to rule-based (RB) and dynamic programming (DP), reveals the significant impact of certain grid-space structures to achieve near optimal results in both energy saving and prediction accuracy at different trip conditions. [ABSTRACT FROM AUTHOR]

Short-term traffic prediction (STTP) is one of the most critical capabilities in Intelligent Transportation Systems (ITS), which can be used to support driving decisions, alleviate traffic congestion and improve transportation efficiency. However, STTP of large-scale road networks remains challenging due to the difficulties of effectively modeling the diverse traffic patterns by high-dimensional time series. Therefore, this paper proposes a framework that involves a deep clustering method for STTP in large-scale road networks. The deep clustering method is employed to supervise the representation learning in a visualized way from the large unlabeled dataset. More specifically, to fully exploit the traffic periodicity, the raw series is first divided into a number of sub-series for triplet generation. The convolutional neural networks (CNNs) with triplet loss are utilized to extract the features of shape by transforming the series into visual images. The shape-based representations are then used to cluster road segments into groups. Thereafter, a model sharing strategy is further proposed to build recurrent NNs-based predictions through group-based models (GBMs). GBM is built for a type of traffic patterns, instead of one road segment exclusively or all road segments uniformly. Our framework can not only significantly reduce the number of prediction models, but also improve their generalization by virtue of being trained on more diverse examples. Furthermore, the proposed framework over a selected road network in Beijing is evaluated. Experiment results show that the deep clustering method can effectively cluster the road segments and GBM can achieve comparable prediction accuracy against the IBM with less number of prediction models. [ABSTRACT FROM AUTHOR]

The Li-ion battery has emerged as the most prominent energy storage device and been widely applied in electric vehicles (EVs). To ensure the safety and performance of batteries, a battery management system (BMS) is incorporated in the EVs. However, how to predict and verify the BMS reliability remains an issue. We conduct the accelerated degradation test on the BMS and develop a semi-parametric framework to predict the reliability. A new measure based on similarity is further proposed for model comparison. A real data set of the BMS is used to verify the validity of the proposed method. A comparison with the other four models shows that our approach can generate a more similar trend to the original data, with 39.23% improvement compared with the best-performed model based on the proposed measure. Besides, the proposed model gives a prediction of lower bound of the reliable lifetime as 10.8 years, which is more reasonable compared with the results (i.e., around one year) given by the conventional models. [ABSTRACT FROM AUTHOR]

Imitation learning for the end-to-end autonomous driving has drawn renewed attention from academic communities. Current methods either only use images as the input, which will yield ambiguities when a vehicle approaches an intersection, or use additional command information to navigate the vehicle but inefficiently. Focusing on making the vehicle automatically drive along the given path, we propose a new and effective navigation command called as subgoal angle which does not require human participation and is calculated by the current position and subgoal of the ego-vehicle. Thus, the subgoal angle contains more information than the navigation command represented as a one-hot vector. Additionally, we propose a model architecture called as angle branched network that makes predictions based on the subgoal angle. In this network, the subgoal angle is not only used for extracting useful features but also for guiding the appropriate prediction layer to make predictions for both the steer angle and the throttle status (which controls the acceleration). Experiments are conducted in a three-dimensional urban simulator. Both quantitive and qualitive results show the effectiveness of the navigation command and the angle branched network. Moreover, the performance can be further boosted by integrating both semantic and depth information into the driving model. Especially by using the depth information, collisions with vehicles and pedestrians can be reduced. [ABSTRACT FROM AUTHOR]

Effective and timely prediction of bit error rate (BER) disturbances is an important means to improve the safety of balise uplink transmission. However, the joint influence of multiple non-stationary factors makes it difficult to predict the abrupt changes of disturbances. To solve this problem, an adaptive attention-based bidirectional stateful encoder-decoder model is proposed for BER disturbances prediction. By persistently transferring the forward and backward states in a batch-to-batch mapping manner, the bidirectional stateful encoder can enhance the capture ability for disturbances. Additionally, an adaptive stateful decoder is used to dynamically synthesize the temporal correlation and contextual summarization to reduce the error accumulation in long-term forecasting. Through effective disturbance discrimination and accumulative error elimination, the proposed model can improve the overall prediction accuracy for BER disturbances. Experiment results on real-world high-speed railway datasets show that the proposed model can achieve superior performance than the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Accurate prediction of remaining useful life (RUL) is of critical significance to the safety and reliability of lithium-ion batteries, which can offer efficient early warning signals for failure. Due to the complicated aging mechanism and realistic noise operation environment, direct predicting RUL with the measured data recorded in practice is challenging. In this work, a novel hybrid approach to forecasting battery future capacity and RUL is proposed by combining the improved variational modal decomposition (VMD), particle filter (PF) and gaussian process regression (GPR). The VMD algorithm is employed to decompose the recorded battery capacity data into an aging trend sequence and several residual sequences, where the number of modal layers is produced by the proposed posterior feedback confidence (PFC) method. The prediction models of PF and GPR algorithm are then respectively established to predict the aging trend sequence and residual sequences. Future capacity and RUL prediction experiments for battery pack and battery cells are performed to verify the effectiveness of the proposed hybrid approach, and the compared experiment results demonstrate that the proposed approach offers wide generality and reduced errors. [ABSTRACT FROM AUTHOR]

The remaining useful life (RUL) prediction of Lithium-ion batteries (LIBs) is of great importance to the health management of electric vehicles and hybrid electric vehicles. However, fluctuation and nonlinearity occur during battery degradation, resulting in difficulties in both model adaptability and RUL prediction accuracy. To face the challenge, we propose a sequence decomposition and deep learning integrated prognostic approach for the RUL prediction of LIBs. Complementary ensemble empirical mode decomposition and principal component analysis are applied to separate the local fluctuations and the global degradation trend from the battery aging data. The long short-term memory neural network combined with fully connected layers is designed as a transfer learning model. The hyperparameter optimization and finetuning strategy of the model is developed based on offline training data. In addition, to further realize the reasonable and effective LIB second-life applications, the RUL corresponding to different failure thresholds is predicted. The performance of the proposed integrated approach in degradation modeling and RUL prediction is evaluated on three publicly available LIB datasets with different degradation characteristics, as well as compared with other prediction algorithms under the same conditions. The illustrative results demonstrate that the proposed approach can achieve accurate, adaptive, and robust prediction for both capacity trajectory and RUL. [ABSTRACT FROM AUTHOR]

Online Ride-Hailing(ORH) service enables riders to enjoy on-demand transportation service with mobile devices. Despite the convenience of the service, privacy leakage could be caused because riders need to submit their locations during ride matching. To protect users’ privacy, much effort has been done to construct a privacy-enhanced ORH system in recent years. However, almost all of current privacy-preserving schemes match the nearest driver to the requesting rider, it is not the optimal matching strategy and unnecessary waste of travel distance may be caused in this way. To solve the problem, we propose a privacy-preserving ORH scheme named pRide, which minimizes the overall travel distance by matching the best driver in global perspective instead of the nearest driver in the local region. To find the best driver, we utilize a deep learning model to predict emergence of ride requests in various regions, and enables the ORH server to perform ride matching leveraging prediction results. Based on the framework, we also propose optimizations to enhance pRide. Moreover, to further defend inference attacks from riders, we propose a private comparing algorithm at the cost of a little communication overhead. Through theoretical analysis and experiments, we prove that pRide is able to decrease the overall travel distance securely and efficiently in ORH matching scenarios. [ABSTRACT FROM AUTHOR]

Reliable trajectory prediction methods are critical in providing predictive safety intelligence for vital decision making in intelligent transportation systems to further enhance the safety of drivers and passengers. To tackle complicated maneuvering and interactions between objects, learning-based algorithms were used to replace classic model-based trajectory prediction algorithms. However, most algorithms implicitly presume that they are executed at centralized processing units after gathering data from edge sensors and delivering the results back to the on-board units. This causes an increase in computation time and latency; thus, reducing the reaction time of the drivers. To reduce the computation time and latency and consider the robustness of local sensors, we propose a decentralized radar-dedicated framework with a deep-learning (DL) model, called predictive RadarNet, to predict future trajectories over binary range angle (RA) maps with a probabilistic representation according to the original radar RA maps for presenting the uncertainty of the estimated trajectories. In addition, to reduce the model size for low-complexity, we designed a prepossessing technique that can largely reduce the size of the input tensors without losing information. Moreover, we found that the functions of the DL-model consist of two operations: future inference of original radar RA maps and transformation to binary RA maps. Thus, we designed two models with different kernels that are suitable for dealing with the two operations. Simulations show that the proposed decentralized framework using predictive RadarNet can provide reliable prediction results with a low computation time. [ABSTRACT FROM AUTHOR]

This article proposes and evaluates a technique to predict the level of interference in wireless networks. We design a recursive predictor that estimates future interference values by filtering measured interference at a given location. The predictor's parameterization is done offline by translating the autocorrelation of interference into an autoregressive moving average (ARMA) representation. This ARMA model is inserted into a steady-state Kalman filter enabling nodes to predict with low computational effort. Results show a good accuracy of predicted values versus true values for relevant time horizons. Although the predictor is parameterized for Poisson-distributed nodes, Rayleigh fading, and fixed message lengths, a sensitivity analysis shows that it also tends to work well in more general network scenarios. Numerical examples for underlay device-to-device communications, a common wireless sensor technology, and coexistence scenarios of Wi-Fi and LTE illustrate its broad applicability. The predictor can be applied as part of interference management to improve medium access, scheduling, and radio resource allocation. [ABSTRACT FROM AUTHOR]

In Internet-of-Vehicles (IoV) ecosystems, intelligent toll gates (ITGs) connect nearby metropolitan cities through smart highways. At ITGs, existing solutions integrate blockchain (BC) and deep-learning schemes to leverage trusted and responsive analytics support for connected smart vehicles (CSVs) at ITGs. BC eliminates third-party intermediaries, and secures payments between vehicle owners (VO) and governing authorities (GA). Deep-Learning, on the other hand, facilitates accurate predictions for diverse and complex urban traffic conditions. However, due to fixed toll pricing schemes based on connected smart vehicles (CSV) type, VOs suffer from variable delays at different lanes due to dynamic congestion scenarios. To address the research gaps of such a fixed pricing schemes, we propose a BC-envisioned scheme DwaRa, that operates in three phases. In the first phase, future traffic is predicted based on Markov queues to balance the congestion at different lanes at ITGs efficiently. Then, we propose a novel spatially induced-long-short term memory (SI-LSTM) model to predict current traffic and weather based on historical repositories. Second, based on inputs by the Markov model, SI-LSTM, lane type, and vehicle type, a dynamic pricing algorithm is presented to improve the quality of experience (QoE) of the VO. Finally, based on dynamic price fixation between the VO and the GA, smart contracts (SCs) are executed and transactional data is secured through BC. The proposed scheme is compared against parameters like average mean-squared error (MSE), predicted traffic, scalability, interplanetary file system (IPFS) storage, computation (CC), and communication cost (CCM). At n = 100 test samples, and arrival rate β = 80, the obtained MSE is 0.0012, with a peak average value of 0.00526. The overall CC is 45.88 milliseconds (ms) and CCM is 53 bytes that indicate the proposed scheme efficacy against conventional approaches. [ABSTRACT FROM AUTHOR]

A novel method is proposed for forecasting the capacity of lithium-ion battery cells. The method uses a Gaussian Process Regression model, a machine learning framework. Besides the high prediction accuracy and robustness the proposed method possesses, the method offers other advantages, namely, it provides uncertainty information, and it has the capability to cross-correlate capacity trends between different battery cells. These two merits make the proposed method a very reliable and practical solution for applications that use battery cell packs with a large number of interconnected battery cells. The proposed method is derived, verified, and compared to benchmark methods on three experimental lithium-ion battery cell datasets. The results show the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

As the population keeps growing, traffic congestion happens more and more often. Consequently, travel time has become an important indicator of driving experience. Accurate travel time information helps drivers plan their route more wisely and thus effectively alleviate traffic congestion. In this research, we propose a vehicle travel time prediction model for freeway traffic. The data used in this research are derived from the traffic dataset of the Taiwan Freeway Bureau, and the travel time prediction is made for the Sun Yat-sen Freeway between Taipei and Hsinchu. First, the missing value of the raw data is imputed by Autoencoder. The data are then segmented according to time series and are used to build the prediction model. To effectively capture the hidden features required to predict the travel time for the vehicle traveling on the freeway, a deep learning architecture is adopted in our system, which includes the GRU neural network model, the XGBoost model, and the Hybrid model that combines the GRU and XGBoost through linear regression. To increase computational efficiency, the travel time predictions for consecutive toll gates every 5 minutes apart are pre-computed offline, so that the online travel time prediction of the whole trip can be obtained by simply summing up a few numbers. Experimental results based on actual traffic data show that the proposed system can achieve good performance in terms of prediction accuracy and execution time. [ABSTRACT FROM AUTHOR]

Autonomous driving has attracted significant attention in recent years. With the booming of artificial intelligence (AI), deep learning technologies have been applied to autonomous driving to help vehicles better perceive the environment. Besides the perceiving environment, predictive driving is another prominent smooth control and safe driving skill for human drivers. In this work, we develop a deep Monte Carlo Tree Search (deep-MCTS) control method for vision-based autonomous driving. Compared with existing deep learning-based autonomous driving control methods, our method can predict driving maneuvers to help improve the stability and performance of driving control. Two deep neural networks (DNNs) are employed for predicting action-state transformation and obtaining action-selection probabilities, respectively. The deep-MCTS utilizes the predicted information of the two DNNs and reconstructs multiple possible trajectories to predict driving maneuvers. An optimal trajectory is selected by the deep-MCTS based on both current road conditions and predicted driving maneuvers. The proposed method achieves high control stability by avoiding sharp turns and driving deviations. We implement our algorithm in the Udacity and Torcs self-driving environments. The testing results show that our algorithm achieves a significant improvement in training efficiency, the stability of steering control, and stability of driving trajectory compared to existing methods. [ABSTRACT FROM AUTHOR]

Reliable and timely delivery of data to mobile targets is a challenging problem in mobile ad-hoc networks, because of the opportunistic and unpredictable nature of the problem. Finding a complete series of mobile-to-mobile contacts and forwarding data toward targets within a designated deadline is even more difficult. We leverage unmanned aerial vehicles (UAVs) as message ferries to travel over highly probable rendezvous points for over-the-air delivery to target users. We propose Rendezvous, an opportunistic yet disciplined data delivery scheme based on trajectory prediction of users and UAV path planning. During the offline learning phase, we identify both temporal and spatial regularities of mobile users from real-world trajectories using sequence-wise clustering, and construct a compact yet well-summarized cluster signature that enables efficient search. In the data delivery phase, given recent movement history of users, UAVs perform their own distributed path planning collaboratively over time. Each UAV finds a chronological sequence of future visiting points where it can make consecutive data delivery efforts to users to overcome the uncertainty in mobility and ensure timely delivery. Real-world trace-driven simulation experiments demonstrate that Rendezvous achieves reliable and punctual data delivery using only few UAVs compared with existing algorithms. [ABSTRACT FROM AUTHOR]

Motion prediction for the leading vehicle is a critical task for connected autonomous vehicles. It provides a method to model the leading-following vehicle behavior and analysis their interactions. In this study, a joint time-series modeling approach for leading vehicle trajectory prediction considering different driving styles is proposed. The proposed method enables a precise and personalized trajectory prediction for the leading vehicle based on limited inter-vehicle communication signals, such as the vehicle speed and acceleration of the front vehicles. Three different driving styles are first recognized based on an unsupervised clustering algorithm, namely, Gaussian Mixture Model (GMM). The GMM generates a specific driving style for each vehicle based on the speed, acceleration, jerk, time, and space headway features of the leading vehicle. The feature importance of driving style recognition is also evaluated based on the Maximal Information Coefficient (MIC) algorithm. Then, a personalized joint time series modeling (JTSM) method based on the Long Short-Term Memory (LSTM) Recurrent Neural Network model (RNN) is proposed to predict the front vehicle trajectories. The JTSM contains a common LSTM layer and different fully connected regression layers for different driving styles. The proposed method is tested with the Next Generation Simulation (NGSIM) data on the US101, and I-80 highway dataset. The JTSM is tested for making predictions one to five seconds ahead. Results indicate that the proposed personalized JTSM approach shows a significant advantage over the baseline algorithms. [ABSTRACT FROM AUTHOR]

Many of today's vehicles come equipped with Advanced Driver Assistance Systems (ADAS). Proactive ADAS have the ability to predict short term driving situations. This provides drivers more time to take adequate actions to avoid or mitigate driving risks. In this work, we address the question of predicting drivers’ imminent maneuvers before they perform an actual steering operation. The proposed system uses deep recurrent neural networks to fuse the information regarding driver observation actions and the driving environment. With new data labeling methods and effective sequential modeling approaches, the system is able to predict with high accuracy driving maneuvers shortly before the actual steering operations. A set of experiments show that the proposed approach anticipates lane change maneuvers 1.50 seconds before cars start to yaw with an accuracy improved to 90.52% and anticipates turn maneuvers at intersections with green lights 2.53 seconds before cars start to yaw with an accuracy improved to 78.59%. We also show in this work how the system can be adapted for driving proficiency assessment. [ABSTRACT FROM AUTHOR]

Dynamic network topology can pose important challenges to communication and control protocols in networks of autonomous vehicles. For instance, maintaining connectivity is a key challenge in unmanned aerial vehicle (UAV) networks. However, tracking and computational resources of the observer module might not be sufficient for constant monitoring of all surrounding nodes in large-scale networks. In this article, we propose an optimal measurement policy for network topology monitoring under constrained resources. To this end, We formulate the localization of multiple objects in terms of linear networked systems and solve it using Kalman filtering with intermittent observation. The proposed policy includes two sequential steps. We first find optimal measurement attempt probabilities for each target using numerical optimization methods to assign the limited number of resources among targets. The optimal resource allocation follows a waterfall-like solution to assign more resources to targets with lower measurement success probability. This provides a $\text{10}\%$ to $\text{60}\%$ gain in prediction accuracy. The second step is finding optimal on-off patterns for measurement attempts for each target over time. We show that a regular measurement pattern that evenly distributed resources over time outperforms the two extreme cases of using all measurement resources either in the beginning or at the end of the measurement cycle. Our proof is based on characterizing the fixed-point solution of the error covariance matrix for regular patterns. Extensive simulation results confirm the optimality of the most alternating pattern with up to 10-fold prediction improvement for different scenarios. These two guidelines define a general policy for target tracking under constrained resources with applications to network topology prediction of autonomous systems. [ABSTRACT FROM AUTHOR]

Despite the increased interest in battery electric vehicles (BEV), limited range abilities unsettle customers, which is often related to range anxiety. A better understanding of energy consumption and the possibility to accurately predict the remaining battery energy along an upcoming route may help to reduce this stress perception by means of advanced in-vehicle information systems. Addressing the trend towards vehicles with on-board cloud communication and information systems, the present research focuses on electric powertrain consumption and speed profile forecasts. A meaningful prediction of a speed profile for a given route is a basic prerequisite for an accurate consumption forecast. This study proposes a methodology to derive such a speed profile from real-time traffic data obtained from HERE Technologies while considering individual driving style characteristics. Given the predicted speed profile, a detailed BEV consumption model which accounts for BEV specific energy management strategies and environmental factors is used to obtain a consumption forecast. Prediction uncertainties are analyzed and parameter sensitivities with respect to energy consumption are derived as a function of the route dependent mean vehicle speed. Within a field study with Mercedes Benz EQC experimental vehicles, covering thirty-two test cycles, it is shown that the proposed methodology can accurately predict energy consumption for long look-ahead horizons and significantly reduces the variance in prediction compared to a typical baseline strategy. [ABSTRACT FROM AUTHOR]

Spatio-temporal dependencies are the key to predicting the traffic parameters of an urban arterial network. However, their inclusion in forecasting traffic states has been hampered due to both the absence of a robust model and the computational burden. Recently, an innovative way to tackle the problem was developed by adopting a convolutional neural network (CNN) to deal with map images representing traffic states. Unlike previous studies that utilized map images only for input, the present study adopted images for both the input and the output of a CNN model to predict traffic speeds. The results show that the performance of the proposed model based on image-to-image learning is superior to that of the existing models. [ABSTRACT FROM AUTHOR]

While vehicular congestion is very often defined in terms of aggregate parameters, such as traffic volume and lane occupancies, based on recent findings, the interpretation that receives most credit is that of a state of a road where traversing vehicles experience a delay exceeding the maximum value typically incurred under light or free-flow traffic conditions. We here propose a new definition according to which a road is in a congested state (be it high or low) only when the likelihood of finding it in the same congested state is high in the near future. Based on this new definition, we devised an algorithm that, exploiting probe vehicles, for any given road 1) identifies if it is congested or not and 2) provides the estimation that a current congested state will last for at least a given time interval. Unlike any other existing approach, an important advantage of ours is that it can generally be applied to any type of road, as it neither needs any a priori knowledge nor requires the estimation of any road parameter (e.g., number of lanes, traffic light cycle, etc.). Further, it allows performing short-term traffic congestion forecasting for any given road. We present several field trials gathered on different urban roads whose empirical results confirm the validity of our approach. [ABSTRACT FROM PUBLISHER]

The high mobility of vehicles and the unreliable wireless communication significantly degrade the performance of data access in vehicular ad hoc networks (VANETs). To address this problem, we propose a novel vehicle-platoon-aware data access solution called V-PADA. In V-PADA, vehicles contribute part of their buffers to replicate data for others in the same platoon and share data with them. When a vehicle leaves the platoon, it prefetches interested data and transfers its buffered data to other vehicles in advance so that they can still access the data after it leaves. To achieve this goal, V-PADA consists of two components: First, a vehicle-platooning protocol is designed to identify platoon formation and predict platoon splits. We use stochastic time series analysis to detect platoon and mobility anomalies and further introduce a two-step split prediction method to reduce the false alarm rate due to road curves. Second, a data management component is designed to guide platoon members to replicate and prefetch the most suitable data so that both high data availability and low data access overhead can be achieved. Extensive simulation results show that V-PADA can effectively improve the data access performance in VANETs. [ABSTRACT FROM PUBLISHER]