2022 IEEE Intelligent Vehicles Symposium (IV), Jun 5, 2022
While privacy concerns entice connected and automated vehicles to incorporate on-board federated ... more While privacy concerns entice connected and automated vehicles to incorporate on-board federated learning (FL) solutions, an integrated vehicle-to-everything communication with heterogeneous computation power aware learning platform is urgently necessary to make it a reality. Motivated by this, we propose a novel mobility, communication and computation aware online FL platform that uses on-road vehicles as learning agents. Thanks to the advanced features of modern vehicles, the on-board sensors can collect data as vehicles travel along their trajectories, while the on-board processors can train machine learning models using the collected data. To take the high mobility of vehicles into account, we consider the delay as a learning parameter and restrict it to be less than a tolerable threshold. To satisfy this threshold, the central server accepts partially trained models, the distributed roadside units (a) perform downlink multicast beamforming to minimize global model distribution delay and (b) allocate optimal uplink radio resources to minimize local model offloading delay, and the vehicle agents conduct heterogeneous local model training. Using real-world vehicle trace datasets, we validate our FL solutions. Simulation shows that the proposed integrated FL platform is robust and outperforms baseline models. With reasonable local training episodes, it can effectively satisfy all constraints and deliver near ground truth multihorizon velocity and vehicle-specific power predictions.
One of the most essential prerequisites behind a successful task execution of a team of agents is... more One of the most essential prerequisites behind a successful task execution of a team of agents is to accurately estimate and track their poses. We consider a cooperative multiagent positioning problem where each agent performs singleagent positioning until it encounters some other agent. Upon the encounter, the two agents measure their relative pose, and exchange particle clouds representing their poses. We propose a cooperative positioning algorithm which fuses the received information with the locally available measurements and infers an agent's pose within Bayesian framework. The algorithm is scalable to multiple agents, has relatively low computational complexity, admits decentralized implementation across agents, and imposes relatively mild requirements on communication coverage and bandwidth. The experiments indicate that the proposed algorithm considerably improves single-agent positioning accuracy, reduces the convergence time of a particle cloud and, unlike its single-agent positioning counterpart, exhibits immunity to an impeding feature-scarce and symmetric environment layout.
This paper investigates a method to improve performance of diffusive molecular communications bet... more This paper investigates a method to improve performance of diffusive molecular communications between biologically-enabled nanomachines in in-vivo aqueous environment. The proposed method exploits periodic flow, e.g., induced by repeated heart pumping. We make an analysis of channel impulse response (CIR) for such drift-diffusion fluid systems. In order to take the cyclic CIR into account, the proposed method optimizes the release timing and size of information molecules so that highest equalization gain can be achieved. We reveal that error rate performance can be significantly improved with adaptive molecule loading by taking care of the cyclic CIR.
This paper considers localization with 28-GHz millimeter wave (mmWave) channel measurements in an... more This paper considers localization with 28-GHz millimeter wave (mmWave) channel measurements in an outdoor environment. Compared with mmWave channel characterization by real-world experiments, localization using real-world 28-GHz experiments has been much less reported. To fill the gap, we report here a preliminary field study of using real-world 28-GHz channel frequency responses (CFR) with a wide bandwidth of 500 MHz for outdoor localization. Specifically, we employ a fingerprinting-based localization approach by registering the location information using multiple wideband CFR measurements and exploring the transmit-receive antenna polarization. Our experimental results demonstrate that, with a full bandwidth of 500 MHz, a correlation-based fingerprinting localization can fully identify all 8 locations with a 1-m separation without any error. The probability of successful localization reduces to 97% or 91.5%, respectively, when two or just one narrowband (< 15 MHz) CFR measurements are used for the training dataset.
Complementary to the fine-grained channel state information (CSI) and coarse-grained received sig... more Complementary to the fine-grained channel state information (CSI) and coarse-grained received signal strength indicator (RSSI) measurements, the mid-grained spatial beam attributes (e.g., beam SNR) during the millimeter-wave (mmWave) beam training phase were recently repurposed for Wi-Fi sensing applications such as human activity recognition and indoor localization. This paper proposes a multi-band Wi-Fi sensing framework to fuse features from both CSI at sub-7 GHz bands and the mid-grained beam SNR at 60 GHz with feature granularity matching that pairs feature maps from the CSI and beam SNR at different granularity levels with learnable weights. To address the issue of limited labeled training data, we propose to pre-train an autoencoder-based multi-band Wi-Fi fusion network in an unsupervised fashion. For specific sensing tasks, separate sensing heads can be attached to the pre-trained fusion network with fine-tuning. The proposed framework is thoroughly validated by three in-house experimental datasets: 1) pose recognition; 2) occupancy sensing; and 3) indoor localization. Comparison to a list of baseline methods demonstrates the effectiveness of granularity matching. Ablation study is performed as a function of the number of labeled data, latent space dimension, and finetuning learning rates.
This paper proposes a learning-based approach to mitigate the shadow effect in the pixel domain f... more This paper proposes a learning-based approach to mitigate the shadow effect in the pixel domain for Terahertz Time-Domain Spectroscopy (THz-TDS) multi-layer imaging. Compared with model-based approaches, this learning-based approach requires no prior knowledge of material properties of the sample. Preliminary simulations confirm the effectiveness of the proposed method.
In this paper, we propose a variational Bayesian inference approach for a low-complexity symbol d... more In this paper, we propose a variational Bayesian inference approach for a low-complexity symbol detection for massive MIMO systems with symbol-dependent transmit-side impairments. This study is motivated by observations that realworld communication transceivers are often affected by the hardware impairments, such as non-linearities of power amplifiers, I/Q imbalance, phase drifts due to non-ideal oscillators, and carrier frequency offsets. Particularly, symbol-dependent perturbations are fully accounted into the designed hierarchical signal model as unknown model parameters. The developed variational Bayesian symbol detector is able to learn the unknown perturbations in an iterative fashion. Numerical evaluation confirms the effectiveness of the proposed approach.
State-of-the-art base stations can be equipped with a massively large number of antenna elements,... more State-of-the-art base stations can be equipped with a massively large number of antenna elements, often several hundreds of elements, thanks to the rapid advancement of wideband radio-frequency (RF) analog circuits and compact antenna design techniques. With massive antenna systems, a relatively large number of users can be served at the same time by means of analog and digital beamforming and spatial multiplexing. We investigate such a largescale multiuser multipleinput multiple-output (MU-MIMO) wireless system employing an orthogonal frequency-division multiplexing (OFDM)-based downlink transmission scheme. The use of OFDM causes a high peak-to-average power ratio (PAPR), which usually calls for expensive and power-inefficient RF components at the base station. In this paper, we propose a nullspace vector perturbation (VP) which integrates both nonlinear lattice and linear subspace precoding approaches. By exploiting high degrees of freedom available in massive MU-MIMO OFDM systems, the signal PAPR can be significantly reduced with the proposed method. We also introduce a Gaussian process (GP) regression approach to be robust against the imperfect channel knowledge, which is required for the VP operation, in time-varying fading channels. Our analysis of outage capacity reveals that the proposed VP with GP regression offers a significant improvement in sum-rate spectral efficiency while reducing the PAPR.
Effective decoding of wireless signals requires various parameter acquisition techniques includin... more Effective decoding of wireless signals requires various parameter acquisition techniques including user activity detection, synchronization, channel estimation, and channel equalization. In traditional systems, these unknown, underlying parameters of the communication channel are individually estimated. This work proposes a novel joint estimation process applying deep learning. The proposed method shows superior performance to traditional methods and is further able to find the multiplicity of collisions, handle synchronization and channel estimation in the case of colliding non-orthogonal transmissions, and is able to discover superior preamble sequences using an auto-encoder structure. The proposed method is intended for decoding transmissions at the base-station in a massive connectivity scenario with many lowcomplexity devices operating concurrently. Excellent performance is demonstrated in estimating Time-of-Arrival (ToA), Carrier-Frequency Offset (CFO), channel gain and collision multiplicity from a received mixture of transmissions using the random access preamble structure structure of the NB-IoT standard. The proposed estimation scheme, employing a convolutional neural network (CNN), achieves a ToA Root-Mean-Square Error (RMSE) of 2.88 µs and a CFO RMSE of 3.44 Hz at 10 dB Signalto-Noise Ratio (SNR), whereas a conventional estimator using two cascaded stages have RM-SEs of 16.20 µs and 7.98 Hz, respectively.
A novel multi-task federated learning (FL) framework is proposed in this paper to optimize the tr... more A novel multi-task federated learning (FL) framework is proposed in this paper to optimize the traffic prediction models without sharing the collected data among traffic stations. In particular, a divisive hierarchical clustering is first introduced to partition the collected traffic data at each station into different clusters. The FL is then implemented to collaboratively train the learning model for each cluster of local data distributed across the stations. Using the multi-task FL framework, the route planning is studied where the road map is modeled as a time-dependent graph and a modified A* algorithm is used to determine the route with the shortest traveling time. Simulation results showcase the prediction accuracy improvement of the proposed multi-task FL framework over two baseline schemes. The simulation results also show that, when using the multi-task FL framework in the route planning, an accurate traveling time can be estimated and an effective route can be selected.
With the increasing development of the IoT applications, heterogeneous wireless networks may coex... more With the increasing development of the IoT applications, heterogeneous wireless networks may coexist. IEEE 802.11ah and IEEE 802.15.4g are two wireless technologies designed for IoT applications. 802.11ah is primarily developed for outdoor applications such as smart city and 802.15.4g is principally developed for large scale outdoor process control applications such as smart utility network. Both technologies have communication range up to 1000 meters. Therefore, 802.11ah network and 802.15.4g network are likely to coexist. Our simulation results show that 802.11ah network can severely interfere with 802.15.4g network since 802.11ah devices are more aggressive than 802.15.4g devices in wireless medium access contention. This capability heterogeneity can lead to significant packet loss in 802.15.4g network. Due to asymmetrical features such as modulation scheme and packet structure, devices in different networks can not understand each other. Thus, the self-transmission control mechanism is needed for more aggressive 802.11ah devices. This paper proposes a learning based self-transmission control method for 802.11ah devices to improve their coexistence with 802.15.4g devices. Using the proposed self-transmission control technique, 802.11ah devices predict the packet transmission of 802.15.4g devices and postpone their transmissions to avoid interference.
This paper proposes innovative anomaly detection technologies for manufacturing systems. We combi... more This paper proposes innovative anomaly detection technologies for manufacturing systems. We combine the event ordering relationship based structuring technique and the deep neural networks to develop the structured neural networks for anomaly detection. The event ordering relationship based neural network structuring process is performed before neural network training process and determines important neuron connections and weight initialization. It reduces the complexity of the neural networks and can improve anomaly detection accuracy. The structured time delay neural network (TDNN) is introduced for anomaly detection via supervised learning. To detect anomaly through unsupervised learning, we propose the structured autoencoder. The proposed structured neural networks outperform the unstructured neural networks in terms of anomaly detection accuracy and can reduce test error by 20%. Compared with popular methods such as one-class SVM, decision trees, and distance-based algorithms, our structured neural networks can reduce anomaly detection misclassification error by as much as 64%.
We analyze brain waves acquired through a consumer-grade EEG device to investigate its capabiliti... more We analyze brain waves acquired through a consumer-grade EEG device to investigate its capabilities for user identification and authentication. First, we show the statistical significance of the P300 component in event-related potential (ERP) data from 14-channel EEGs across 25 subjects. We then apply a variety of machine learning techniques, comparing the user identification performance of various different combinations of a dimensionality reduction technique followed by a classification algorithm. Experimental results show that an identification accuracy of 72% can be achieved using only a single 800 ms ERP epoch. In addition, we demonstrate that the user identification accuracy can be significantly improved to more than 96.7% by joint classification of multiple epochs.
Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, Jul 1, 2022
Graph neural networks have been widely used for a variety of learning tasks. Link prediction is a... more Graph neural networks have been widely used for a variety of learning tasks. Link prediction is a relatively under-studied graph learning task, with current state-of-the-art models based on oneor two-layer shallow graph auto-encoder (GAE) architectures. In this paper, we overcome the limitation of current methods for link prediction of non-Euclidean network data, which can only use shallow GAEs and variational GAEs. Our proposed methods innovatively incorporate standard auto-encoders (AEs) into the architectures of GAEs to capitalize on the intimate coupling of node and edge information in complex network data. Empirically, extensive experiments on various datasets demonstrate the competitive performance of our proposed approach. Theoretically, we prove that our deep extensions can inclusively express multiple polynomial filters with different orders. The codes of this paper are available at https://github.com/xinxingwu-uk/DGAE.
We present a method for separating collided signals from multiple users in the presence of strong... more We present a method for separating collided signals from multiple users in the presence of strong and wideband interference/jamming signal. More specifically, we consider a massive connectivity setup where few, out of a large number of users, equipped with spreading codes, synchronously transmit symbols. The received signal is a noisy mixture of symbols transmitted through users' flat fading channels, impaired by fast frequency hopping jamming signal of relatively large power. In the absence of any conventional technique suitable for the considered setup, we propose a "model-driven" deep learning method, based on convolution neural network, to suppress jamming signal from the received signal, and detect active users together with their transmitted symbols. A numerical study of the proposed method confirms its effectiveness in scenarios where classical techniques fail. As such, in a two user scenario with wideband jamming signal of power 20 dB above the power any active user, the proposed algorithm achieves error rates 10 −2 for a wide range of AWGN variances.
Wireless networked control system is gaining momentum in industrial cyber-physical systems, e.g.,... more Wireless networked control system is gaining momentum in industrial cyber-physical systems, e.g., smart factory. Suffering from limited bandwidth and nondeterministic link quality, a critical challenge in its deployment is how to optimize the closed-loop control system performance as well as maintain stability. In order to bridge the gap between network design and control system performance, we propose an optimal dynamic scheduling strategy that optimizes performance of multi-loop control systems by allocating network resources based on predictions of both link quality and control performance at run-time. The optimal dynamic scheduling strategy boils down to solving a nonlinear integer programming problem, which is further relaxed to a linear programming problem. The proposed strategy provably renders the closed-loop system meansquare stable under mild assumptions. Its efficacy is demonstrated by simulating a four-loop control system over an IEEE 802.15.4 wireless network simulator-TOSSIM. Simulation results show that the optimal dynamic scheduling can enhance control system performance and adapt to both constant and variable network background noises as well as physical disturbance.
This paper considers mutual interference mitigation among automotive radars using frequency-modul... more This paper considers mutual interference mitigation among automotive radars using frequency-modulated continuous wave (FMCW) signal and multiple-input multiple-output (MIMO) virtual arrays. For the first time, we derive a general interference signal model that fully accounts for not only the time-frequency incoherence, e.g., different FMCW configuration parameters and time offsets, but also the slow-time code MIMO incoherence and array configuration differences between the victim and interfering radars. Along with a standard MIMO-FMCW object signal model, we turn the interference mitigation into a spatial-domain object detection under incoherent MIMO-FMCW interference described by the explicit interference signal model, and propose a constant false alarm rate (CFAR) detector. More specifically, the proposed detector exploits the structural property of the derived interference model at both transmit and receive steering vector space. We also derive analytical closedform expressions for probabilities of detection and false alarm. Performance evaluation using both synthetic-level and phased array system-level simulation confirms the effectiveness of our proposed detector over selected baseline methods.
This paper proposes an anomaly detection algorithm for a factory automation system, which jointly... more This paper proposes an anomaly detection algorithm for a factory automation system, which jointly performs data pre-processing and time-delay autoencoder (TDAE) with a hybrid loss function. The source data are pre-processed by digital filters before feeding into a TDAE for anomaly detection. The digital filters extract analog signals from a variety of frequency bands to facilitate identifying anomalies. The pre-processed data then takes time-delay reform to explore temporal relationship of data signals. In addition, two anomaly diagnosis algorithms, a statistical based method and an autoencoder based method, are presented. Numerical results show that time-delay reform can improve the anomaly detection accuracy compared to the conventional autoencoder. Data pre-processing can further improve the anomaly detection accuracy. Moreover, we confirm that our anomaly diagnosis algorithms outperform traditional method that does not perform data pre-processing and time-delay reform.
2022 IEEE Intelligent Vehicles Symposium (IV), Jun 5, 2022
While privacy concerns entice connected and automated vehicles to incorporate on-board federated ... more While privacy concerns entice connected and automated vehicles to incorporate on-board federated learning (FL) solutions, an integrated vehicle-to-everything communication with heterogeneous computation power aware learning platform is urgently necessary to make it a reality. Motivated by this, we propose a novel mobility, communication and computation aware online FL platform that uses on-road vehicles as learning agents. Thanks to the advanced features of modern vehicles, the on-board sensors can collect data as vehicles travel along their trajectories, while the on-board processors can train machine learning models using the collected data. To take the high mobility of vehicles into account, we consider the delay as a learning parameter and restrict it to be less than a tolerable threshold. To satisfy this threshold, the central server accepts partially trained models, the distributed roadside units (a) perform downlink multicast beamforming to minimize global model distribution delay and (b) allocate optimal uplink radio resources to minimize local model offloading delay, and the vehicle agents conduct heterogeneous local model training. Using real-world vehicle trace datasets, we validate our FL solutions. Simulation shows that the proposed integrated FL platform is robust and outperforms baseline models. With reasonable local training episodes, it can effectively satisfy all constraints and deliver near ground truth multihorizon velocity and vehicle-specific power predictions.
One of the most essential prerequisites behind a successful task execution of a team of agents is... more One of the most essential prerequisites behind a successful task execution of a team of agents is to accurately estimate and track their poses. We consider a cooperative multiagent positioning problem where each agent performs singleagent positioning until it encounters some other agent. Upon the encounter, the two agents measure their relative pose, and exchange particle clouds representing their poses. We propose a cooperative positioning algorithm which fuses the received information with the locally available measurements and infers an agent's pose within Bayesian framework. The algorithm is scalable to multiple agents, has relatively low computational complexity, admits decentralized implementation across agents, and imposes relatively mild requirements on communication coverage and bandwidth. The experiments indicate that the proposed algorithm considerably improves single-agent positioning accuracy, reduces the convergence time of a particle cloud and, unlike its single-agent positioning counterpart, exhibits immunity to an impeding feature-scarce and symmetric environment layout.
This paper investigates a method to improve performance of diffusive molecular communications bet... more This paper investigates a method to improve performance of diffusive molecular communications between biologically-enabled nanomachines in in-vivo aqueous environment. The proposed method exploits periodic flow, e.g., induced by repeated heart pumping. We make an analysis of channel impulse response (CIR) for such drift-diffusion fluid systems. In order to take the cyclic CIR into account, the proposed method optimizes the release timing and size of information molecules so that highest equalization gain can be achieved. We reveal that error rate performance can be significantly improved with adaptive molecule loading by taking care of the cyclic CIR.
This paper considers localization with 28-GHz millimeter wave (mmWave) channel measurements in an... more This paper considers localization with 28-GHz millimeter wave (mmWave) channel measurements in an outdoor environment. Compared with mmWave channel characterization by real-world experiments, localization using real-world 28-GHz experiments has been much less reported. To fill the gap, we report here a preliminary field study of using real-world 28-GHz channel frequency responses (CFR) with a wide bandwidth of 500 MHz for outdoor localization. Specifically, we employ a fingerprinting-based localization approach by registering the location information using multiple wideband CFR measurements and exploring the transmit-receive antenna polarization. Our experimental results demonstrate that, with a full bandwidth of 500 MHz, a correlation-based fingerprinting localization can fully identify all 8 locations with a 1-m separation without any error. The probability of successful localization reduces to 97% or 91.5%, respectively, when two or just one narrowband (< 15 MHz) CFR measurements are used for the training dataset.
Complementary to the fine-grained channel state information (CSI) and coarse-grained received sig... more Complementary to the fine-grained channel state information (CSI) and coarse-grained received signal strength indicator (RSSI) measurements, the mid-grained spatial beam attributes (e.g., beam SNR) during the millimeter-wave (mmWave) beam training phase were recently repurposed for Wi-Fi sensing applications such as human activity recognition and indoor localization. This paper proposes a multi-band Wi-Fi sensing framework to fuse features from both CSI at sub-7 GHz bands and the mid-grained beam SNR at 60 GHz with feature granularity matching that pairs feature maps from the CSI and beam SNR at different granularity levels with learnable weights. To address the issue of limited labeled training data, we propose to pre-train an autoencoder-based multi-band Wi-Fi fusion network in an unsupervised fashion. For specific sensing tasks, separate sensing heads can be attached to the pre-trained fusion network with fine-tuning. The proposed framework is thoroughly validated by three in-house experimental datasets: 1) pose recognition; 2) occupancy sensing; and 3) indoor localization. Comparison to a list of baseline methods demonstrates the effectiveness of granularity matching. Ablation study is performed as a function of the number of labeled data, latent space dimension, and finetuning learning rates.
This paper proposes a learning-based approach to mitigate the shadow effect in the pixel domain f... more This paper proposes a learning-based approach to mitigate the shadow effect in the pixel domain for Terahertz Time-Domain Spectroscopy (THz-TDS) multi-layer imaging. Compared with model-based approaches, this learning-based approach requires no prior knowledge of material properties of the sample. Preliminary simulations confirm the effectiveness of the proposed method.
In this paper, we propose a variational Bayesian inference approach for a low-complexity symbol d... more In this paper, we propose a variational Bayesian inference approach for a low-complexity symbol detection for massive MIMO systems with symbol-dependent transmit-side impairments. This study is motivated by observations that realworld communication transceivers are often affected by the hardware impairments, such as non-linearities of power amplifiers, I/Q imbalance, phase drifts due to non-ideal oscillators, and carrier frequency offsets. Particularly, symbol-dependent perturbations are fully accounted into the designed hierarchical signal model as unknown model parameters. The developed variational Bayesian symbol detector is able to learn the unknown perturbations in an iterative fashion. Numerical evaluation confirms the effectiveness of the proposed approach.
State-of-the-art base stations can be equipped with a massively large number of antenna elements,... more State-of-the-art base stations can be equipped with a massively large number of antenna elements, often several hundreds of elements, thanks to the rapid advancement of wideband radio-frequency (RF) analog circuits and compact antenna design techniques. With massive antenna systems, a relatively large number of users can be served at the same time by means of analog and digital beamforming and spatial multiplexing. We investigate such a largescale multiuser multipleinput multiple-output (MU-MIMO) wireless system employing an orthogonal frequency-division multiplexing (OFDM)-based downlink transmission scheme. The use of OFDM causes a high peak-to-average power ratio (PAPR), which usually calls for expensive and power-inefficient RF components at the base station. In this paper, we propose a nullspace vector perturbation (VP) which integrates both nonlinear lattice and linear subspace precoding approaches. By exploiting high degrees of freedom available in massive MU-MIMO OFDM systems, the signal PAPR can be significantly reduced with the proposed method. We also introduce a Gaussian process (GP) regression approach to be robust against the imperfect channel knowledge, which is required for the VP operation, in time-varying fading channels. Our analysis of outage capacity reveals that the proposed VP with GP regression offers a significant improvement in sum-rate spectral efficiency while reducing the PAPR.
Effective decoding of wireless signals requires various parameter acquisition techniques includin... more Effective decoding of wireless signals requires various parameter acquisition techniques including user activity detection, synchronization, channel estimation, and channel equalization. In traditional systems, these unknown, underlying parameters of the communication channel are individually estimated. This work proposes a novel joint estimation process applying deep learning. The proposed method shows superior performance to traditional methods and is further able to find the multiplicity of collisions, handle synchronization and channel estimation in the case of colliding non-orthogonal transmissions, and is able to discover superior preamble sequences using an auto-encoder structure. The proposed method is intended for decoding transmissions at the base-station in a massive connectivity scenario with many lowcomplexity devices operating concurrently. Excellent performance is demonstrated in estimating Time-of-Arrival (ToA), Carrier-Frequency Offset (CFO), channel gain and collision multiplicity from a received mixture of transmissions using the random access preamble structure structure of the NB-IoT standard. The proposed estimation scheme, employing a convolutional neural network (CNN), achieves a ToA Root-Mean-Square Error (RMSE) of 2.88 µs and a CFO RMSE of 3.44 Hz at 10 dB Signalto-Noise Ratio (SNR), whereas a conventional estimator using two cascaded stages have RM-SEs of 16.20 µs and 7.98 Hz, respectively.
A novel multi-task federated learning (FL) framework is proposed in this paper to optimize the tr... more A novel multi-task federated learning (FL) framework is proposed in this paper to optimize the traffic prediction models without sharing the collected data among traffic stations. In particular, a divisive hierarchical clustering is first introduced to partition the collected traffic data at each station into different clusters. The FL is then implemented to collaboratively train the learning model for each cluster of local data distributed across the stations. Using the multi-task FL framework, the route planning is studied where the road map is modeled as a time-dependent graph and a modified A* algorithm is used to determine the route with the shortest traveling time. Simulation results showcase the prediction accuracy improvement of the proposed multi-task FL framework over two baseline schemes. The simulation results also show that, when using the multi-task FL framework in the route planning, an accurate traveling time can be estimated and an effective route can be selected.
With the increasing development of the IoT applications, heterogeneous wireless networks may coex... more With the increasing development of the IoT applications, heterogeneous wireless networks may coexist. IEEE 802.11ah and IEEE 802.15.4g are two wireless technologies designed for IoT applications. 802.11ah is primarily developed for outdoor applications such as smart city and 802.15.4g is principally developed for large scale outdoor process control applications such as smart utility network. Both technologies have communication range up to 1000 meters. Therefore, 802.11ah network and 802.15.4g network are likely to coexist. Our simulation results show that 802.11ah network can severely interfere with 802.15.4g network since 802.11ah devices are more aggressive than 802.15.4g devices in wireless medium access contention. This capability heterogeneity can lead to significant packet loss in 802.15.4g network. Due to asymmetrical features such as modulation scheme and packet structure, devices in different networks can not understand each other. Thus, the self-transmission control mechanism is needed for more aggressive 802.11ah devices. This paper proposes a learning based self-transmission control method for 802.11ah devices to improve their coexistence with 802.15.4g devices. Using the proposed self-transmission control technique, 802.11ah devices predict the packet transmission of 802.15.4g devices and postpone their transmissions to avoid interference.
This paper proposes innovative anomaly detection technologies for manufacturing systems. We combi... more This paper proposes innovative anomaly detection technologies for manufacturing systems. We combine the event ordering relationship based structuring technique and the deep neural networks to develop the structured neural networks for anomaly detection. The event ordering relationship based neural network structuring process is performed before neural network training process and determines important neuron connections and weight initialization. It reduces the complexity of the neural networks and can improve anomaly detection accuracy. The structured time delay neural network (TDNN) is introduced for anomaly detection via supervised learning. To detect anomaly through unsupervised learning, we propose the structured autoencoder. The proposed structured neural networks outperform the unstructured neural networks in terms of anomaly detection accuracy and can reduce test error by 20%. Compared with popular methods such as one-class SVM, decision trees, and distance-based algorithms, our structured neural networks can reduce anomaly detection misclassification error by as much as 64%.
We analyze brain waves acquired through a consumer-grade EEG device to investigate its capabiliti... more We analyze brain waves acquired through a consumer-grade EEG device to investigate its capabilities for user identification and authentication. First, we show the statistical significance of the P300 component in event-related potential (ERP) data from 14-channel EEGs across 25 subjects. We then apply a variety of machine learning techniques, comparing the user identification performance of various different combinations of a dimensionality reduction technique followed by a classification algorithm. Experimental results show that an identification accuracy of 72% can be achieved using only a single 800 ms ERP epoch. In addition, we demonstrate that the user identification accuracy can be significantly improved to more than 96.7% by joint classification of multiple epochs.
Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, Jul 1, 2022
Graph neural networks have been widely used for a variety of learning tasks. Link prediction is a... more Graph neural networks have been widely used for a variety of learning tasks. Link prediction is a relatively under-studied graph learning task, with current state-of-the-art models based on oneor two-layer shallow graph auto-encoder (GAE) architectures. In this paper, we overcome the limitation of current methods for link prediction of non-Euclidean network data, which can only use shallow GAEs and variational GAEs. Our proposed methods innovatively incorporate standard auto-encoders (AEs) into the architectures of GAEs to capitalize on the intimate coupling of node and edge information in complex network data. Empirically, extensive experiments on various datasets demonstrate the competitive performance of our proposed approach. Theoretically, we prove that our deep extensions can inclusively express multiple polynomial filters with different orders. The codes of this paper are available at https://github.com/xinxingwu-uk/DGAE.
We present a method for separating collided signals from multiple users in the presence of strong... more We present a method for separating collided signals from multiple users in the presence of strong and wideband interference/jamming signal. More specifically, we consider a massive connectivity setup where few, out of a large number of users, equipped with spreading codes, synchronously transmit symbols. The received signal is a noisy mixture of symbols transmitted through users' flat fading channels, impaired by fast frequency hopping jamming signal of relatively large power. In the absence of any conventional technique suitable for the considered setup, we propose a "model-driven" deep learning method, based on convolution neural network, to suppress jamming signal from the received signal, and detect active users together with their transmitted symbols. A numerical study of the proposed method confirms its effectiveness in scenarios where classical techniques fail. As such, in a two user scenario with wideband jamming signal of power 20 dB above the power any active user, the proposed algorithm achieves error rates 10 −2 for a wide range of AWGN variances.
Wireless networked control system is gaining momentum in industrial cyber-physical systems, e.g.,... more Wireless networked control system is gaining momentum in industrial cyber-physical systems, e.g., smart factory. Suffering from limited bandwidth and nondeterministic link quality, a critical challenge in its deployment is how to optimize the closed-loop control system performance as well as maintain stability. In order to bridge the gap between network design and control system performance, we propose an optimal dynamic scheduling strategy that optimizes performance of multi-loop control systems by allocating network resources based on predictions of both link quality and control performance at run-time. The optimal dynamic scheduling strategy boils down to solving a nonlinear integer programming problem, which is further relaxed to a linear programming problem. The proposed strategy provably renders the closed-loop system meansquare stable under mild assumptions. Its efficacy is demonstrated by simulating a four-loop control system over an IEEE 802.15.4 wireless network simulator-TOSSIM. Simulation results show that the optimal dynamic scheduling can enhance control system performance and adapt to both constant and variable network background noises as well as physical disturbance.
This paper considers mutual interference mitigation among automotive radars using frequency-modul... more This paper considers mutual interference mitigation among automotive radars using frequency-modulated continuous wave (FMCW) signal and multiple-input multiple-output (MIMO) virtual arrays. For the first time, we derive a general interference signal model that fully accounts for not only the time-frequency incoherence, e.g., different FMCW configuration parameters and time offsets, but also the slow-time code MIMO incoherence and array configuration differences between the victim and interfering radars. Along with a standard MIMO-FMCW object signal model, we turn the interference mitigation into a spatial-domain object detection under incoherent MIMO-FMCW interference described by the explicit interference signal model, and propose a constant false alarm rate (CFAR) detector. More specifically, the proposed detector exploits the structural property of the derived interference model at both transmit and receive steering vector space. We also derive analytical closedform expressions for probabilities of detection and false alarm. Performance evaluation using both synthetic-level and phased array system-level simulation confirms the effectiveness of our proposed detector over selected baseline methods.
This paper proposes an anomaly detection algorithm for a factory automation system, which jointly... more This paper proposes an anomaly detection algorithm for a factory automation system, which jointly performs data pre-processing and time-delay autoencoder (TDAE) with a hybrid loss function. The source data are pre-processed by digital filters before feeding into a TDAE for anomaly detection. The digital filters extract analog signals from a variety of frequency bands to facilitate identifying anomalies. The pre-processed data then takes time-delay reform to explore temporal relationship of data signals. In addition, two anomaly diagnosis algorithms, a statistical based method and an autoencoder based method, are presented. Numerical results show that time-delay reform can improve the anomaly detection accuracy compared to the conventional autoencoder. Data pre-processing can further improve the anomaly detection accuracy. Moreover, we confirm that our anomaly diagnosis algorithms outperform traditional method that does not perform data pre-processing and time-delay reform.
Uploads
Papers by Philip Orlik