Generative AI, exemplified by large language models like the OpenAI GPT and Meta LLaMA families, can produce diverse content in response to prompts. This capability offers a promising solution to challenges in precision medicine, which seeks to tailor treatments to individual clinical profiles but often struggles with data collection, cost, and privacy concerns. By generating realistic, privacy-preserving patient data, generative AI has the potential to transform patient-centric healthcare. With such motivation, this research develops a comprehensive Generative AI pipeline emphasizing data granularity for accurate prediction of personalized treatments. The pipeline features a central Large Language Model interacting with a Machine Learning agent to determine key factors affecting a patient’s condition. This information is compiled into a query to retrieve personalized suggestions from a guideline database. We exam- in model development, experimental processes, and concerns such as data quality, response evaluation, trust, and reliability. Then, we apply our developed framework on three chronic diseases, namely diabetes, heart disease, and mental illness, on the capability of generating tailored treatment recommendations. Experiments show the proposed framework is a promising step toward explainable, personalized, and clinically aligned AI-driven treatment planning, laying the foundation for future trustworthy, patient-focused medical AI systems.

In this work, we present a novel approach to enhance the robustness of autonomous robotic Radio Frequency Identification (RFID) inventory systems using Conformal Prediction (CP). Recent AI-driven approaches, especially deep-learning models, have made significant advances in performing inventory strategies and action planning. However, these models lack the capability to measure uncertainty during the prediction process, which can result in accumulated errors and lead to catastrophic failures. To address the above challenge, we propose a confidenceguaranteed policy using CP to ensure reliable predictions in RFID inventory tasks. Our method focuses on managing the uncertainty in sub-goal estimation for a trained model, ensuring that predictions can meet or exceed a user-specific confidence level. We conduct extensive experiments to assess the proposed method by regulating an existing model and evaluate its effectiveness in identifying uncertain predictions. The experimental results demonstrate the effectiveness of our approach in improving both the reliability and efficiency of RFID inventory tasks, ensuring consistent and trustworthy operation.

Large Language Models (LLMs) have revolutionized natural language processing but remain resource-heavy and impractical for many organizations to deploy locally. Smaller specialist models offer a viable alternative, often developed using Knowledge Distillation (KD). Traditional KD methods, however, rely on static source datasets to elicit knowledge from the teacher model, limiting their ability to dynamically address student model weaknesses during training. This research introduces two adaptive knowledge elicitation methods: "Feedback-Driven Question Generation" and "Targeted Prompt Question Generation". Both methods iteratively expand the training dataset based on student model performance, leveraging a teacher model to target specific deficiencies. Using a Python QA task as a case study, our results show that both our methods enhance the student model's accuracy and response quality, with Method 2, which incorporates specialized prompt configurations, outperforming Method 1. These findings highlight the promise of adaptive KD in bridging the gap between large generalist AI models and smaller domain-specific models. Furthermore, these methods demonstrate an effective data augmentation technique for generating synthetic data specifically tailored to address student model weaknesses without overfitting or resulting in catastrophic forgetting.

The rise of Edge AI necessitates energy-efficient models like Spiking Neural Networks (SNNs), often trained using Federated Learning (FL) to preserve data privacy. However, FL is vulnerable to Byzantine attacks, where malicious clients disrupt training. While SNNs offer potential energy benefits due to their event-driven nature, their unique training mechanisms, particularly the use of surrogate gradients to handle non-differentiable spike events, raise questions about their inherent robustness in adversarial FL settings. We evaluate the robustness of SNNs employing 5 surrogate gradients (distinct by function shape) against 7 diverse Byzantine attacks and assess recovery potential using 5 robust aggregation rules (AGRs). Our extensive experiments (1032 runs) reveal that SNNs are not universally more robust than ANNs; they show resilience to certain structured attacks (e.g., MinMax) but vulnerability to others (e.g., Label Flip). We find a moderate positive correlation between surrogate gradient choice and recovery effectiveness using AGRs, with Triangle and Rectangle surrogates often enabling better recovery, though this advantage is context-dependent. Our results underscore that robust AGRs (like DnC and RFA) are essential for mitigating attacks in SNN-based FL, regardless of the surrogate gradient used. We conclude that achieving reliable SNN deployment in adversarial FL requires a holistic, context-aware approach, carefully considering the interplay between network type, surrogate gradient, threat model, and defense mechanisms. Our code is open-sourced for reproducibility1.

Spiking Neural Networks (SNN), which offer exceptional energy efficiency for inference, and Federated Learning (FL), which offers privacy-preserving training, is a rising area of interest that highly beneficial towards Internet of Things (IoT) devices. Despite this, research that tackles Byzantine attacks and bandwidth limitation in FL-SNN, both poses significant threats on model convergence and training times, still remains largely unexplored. In this paper, we first systematically evaluate the robustness of ANN and SNN in the FL context under four model-poisoning Byzantine attacks. We find that FL-SNN demonstrate better reliability than FL-ANN against most Byzantine attacks except MinMax. We then propose the$Top-\kappa$sparsification approach for better robustness in FL-SNN and to reduce communication overhead. Using this simple compression method, we observe ~40% accuracy enhancement in FL-SNN training under the lethal MinMax attack, leading to FL-SNN being more robust than FL-ANN in all four model-poisoning Byzantine attacks. This study highlights the dual benefits of FL-SNN with$Top-\kappa$sparsification in significantly reduce energy consumption and provide better robustness to Byzantine attacks for edge AI applications.

When multimodal AI systems increasingly utilize diverse data sources to achieve advanced understanding and interaction, they inevitably collect vast amounts of sensitive information, thus highlighting the urgent need for robust privacy safeguards, especially as these technologies expand into fields like healthcare, finance, and education. Existing research on data privacy in AI, encompassing adversarial training-based models, differential privacy-based models, and differentially private transform-based models, often neglects the inter-correlation inherent in multi-sensor data. To address this gap, we propose the differentiAl Private OnLine muLti-sensor data predictiOn model (APOLLO), which simultaneously considers intra-correlation and inter-correlation to enhance privacy protection while maintaining predictive performance. Under the proposed APOLLO frame-work, we design two implementations: APOLLO I, which ensures$\epsilon$-differential privacy by adding Laplace noise to each correlated data segment, and APOLLO II, which applies additional noise to make the concatenated multi-sensor data realize$\epsilon{-}$differential privacy. Furthermore, we conduct the theoretical analysis to reveal the relationship between performance influence and the privacy budget, providing guidelines for noise addition with the aim of achieving certified performance. Comprehensive experiments validate the effectiveness of the APOLLO model, establishing a new standard for privacy-preserving multi-sensor data prediction.

We develop a new framework, Chrono Clustering, to uncover and portray the progression of topics over time, where dual clustering, enhanced keyword extraction and hypergraph visualization are integrated. With K-Means on embedding space for base topic discovery and one-iteration K-Medoids for aligning base topics to other timeframes, our novel method could effectively quantify temporal topic shifts. Evaluated on AI conference datasets, Chrono Clustering boosts detection of trending topics that match real-world advances. It generates novel Chrono Graph to intuitively show dynamic topic progressions, demonstrating promising values for temporal topic modeling

Spiking Neural Networks (SNNs) have recently gained significant interest in on-chip learning in embedded devices and emerged as an energy-efficient alternative to conventional Artificial Neural Networks (ANNs). However, to extend SNNs to a Federated Learning (FL) setting involving collaborative model training, the communication between the local devices and the remote server remains the bottleneck, which is often restricted and costly. In this paper, we first explore the inherent robustness of SNNs under noisy communication in FL. Building upon this foundation, we propose a novel Federated Learning with Top-κ Sparsification (FLTS) algorithm to reduce the bandwidth usage for FL training. We discover that the proposed scheme with SNNs allows more bandwidth savings compared to ANNs without impacting the model’s accuracy. Additionally, the number of parameters to be communicated can be reduced to as low as 6% of the size of the original model. We further improve the communication efficiency by enabling dynamic parameter compression during model training. Extensive experiment results demonstrate that our proposed algorithms significantly outperform the baselines in terms of communication cost and model accuracy and are promising for practical network-efficient FL with SNNs.

Artificial Intelligence of Things (AIoT) represents a collaborative fusion of artificial intelligence (AI) and the Internet of Things (IoT). AIoT systems enable real-time data acquisition through IoT sensors and conduct intelligent data analysis tasks across the entire spectrum from terminal to edge to cloud, creating a dynamic and empowering ecosystem. However, the evolving landscape of AIoT faces a perplexing challenge: how to effectively sense the geographically diverse and highly variable environment, accurately gather massive and diverse IoT data with varying value density, and intelligently integrate multisource data to deliver real-time, intelligent, and high-quality applications.

The fast development in Deep Learning (DL) has made it a promising technique for various autonomous robotic systems. Recently, researchers have explored deploying DL models, such as Reinforcement Learning and Imitation Learning, to enable robots for Radio-frequency Identification (RFID) based inventory tasks. However, the existing methods are either focused on a single field or need tremendous data and time to train. To address these problems, this paper presents a Cross-Modal Reasoning Model (CMRM), which is designed to extract high-dimension information from multiple sensors and learn to reason from spatial and historical features for latent cross-modal relations. Furthermore, CMRM aligns the learned tasking policy to high-level features to offer zero-shot generalization to unseen environments. We conduct extensive experiments in several virtual environments as well as in indoor settings with robots for RFID inventory. The experimental results demonstrate that the proposed CMRM can significantly improve learning efficiency by around 20 times. It also demonstrates a robust zero-shot generalization for deploying a learned policy in unseen environments to perform RFID inventory tasks successfully.

Millimeter wave (mmWave) based human activity recognition is important in smart health in terms of studying user lifestyle. In practical health IoT scenarios, fast and accurate human activity recognition is critically important. In this work, we design and implement a lightweight deep learning solution for human activity recognition based on discrete Fourier transformation. The model has a fairly small number of model parameters while offering high accuracy in activity recognition. The core of the solution is a discrete Fourier transform module inside a neural network, which converts the temporal features of mmWave radar activity data into frequency features before activity recognition is performed by a simple classifier. We have extensively evaluated this solution against other traditional deep learning models in mmWave human activity recognition. The evaluation demonstrates that the DFT-based network can achieve the same accuracy as other traditional neural network models, but with a very small computational load.

Human daily activity monitoring has its particular significance in smart health. Human activity recognition based on mmWave has drawn enormous research efforts and achieved significant progress. Most of these solutions, however, work on data that has been manually segmented for each piece to contain only a single activity, which is impractical in reality where the sensor continuously generates data containing a series of activities. To address this challenge, this paper proposes a multi-head attention model that can detect the transition from one activity to another in a stream of mmWave sensor data of various human activities by analyzing the inner correlation of mmWave radar data fragments with a sliding window mechanism. Furthermore, the model then recognizes the new activity type in the data once it detects an activity transition. The solution has been extensively evaluated with a sparse point cloud dataset generated by a mmWave radar, which contains five types of activities. The experiment results show that the solution can achieve an accuracy of 98% in detecting activity transition at its best.

Reinforcement learning has shown encouraging results in decision-making tasks such as those found in gaming and robotics. However, when applied in safety-critical applications, like navigation, reinforcement learning can cause serious safety problems due to its trial-and-error nature. In this paper, we propose a safe reinforcement learning framework for LiDAR-based navigation. The proposed framework incorporates control barrier function theory with reinforcement learning to mitigate safety risks while learning a navigation strategy. Furthermore, by designing an online learning control barrier function, the proposed method can adapt to different obstacle settings through learning from past experiences. As a map-less solution, our approach uses only LiDAR as the perception system and requires no additional localization information. We conduct simulated experiments in ROS Gazebo under different indoor environments. The results show that our method outperforms other approaches by achieving up to twice the reward and five times fewer collisions when compared to the runner-up.

Insecure and portable devices in the smart city’s Internet of Things (IoT) network are increasing at an incredible rate. Various distributed and centralized platforms against cyber attacks have been implemented in recent years, but these platforms are inefficient due to their constrained levels of storage, high energy consumption, the central point of failure, underutilized resources, high latency, etc. In addition, the current architecture confronts the problems of scalability, flexibility, complexity, monitoring, managing and collecting of IoT data, and defend against cyber threats. To address these issues, the authors present a distributed and decentralized blockchain-software-defined networking (SDN)-based energy-aware architecture for IoT in smart cities. Thus, SDN is continuously observing, controlling, and managing IoT devices activities and detects possible attacks in the network; blockchain provides adequate security and privacy against cyber attacks, and reduces the central point of failure issues; network function virtualization (NFV) is used to saving energy, load balancing, as well as increasing the lifetime of the entire network. Also, we introduce a cluster head selection (CHS) algorithm to reduce the energy consumption in the presented model. Finally, we analyze the performance using various parameters (e.g., throughput, response time, gas consumption, and communication overhead) and demonstrate the result that provides higher throughput, lower response time, and lower gas consumption than existing works for smart cities.

The Ever-Increasing evolution of technologies in communication, artificial intelligence (AI), manufacturing, etc., is promoting a new wave of industrial revolution. Industrial Internet of Things (IIoT) has been considered as a critical stimulator for both science and economics by amounts of countries.

Primary Hyperparathyroidism(PHPT) is a relatively common disease, affecting about one in every 1,000 adults. However, screening for PHPT can be difficult, meaning it often goes undiagnosed for long periods of time. While looking at specific blood test results independently can help indicate whether a patient has PHPT, often these blood result levels can all be within their respective normal ranges despite the patient having PHPT. Based on clinical data from the real world, in this work, we propose a novel approach to screening PHPT with neural network (NN) architectures, achieving over 97% accuracy with common blood values as inputs. Further, we propose a second model achieving over 99% accuracy with additional lab test values as inputs. Moreover, compared to traditional PHPT screening methods, our NN models can reduce the false negatives of traditional screening methods by 99%.

In order to solve the large-scale, strong coupling, and nonlinear optimization problems in many Internet-of-Things (IoT) applications, such as intelligent infrastructure and smart city, this article proposes a real-coded genetic algorithm based on an auxiliary-individual-directed crossover operator (AIDX). AIDX is an alternative offspring framework of the directed crossover which uses auxiliary individual to reduce the search space of alternative offspring for the rapid optimization of multidimensional problems. In our solution AIDX, the parents-center distribution is adopted to reduce the risk of convergence to local optimization and enhance the stability of the algorithm. In addition, in order to increase the diversity of the population at the late stage of optimization, K-Bit-Swap (KBS) is used as a supplement for the exchange of genetic information among individuals in different dimensions. In the extensive experiments, 24 benchmarks with different dimensions are used to evaluate the performance of AIDX-GA. The results show that the proposed AIDX-GA has a significantly improved optimization effect on multidimensional optimization problems, and the stability of the algorithm is largely enhanced even when the global optimum is located near the boundary. AIDX-GA has also been evaluated in an industrial IoT case that identifies the resistance coefficients of a pipe network in the city infrastructure. The results show that the accuracy of AIDX-GA is high, and it has excellent universality and stability, which can be used to solve many IoT problems.

Imitation learning holds the promise to address challenging robotic tasks such as autonomous navigation. It however requires a human supervisor to oversee the training process and send correct control commands to robots without feedback, which is always prone to error and expensive. To minimize human involvement and avoid manual labeling of data in the robotic autonomous navigation with imitation learning, this paper proposes a novel semi-supervised imitation learning solution based on a multi-sensory design. This solution includes a suboptimalsensor policybased on sensor fusion to automatically label states encountered by a robot to avoid human supervision during training. In addition, arecording policyis developed to throttle the adversarial affect of learning too much from the suboptimal sensor policy. As a result, this solution allows the robot to learn a navigation policy in a self-supervised manner without human intervention after the initial data collection. With extensive experiments in indoor environments, this solution can achieve near human performance in most of the tasks and even surpasses human performance in case of unexpected events such as hardware failures or human operation errors. To best of our knowledge, this is the first work that synthesizes sensor fusion and imitation learning to enable robotic autonomous navigation in the real world without human supervision.

Human activity recognition based on wireless sensing is advantageous at various features such as privacy preservation, but also very challenging due to the instability of wireless signals. This paper proposes a deep learning driven wireless human activity recognition solution based on Multiple-Input-Multiple-Output (MIMO) radar sensing. User activities are first sensed by a low-power Frequency-Modulated Continuous Wave (FMCW) MIMO radar. Then a sequence of 3D images are generated out of the reflected signal strength. Next, deep neural networks (DNNs) are designed to analyze the correlation among the sequential 3D images to recognize various types of human activities. This work has developed: 1) a large dataset containing over 1,500 training videos of six different types of indoor activities, 2) a customized deep learning video data-loader to select proper training data in each training epoch, 3) a deep recurrent neural network (RNN) model to recognize human activities based on radar imaging results. This solution has been extensively evaluated in a research lab room. The results show that the solution is able to generate wireless imaging frame-by-frame, and it can achieve over 86.7% accuracy in recognizing six different types of human activities.

In-band full duplex (IBFD) wireless is of utmost interest to future wireless communication and networking due to great potentials of spectrum efficiency. IBFD wireless, how- ever, is throttled by its key challenge, namely self-interference. Therefore, effective self- interference cancellation is the key to enable IBFD wireless. This paper proposes a real-time non- linear self-interference cancellation solution: Deep learning based Self-Interference Cancellation (DSIC) to enable IBFD wireless. In this solution, a self-interference channel is modeled by a deep neural network (DNN). Synchronized self- interference channel data is first collected to train the DNN of the self-interference channel. Afterwards, the trained DNN is used to cancel the self-interference at a wireless node. This solution has been implemented on a USRP SDR testbed and evaluated in real world in multiple scenarios with various modulations in transmitting information including numbers, texts as well as images. It results in the performance of 17dB in digital cancellation, which is very close to the self-interference power and nearly cancels the self- interference at a SDR node in the testbed. The solution yields an average of 8.5% bit error rate (BER) over many scenarios and different modulation schemes.

Compared to traditional camera-based computer vision and imaging, radio imaging based on wireless sensing does not require lighting and is friendly to privacy. This work proposes a deep learning radio imaging solution to visualize real-time user indoor activities. The proposed solution uses a low-power, MIMO Frequency Modulated Continuous Wave (FMCW) radar array to capture the reflected signals from human objects, and then constructs 3D human visualization through a serials of data analytics including: 1) a data preprocessing mechanism to remove background static reflection, 2) a signal processing mechanism to transfer received complex radar signals to a matrix containing spatial information, and 3) a deep learning scheme to filter abnormal frames resulted from rough surface of human body. This solution has been extensively evaluated in an indoor research lab. The constructed real-time human images are compared to the camera images captured at the same time. The results show that the proposed radio imaging solution can result in significantly high accuracy.

The fifth generation (5G) mobile communication system is regard as a key enabler in promoting the deployment of Internet of Things (IoT), which is accompanied by the increasing service demands such like high data rate, enormous connection, and low latency. To meet these demands, network slicing has been envisioned as an efficient technology to customize infrastructures and allocate resources for 5G IoT services. However, due to various application backgrounds and ubiquitous social interactions of IoT services, the heterogeneous and social-driven resource requirement of users should be carefully assessed in resource allocation for the sliced 5G wireless network. In this paper, a novel nature-inspired wireless resource allocation scheme with slice characteristic perception is proposed, which comprehensively analyzes the properties of slices and converts them into a network profit model of resource utilization. Specifically, personalized service preferences and evolutionary interest relationships of users are exploited to model the complex and dynamic network environment with cellular automaton, and a biologically inspired allocation strategy of virtual wireless resource is proposed on the requirements of continuously updated user groups. Simulation results show that the proposed scheme achieves favorable resource utilization and low computational complexity, which favors the dynamic IoT slicing architecture and improves the efficiency and flexibility of resource allocation.

Smart health calls for novel approaches to detect vital signs in non- contact, non-invasive and non-intrusive matters. In this work, we design a solution that monitors the rates of heartbeats and respiration simultaneously by using a Frequency Modulated Continuous Wave (FMCW) radar with multiple antennas. This solution measures the reflections from heartbeats and respiration at a high frequency of 4 K H z to capture fine dynamics of motions with big data. It employs multiple antennas and superposition to reduce the interference noises from unwanted motions in the background and any detection defects. The heart and respiration rates are detected in the frequency domains after a chain of preprocessing techniques on the sensed big data. With extensive experiments in a lab office, this system demonstrates high accuracies in various cases: 98% in the still case, 95% with finger motions and 96% with body motions. The tests also confirm that multiple antennas and signal superposition improve the detection accuracy and reliability.

Deep imitation learning enables robots to learn from expert demonstrations to perform tasks such as lane following or obstacle avoidance. However, in the traditional imitation learning framework, one model only learns one task, and thus it lacks of the capability to support a robot to perform various different navigation tasks with one model in indoor environments. This paper proposes a new framework, Shared Multi-headed Imitation Learning (SMIL), that allows a robot to perform multiple tasks with one model without switching among different models. We model each task as a sub-policy and design a multi-headed policy to learn the shared information among related tasks by summing up activations from all sub-policies. Compared to single or non-shared multi-headed policies, this framework is able to leverage correlated information among tasks to increase performance. We have implemented this framework using a robot based on NVIDIA TX2 and performed extensive experiments in indoor environments with different baseline solutions. The results demonstrate that SMIL has doubled the performance over non-shared multi-headed policy.

Indoor human activity recognition has been extensively investigated. However, most of the solutions require sensors e.g. 9-axis IMU be equipped on human body or use image processing that presents privacy issues. This work proposes an ambient radar sensor based a solution to recognize the activities that humans normally perform in indoor environments. This solution uses a 7.8 GHz radar to emit 16 pulse signals every second and samples the reflected signals at 128 KHz to capture the fine dynamics of human activities. This solution designs a set of data preprocessing algorithms, including a data refining algorithm to filter outlier data, a contrastive divergence algorithm to remove background static reflection, and a transformation algorithm to convert the signal data into feature- rich spatial location changes. This solution also develops schemes to separate a collection of various activities into individuals. A lowpass frequency filter is designed to remove unwanted noisy data and the motion intensity is used to classify the activities into two high-level groups. It uses a slope-based approach and a k- means clustering to further finely recognize each activity. This solution has been extensively evaluated in a spacious research lab room and shows outstanding accuracy.

In this paper, we introduce Time-Constrained Vehicle Routing Problem (TCVRP), a variation of the Vehicle Routing Problem (VRP) with time constraints. We formulate the problem as Integer Linear Programming with the two objectives: minimizing the number of agents used in routing, and minimizing the time spent in routing. We propose an efficient heuristic using a combination of A*Search and Ruin and Recreate algorithms to solve TCVRP Finally, We conduct an experimental evaluation of the proposed heuristic and provide the performance results.

Malicious network nodes often incur problems to network and data privacy by distributing forged public keys. To address this issue, this paper proposes a dynamic trust relationships aware data privacy protection (DTRPP) mechanism for mobile crowd-sensing. In this mechanism, combining key distribution with trust management, the trust value of a public key is evaluated according to both the number of supporter and the trust degree of the public key. The trust value is estimated from the accuracy of the public key provided by the encountering nodes. DTRPP achieves the dynamic management of nodes and estimates the trust degree of the public key. In addition, by classifying traffic data into different types and selecting a proper relay node to forward the data according to data types, it is more effective to use the network resource with the trust degree and centrality of the relays. With extensive evaluations, results show that the proposed mechanism protects the data privacy effectively and has better performance on the average delay, the delivery rate and the loading rate when compared to traditional mechanisms.

Wireless Body Area Networks (WBAN) have broad prospects for the use in mobile health, sports training support, etc. One critical research problem on WBAN is the cross-interference among multiple WBANs when they are close to each other, because they work on unlicensed open wireless frequency bands. This paper proposes a social group interaction power control game model to mitigate inter- WBAN interference, which consists of new utility and cost functions designed to accommodate both convergence speed and quality. This work proves that only one Nash equilibrium (NE) point exists for this game model, which guarantees its convergence. Extensive simulation has been performed to evaluate the performance and the results demonstrate that the proposed algorithm is highly effective and the convergence is rapid.

The world has entered into a “smart” era. One area becoming smart is the place where we live-homes. Smart homes are expected to be equipped with numerous sensors to continually monitor, sense, and actuate the space. The data from these sensors can be used to provide various types of services by automating common tasks while causing minimal disruption to daily life. In order to provide these services, a system must have sufficient intelligence to predict future events based on its observations. This paper first examines the requirements for smart home predictions. It then comprehensively reviews prediction algorithms and variations that have been proposed and investigated in smart environments, such as smart homes. It is these prediction algorithms that provide the intelligence required by a smart home. Comparisons are also made upon these prediction algorithms on their features and models.

Internet of Things (IoT) heavily relies on wireless communication to interconnect various sensors and hubs. This paper proposes a symbol-level information multiplexing mechanism (SIMPLEX) that exploits link margin in wireless networks to minimize channel underutilization. Multiplexing is achieved by carrying extra information through a type of specially designed symbols inserted. The key enabler of the inserting and detecting such specially symbols is a per-bit channel assessment scheme that hierarchically estimates the error probability of a received symbol. On the GNU SDR testbed experiments, SIMPLEX shows an accuracy of 97% in recognizing the special symbols in demultiplexing. By varying the frequency domain indices of such specially symbols and their positions on I-Q constellation map, SIMPLEX provides a series of multiplexing rates that carry different amount of extra information. We also design an adaptive multiplexing rate selection scheme to dynamically achieve optimal exploitation of link margin upon instant channel conditions, which can find the optimal multiplexing rates over 90% of the time on the testbed. Multiplexing throughput gain has been theoretically derived and empirically validated of high accuracy with only a negligible deviation to experiment results. A throughput gain can be obtained as much as up to 55%.

Cyber-physical systems (CPS) are emerging research areas that involve multiple disciplines. Two critical components in CPS are networking technologies and security. Because of the multi-disciplinary nature, the networking and security of CPS expand beyond traditional computing domains and have to consider the impact of applied physical systems. Therefore, new innovations are required such as novel transmission technologies, networking protocols, architectures, and security solutions. As a result, a significant amount of research work is expected for new models, performance analysis as well as evaluation, prototypes, and testbeds. This special issue focuses on research interests and activities related to networking and security in SmartGrid, Transportation and Medical Systems, with an emphasis on original analytical, experimental, and systems-related papers in these target domains.

Multi-channel media access control (MAC) is important in wireless sensor networks because it allows parallel data transmissions and resists external wireless interference. Existing multi-channel MAC protocols, however, do not efficiently support delay-sensitive applications that require reliable and timely data transmissions. In addition, multi-channel operation is inherently deficient for supporting multi-hop broadcasting, due to independent waking-up schedules on sensors. To address these issues, we present a routing-enhanced multi-channel MAC (RM-MAC) which allows nodes to coordinately select their channel polling times based on cross-layer routing information. RMMAC also supports a ripple broadcast mechanism which achieves efficient multi-hop broadcast among sensors. Simulation results show that RM-MAC provides significant improvement over the MuCHMAC [1], in terms of end-to-end delay, under a wide range of traffic loads including both unicast and broadcast traffic.

This paper proposes a wireless channel assessment metric A posteriori Bit Error Probability (ABEP) that indicates the error likelihood of each bit during demodulation. Estimated upon constellations, ABEP provides the finest granularity (per bit) in revealing dynamic channel conditions. With comparative validation on an implemented GNU SDR testbed, ABEP significantly outperforms a pioneer flagship work SoftPHY in fidelity and channel predictability. An ABEP based channel assessment algorithm is devised to recognize channel conditions such as channel fading and collision, and it demonstrates a recognition accuracy up to 99% across various channel conditions. Based on the ABEP channel assessment, a cross-layer protocol-ABEP-enabled adaptive rate control (AARC)-is designed for high wireless network performance. AARC is implemented and evaluated on ns-3 simulator with a newly developed physical layer. Extensive experiments show that AARC significantly outperforms prior rate control solutions in rate selection accuracy and throughput.

Hadoop is the de facto engine that drives current cloud computing practice. Current Hadoop architecture suffers from single point of failure problems: its job management lacks of fault tolerance. If a job management fails, even if its tasks remains still active on cloud nodes, this job loses all state information and has to restart from scratch. In this work, we propose a distributed MapReduce engine for Hadoop with the Distributed Hash Table (DHT) algorithm that drives the scalable peer-to-peer networks today. The distributed Hadoop engine provides the fault-tolerance capability necessary to support efficient job computation required in the cloud computing with numerous jobs running at a moment. We have implemented the proposed distributed solution into Hadoop and evaluated its performance in job failures under various network deployments.

Efficient and accurate target localization is one of the most fundamental problems in Wireless Sensor Networks (WSN), and has been studied for several years. Due to its simplicity and low energy consumption, binary sensing model is widely used in the literature to achieve fast target localization but with low accuracy. To improve localization accuracy, we propose a novel multi-bit sensing model where multi-bit information is sent by sensors to report the relative distances between a target and the sensors. Based on this sensing model, a new target localization algorithm is proposed, which can improve localization precision by estimating a targets position within a reduced area. Furthermore, the proposed algorithm works well with irregular sensing boundary caused by noise, channel fading, and obstacles. Simulation results show that the multi-bit sensing based target localization (MSTL) algorithm could improve localization accuracy by 50%.

Wireless information security generates shared secret keys from reciprocal channel dynamics. Current solutions are mostly based on temporal per-frame channel measurements of signal strength and suffer from low key generate rate (KGR), large budget in channel probing, and poor secrecy if a channel does not temporally vary significantly. This paper designs a cross-layer solution that measures noise-free per-symbol channel dynamics across both time and frequency domain and derives keys from the highly fine-grained per-symbol reciprocal channel measurements. This solution consists of merits that: (1) the persymbol granularity improves the volume of available uncorrelated channel measurements by orders of magnitude over per-frame granularity in conventional solutions and so does KGR; 2) the solution exploits subtle channel fluctuations in frequency domain that does not force users to move to incur enough temporal variations as conventional solutions require; and (3) it measures noise-free channel response that suppresses key bit disagreement between trusted users. As a result, in every aspect, the proposed solution improves the security performance by orders of magnitude over conventional solutions. The performance has been evaluated on both a GNU SDR testbed in practice and a local GNU Radio simulator. The cross-layer solution can generate a KGR of 24.07 bits per probing frame on testbed or 19 bits in simulation, although conventional optimal solutions only has a KGR of at most one or two bit per probing frame. It also has a low key bit disagreement ratio while maintaining a high entropy rate. The derived keys show strong independence with correlation coefficients mostly less than 0.05. Furthermore, it is empirically shown that any slight physical change, e.g. a small rotation of antenna, results in fundamentally different cross-layer frequency measurements, which implies the strong secrecy and high efficiency of the proposed solution.

This paper proposes a bit rate adaptation scheme based on frame transmission statistics. The bit adaptation scheme judiciously exploits the fragmentation mechanism in the IEEE 802.11 to accurately diagnose the cause of a frame transmission failure and accordingly adjusts the frame delivery statistics. We analytically prove that the fragmentation mechanism incurs less overhead and is more effective than the use of RTS/CTS for rate adaptation in a collisions dominated environments. We implemented and evaluated this scheme and four other recent and representative rate adaptations schemes on a Linux based testbed. We also discuss some anomalies plaguing these rate adaptation schemes observed while implementing them. With extensive experiments, we observe that these rate adaptation schemes have strengths and weaknesses under different scenarios, although our scheme performs very well in most of cases.

Wireless communication is indispensable to Internet of Things (IoT). Carrier sensing multiple access/collision avoidance (CSMA/CA) is a well-proven wireless random access protocol and allows each node of equal probability in accessing wireless channel, which incurs equal throughput in long term regardless of the channel conditions. To exploit node diversity that refers to the difference of channel condition among nodes, this paper proposes two opportunistic random access mechanisms: overlapped contention and segmented contention, to favor the node of the best channel condition. In the overlapped contention, the contention windows of all nodes share the same ground of zero, but have different upper bounds upon channel condition. In the segmented contention, the contention window upper bound of a better channel condition is smaller than the lower bound of a worse channel condition; namely, their contention windows are segmented without any overlapping. These algorithms are also polished to provide temporal fairness and avoid starving the nodes of poor channel conditions. The proposed mechanisms are analyzed, implemented, and evaluated on a Linux-based testbed and in the NS3 simulator. Extensive comparative experiments show that both opportunistic solutions can significantly improve the network performance in throughput, delay, and jitter over the current CSMA/CA protocol. In particular, the overlapped contention scheme can offer 73.3% and 37.5% throughput improvements in the infrastructure-based and ad hoc networks, respectively.

Cooperative communication can achieve diversity gain and increase the channel capacity. This paper proposes a novel cooperative scheme called Cooperative Binary Relaying and Combining (CBRC) for multi-hop wireless networking systems where the nodes cooperatively demodulate high-order modulated signal symbols with low-order robust modulations. Low-order demodulation schemes make partial decision at each relay node and avoid propagating the errors that can result from high-order demodulation schemes in conventional cooperative relaying. Therefore, CBRC supports high bit rate transmission at high-order modulations over low SNR links. Extensive simulations are conducted to evaluate the bit error rate performance of CBRC and conventional cooperative strategies. The results show that CBRC can significantly improve the performance.

There is a strong need to explore green and harvestable energy in computer communications. However, adapting wireless network performance to harvested energy has largely been ignored in literature. In this paper, we propose a new resource allocation scheme to improve data delivery quality in energy harvesting enabled wireless networks. In the proposed approach, packet Automatic Repeat reQuest (ARQ) limit of each wireless node is adaptively adjusted according to harvested energy. To achieve such optimal retry adaptation, energy neutrality constraint is considered in the overall optimization process. Simulation results show that the proposed retry adaptation approach significantly improves packet delivery ratio by exploring the harvested energy.

Understanding channel dynamics is essential and critical to a variety of research orientations on wireless networks. This paper presents observations and analysis from extensive measurements on IEEE 802.11g channels in an indoor environment with a customized testbed and measurement tools. We obtained the following major observations: 1) delivery ratio varies smoothly and large time scale delivery ratios change largely; 2) however, Signal-to-Noise Ratio (SNR) varies largely in micro time scale, but stable in large time scale; 3) conformable to what is observed by other researchers, frame delivery ratio is not strongly correlated with SNR; 4) large time scale loss rate information is not informative.

Rate adaptation is the determination of the optimal data transmission rate most appropriate for current wireless channel conditions. It consists of assessing channel conditions and accordingly adjusting the rate. Rate adaptation is fairly challenging due to wild channel conditions fluctuations. In the last decade, rate adaptation for IEEE 802.11 networks has been extensively investigated. This paper presents a comprehensive and detailed study of the advances of rate adaptation schemes proposed for IEEE 802.11 networks, and summarizes their characteristics. We also categorize these rate adaptation schemes based on their support of loss differentiation and their methods to sense the channel conditions. Then, this paper compares the performance of three representative schemes through simulations. Finally, open issues for rate adaptation are raised.

With opportunistic transmissions, a node exploits a stable high data rate to transmit multiple frames (instead of one) whenever it captures a contended medium. The IEEE 802.11 networks could support such multi-frame transmissions at high data rates if the channel remains stable long enough. Opportunistic transmission has been studied and shown to improve IEEE 802.11 networks performance. However, to our knowledge all opportunistic schemes require RTS/CTS control frames while RTS/CTS control frames are usually not used. First, this work introduces a Basic Opportunistic Transmission (BOT) scheme that works without RTS/CTS control frames. However, the lack of RTS/CTS control frames may lead to high congestion loss rates that defeat any opportunistic scheme and drastically lower network performance. Channel fading exacerbates the frame loss. This work addresses this weakness by proposing an effective loss recovery strategy to enable BOT to cope with frame losses from congestion and channel fading. BOT with loss recovery constitutes the OTLR. Extensive ns-2 simulations with CBR and TCP traffic illustrate that 1) BOT yields a dramatic improvement over traditional single frame transmission, 2) BOT is vulnerable to high frame loss rates stemming from collisions or channel fading, and 3) OTLR improves BOT's performance in lossy environments.

Data rate adaptation aims to select the optimal data rates for current channel conditions leading to substantial performance improvement. This paper proposes a data rate adaptation technique that: (1) exploits the periodic IEEE 802.11 beacons; (2) discriminates between frame losses due to channel fading and those due to collisions and takes actions appropriate for each type of loss; and (3) recommends and justifies the use of the lowest data rate for the very first retransmission after a frame loss. The last feature - namely retransmitting at the lowest data rate - helps in diagnosing the real cause of a frame loss. Moreover, this work analytically shows that retransmitting at the lowest data rate is more efficient, especially in poor SNR environments or when there is no knowledge of the cause of a loss (channel degradation or transmission collision). This scheme, dubbed loss differentiated rate adaptation (LDRA), is extensively evaluated through simulations and shown to perform better especially when network traffic is heavy.

Location determination of indoor mobile users is challenging due the complex and volatile indoor radio propagation signals. A radio-frequency (RF) based indoor localization system, like RADAR or ARIADNE, typically operates by first constructing a lookup table mapping the radio signal strength at different known locations in the building, and then a mobile user's location at an arbitrary point in the building is determined by measuring the signal strength at the location in question and searching the corresponding location from the above lookup table. Usually, the mobile's signal strength is measured by three or more sniffers deployed inside the building. Obviously, the number of sniffers and their positions greatly affect the localization performance. This paper presents a detailed analysis and experimental results that explore the impact of the sniffers deployment on the performance of the indoor localization. The results demonstrate that the best localization performance is obtained when the center of gravity of the equilateral triangular (formed by three sniffers) coincides with that of the floor plan; and in order to provide optimal localization for all positions of a large floor, it is necessary to deploy more than three sniffers in a semi-mesh style such that any position in the building is always covered by three nearby sniffers.