Channel state information (CSI) as essential knowledge in communications has recently been exploited by WiFi and Internet-of-Things (IoT) devices to achieve sensing such as human activity and event detection. This passive sensing paradigm in WiFi/IoT systems poses a new privacy threat since CSI is easily accessible by user/IoT devices (UDs), but it is hard to know whether a UD analyzes CSI to extract environmental information and to prevent it from doing so. This paper presents a theoretical study of how a WiFi/IoT access point (AP) can privately precode signals, coined anti-sensing precoding, to impact a UD's CSIbased sensing upon AP-UD communication. To draw insights into environmental information privacy, we model and derive UD's sensing performance metrics in closed form. Two possible cases in which a UD has prior or online CSI knowledge as a basis for sensing decisions are studied. The optimal anti-sensing precoding that minimizes UD's correct decision probability in each case is derived and analyzed. Further, the influences of the precoding on UD's communication performance are discussed. In simulations, we validate our theoretical results and reveal the benefit and cost of the precoding to anti CSI-based passive sensing.

Passive sensing has garnered substantial attention in e-healthcare and smart home applications for the appealing concept of detecting human occupancy and activities based on channel state information (CSI) in wireless signal transmissions. Despite the promising benefits of passive sensing, what is less mentioned is the privacy issue raised. It is difficult to prevent a communication device from analyzing CSI to learn private environmental information. This paper investigates the use of a reconfigurable intelligent surface (RIS) by a system planner (SP) to assist a user device (UD) Communication and meanwhile Anti its Passive Sensing based on CSI, named RIS-CAPS. We address the fundamental UD passive sensing performance in identifying changes in the environment, called environmental states, by CSI estimation and reveal the tradeoff between communication and system privacy when a RIS is controlled to influence wireless channels. We report a shortboard effect that to anti UD’s sensing, the UD’s communication performance in different environmental states will be determined by the state having the worst channel condition. Furthermore, we analyze and tackle the new RIS configuration optimization problem to improve UD’s communication while constraining its passive sensing performance. A solution algorithm that exploits alternating optimization and semidefinite relaxation techniques is developed to address the challenging RIS configuration optimization. Finally, extensive simulations validate the performance modeling and analysis in our study. Insights into when RIS is more capable to anti passive sensing are provided.

Passive sensing that extracts human occupancy and activity information from wireless signal propagation properties has drawn growing interest in recent years. With wireless signals pervasively accessible, passive sensing brings both opportunities and threats to privacy. To leverage the sensing benefits, improving passive sensing accuracy and granularity is pursued. However, in turn, this increases the difficulty of protecting information. In this study, we investigate the use of reconfigurable intelligent surface (RIS) to influence passive sensing for different needs, i.e., to support or to prevent sensing. This paper presents the first theoretical performance analysis of RIS-enabled passive sensing and thereof, addresses the optimal RIS configurations that can improve or degrade the sensing accuracy as desired. In specific, with knowledge of channel features in target sensing scenes of interest, we model the passive sensing performance by a signal receiver to identify the target scenes when a RIS is utilized in the environment. We derive closed-form probability metrics of correct sensing, i.e., the accuracy, and false sensing of the scenes, with which we optimize RIS configuration to maximize/minimize the sensing accuracy. A heuristic algorithm, termed RIS- PS, is developed, solving the RIS configurations in a low complexity. In simulations, we verify our theoretical results and the performance of RIS-PS. Results show RIS's ability in both light-of-sight and non-light-of-sight passive sensing scenarios.

Machine learning (ML)-enhanced future communication networks are able to advance human-to-machine (H2M) applications by intelligent bandwidth prediction techniques to achieve bandwidth pre-allocation. Existing methods of H2M bandwidth prediction typically assume the stationary data stream over time. However, in the near future, communication networks are expected to support dynamic and heterogeneous applications. Since different H2M applications will exhibit different traffic distributions and loads across the day, an ML model learned on a specific H2M application at a particular network load will, therefore, be unable to adapt to changing applications and network loads. This will give rise to the phenomenon known as concept drift. This paper addresses concept drift in dynamic and heterogeneous networks supporting H2M applications by proposing a novel framework, the concept drift detection and adaptation (CDDA) framework, to respond and adapt to the concept drift rapidly. CDDA learns the traffic characteristics of H2M applications and combines offline and online learning processes to enhance H2M traffic prediction and improve band-width prediction performance. Results from our investigation using experimental traffic from H2M applications over a 10Gb/s passive optical network simulator show that CDDA can more rapidly respond to concept drift and better predict the bandwidth of changing H2M applications and network load.

Reinforcement learning (RL) has recently emerged as a promising solution for intelligence bandwidth decisions that reduce latency in optical access networks. Even though RL drives model-free self-adaptive bandwidth decisions, the learning time cost and the widely-known exploration-exploitation dilemma of when to apply the best decision learnt are challenging to address in the bandwidth decision context. This paper for the first time exploreshow to rapidly learn an optimal bandwidth decision with a known confidence level of the decisionfor minimizing optical access network latency. We investigate critical aspects, including reward acquisition and strategies to explore decisions, in an RL-based bandwidth allocation scheme. Applying renewal theory, we address the timing for the central office (CO) to acquire rewards from optical network units for accurate decision value evaluation. Further, we derive the relationship between the decision practice times and the confidence of the optimal decision in closed-form. A reward variance-oriented (RVO) exploration strategy is proposed, in which the CO selects bandwidth decisions with probabilities proportional to the reward variances. We prove that the RVO is the most time efficient in learning an optimal decision with a confidence guarantee. With numerical and extensive simulations, we validate the theory and compare several common strategies with the RVO.

Passive optical networks are envisioned to become increasingly complex as they support more and more diverse and immersive services that have different capacity, latency, and reliability needs. In the near term, they are expected to support the delivery of a diverse and immersive set of services including mixed reality, holographic communication, human-to-machine/robot communications, Tactile Internet, and digital sensing. However, in supporting these diverse and immersive services, traffic on the network will become increasingly dynamic across a range of different time scales. The upstream bandwidth in a passive optical network is typically shared by a group of end users, meaning that the uplink latency performance as experienced by each end user is thus highly dependent on the amount and when bandwidth to that end user is allocated. Machine learning enhanced bandwidth allocation algorithms have been proposed but are typically stationary, primarily-designed or pre-trained based on certain network configurations. In dynamic network conditions where traffic can evolve over time, concept drift, a phenomenon whereby the underlying distribution of the training data will no longer be representative of that in deployment, may occur. In view of future dynamic network conditions, we present a novel online reinforcement learning based bandwidth allocation scheme to address concept drift in machine learning enhanced passive optical network. The scheme facilitates self-adaptive decisions in real-time to accommodate dynamic network environments with varying traffic types and network loads. Results from comprehensive performance evaluation of the scheme show that rapid and self-adaptive bandwidth decisions can be achieved, yielding ~ 60% latency improvement in dynamic traffic environments.

Due to its mobility and agility, unmanned aerial vehicle (UAV) has emerged as a promising technology for various tasks, such as sensing, inspection and delivery. However, a typical UAV has limited energy storage and cannot fly a long distance without being recharged. This motivates several existing proposals to use trucks and other ground vehicles to offer riding to help UAVs save energy and expand the operation radius. We present the first theoretical study regarding how UAVs should optimally hitch on ground vehicles, considering vehicles' different travelling patterns and supporting capabilities. For a single UAV, we derive closed-form optimal vehicle selection and hitching strategy. When vehicles only support hitching, a UAV would prefer the vehicle that can carry it closest to its final destination. When vehicles can offer hitching plus charging, the UAV may hitch on a vehicle that carries it farther away from its destination and hitch a longer distance. The UAV may also prefer to hitch on a slower vehicle for the benefit of battery recharging. For multiple UAVs in need of hitching, we develop the max-saving algorithm (MSA) to optimally match UAV-vehicle collaboration. We prove that the MSA globally optimizes the total hitching benefits for the UAVs.

Human-controlled and haptic feedback data in emerging Tactile Internet human-to-machine (H2M) applications require stringent low-latency transmission. Understanding the traffic features of the new applications is vital in innovating network control and resource allocation strategies to meet their latency demand. In this article, we present our experimental study on human control and haptic feedback traffic in H2M applications and investigate novel bandwidth allocation schemes in supporting converged H2M application delivery over access networks. We introduce our haptic experiment system, the developed H2M applications, and analyze the control and feedback traffic traces collected. Then, exploiting the correlation between real-time control and feedback reported in our analysis, we propose an artificial intelligence-facilitated low-latency bandwidth allocation (ALL) scheme for emerging H2M applications. ALL provisions priority-differentiated bandwidth allocation for aggregated H2M and conventional content-centric applications over future access networks. By using ALL, the central office preallocates bandwidth for control and its corresponding feedback traffic interactively and prioritizes their transmission over content traffic. This expedites H2M application delivery by eliminating the report-then-grant process in the existing bandwidth allocation schemes. Via extensive simulations injected with experimental traffic traces, we comprehensively investigate the latency performance of ALL and existing schemes. Our results validate the superior capability of ALL in reducing and constraining latency for H2M applications.

This paper presents the exploitation of an artificial neural network (ANN) to facilitate insights into existing bandwidth allocation schemes in optical access networks and supervise bandwidth allocation decisions that reduce the latency. Specifically, based on the classic and predictive dynamic bandwidth allocation (DBA) schemes, we train a multi-layered ANN at the central office (CO) to learn the uplink latency corresponding to varying bandwidth allocation decisions. Multiple network feature knowledge, such as network load, traffic/packet statistics, fiber link distances and the number of optical network units (ONUs), is for the first time considered and utilized in the training process. Then, with the dependency between bandwidth allocation and the resultant latency learned by the ANN, we numerically analyze the latency performance of existing DBA schemes and show the optimal bandwidth decisions supervised by the ANN in achieving low latency. With extensive simulations, we show that exploiting the ANN to supervise bandwidth allocation at the CO, termed as ANN-DBA scheme, effective improvement in latency performance is realized.

In anticipation of future ubiquitously-connected healthcare services with stringent delay requirements, we outline the main characteristics, design challenges and existing open issues in the medium access control (MAC) layer design of wireless body area networks (WBANs) and highlight the need to define hybrid MAC frame that enables flexible access according to traffic and services. For the first time, a thorough investigation of two WBAN standards, namely the IEEE 802.15.6 WBAN and the recently-proposed ETSI SmartBAN, in terms of delay and energy, is presented. We provide insights into the impact of MAC frame timing structure on delay and energy performances of these existing protocols. We compare the selections of access durations for the SmartBAN hybrid MAC frame and IEEE 802.15.6 superframe. Then, we present our simulation comparisons of the uplink delay and energy consumption in a SmartBAN and a IEEE 802.15.16 WBAN for healthcare, taking into account periodic monitoring and health-critical emergency traffic patterns. Our results emphasize that compared to IEEE 802.15.6-defined WBANs, SmartBANs are advantageous in energy-saving. Moreover, with a time-optimized MAC, SmartBANs reduce the delay for both periodic monitoring and emergency report.

Human-to-machine (H2M) communications in emerging tactile-haptic applications are characterized by stringent low-latency transmission. To achieve low-latency transmissions over existing optical and wireless access networks, this paper proposes a machine learning-based predictive dynamic bandwidth allocation (DBA) algorithm, termed MLP-DBA, to address the uplink bandwidth contention and latency bottleneck of such networks. The proposed algorithm utilizes an artificial neural network (ANN) at the central office (CO) to predict H2M packet bursts arriving at each optical network unit wireless access point (ONU-AP), thereby enabling the uplink bandwidth demand of each ONU-AP to be estimated. As such, arriving packet bursts at the ONU-APs can be allocated bandwidth for transmission by the CO without having to wait to transmit in the following transmission cycles. Extensive simulations show that the ANN-based prediction of H2M packet bursts achieves >90% accuracy, significantly improving bandwidth demand estimation over existing prediction algorithms. MLP-DBA also makes adaptive bandwidth allocation decisions by classifying each ONU-AP according to its estimated bandwidth, with results showing reduced uplink latency and packet drop ratio as compared to conventional predictive DBA algorithms.

The smart body area network (SmartBAN) is a recently proposed system for realizing low complexity and ultralow power body area network. In this paper, we present the first study on SmartBAN downlink delay and energy performances considering the emerging control/actuation applications in future e-health. A novel downlink transmission framework arising from our performance study is presented, which reduces the delay and energy consumption for downlink-dominated SmartBANs. In this paper, we first investigate the delay of the supplementary downlink mode (SDM) specified in the SmartBAN medium access control (MAC) protocol and propose an improved SDM (ISDM), showing that reordering the access periods in the SmartBAN MAC frame can effectively reduce the delay. To address the delay bottleneck in SDM and ISDM, we further propose limited-exhaustive (LEDM) and fully exhaustive (FEDM) downlink mechanisms. Then, based on SDM, ISDM, LEDM, and FEDM, energy-saving mechanisms in the SmartBAN downlink are discussed. Moreover, to critically evaluate the delay and energy-savings of LEDM and FEDM, we develop two embedded Markov chains that suit the SmartBAN beacon-enabled MAC. Finally, based on the above performances study, a novel downlink transmission framework that selects suitable transmission mechanism and access durations for delay-constraint SmartBAN applications is proposed. Extensive simulations show the effectiveness of our proposed mechanisms and transmission framework.

The smart body area network (SmartBAN) is a recently proposed system for wireless body area networks (WBANs). Compared to conventional WBANs, it is designed to support lower system complexity and ultralow power consumption. In a SmartBAN, the sensors' access is scheduled upon receiving a beacon on the data channel at the beginning of each working cycle, termed as interbeacon interval (IBI). As network performance, including delay and energy consumption, is highly dependent on the length of IBI, we present, in this paper, an optimal IBI frame for SmartBAN. Our focus is on the delivery of uplink periodically-generated sensor data with low delay and high energy efficiency. As periodic traffic is a common traffic pattern widely generated in e-health applications, for which most previously proposed Markov models for WBANs are inapplicable, a closed-form analytical delay model for periodic SmartBAN transmission is derived. Exploiting this model, a time-optimized framework that minimizes average uplink delay is formulated. An adaptive IBI algorithm is then proposed to determine the optimal IBI during the network connection between hub and sensors. Finally, sleep mode in sensors and doze mode in the hub are introduced to reduce energy consumption under the proposed framework. With optimal IBI, the percentage of energy savings and channel efficiency is evaluated. Simulation and theoretical results show that by using the proposed time-optimized framework, both delay and energy consumption of periodic traffic are significantly reduced. Comparisons with the IEEE 802.15.6 WBAN show performance improvements in periodic uplink delay and energy consumption with our proposed time-optimized framework for SmartBANs.

Wireless body area networks (WBANs) are expected to support control/steering and haptic communication via on-body actuators in future Tactile Internet enabled e-health systems. However, little is known to date about the capability of WBANs in realizing remotely-controlled applications since current WBANs are uplink-transmission dominated for the purpose of health monitoring and supervision. In this paper, we present the first downlink delay performance evaluation based on the recently-proposed Smart Body Area Network (SmartBAN). To meet the stringent 1-ms Tactile Internet delay requirement for real-time tactile feedback and control delivery, we make comparisons between two downlink transmission mechanisms: (a) conventional exhaustive transmission; and (b) fixed-length exhaustive transmission based on the SmartBAN medium access control (MAC) layer protocol. M/D/1 and embedded Markov chain models are developed to evaluate downlink delay for SmartBANs adopting the transmission mechanisms above. The accuracy of our models is validated by simulations. Analytical and simulation results show that compared to the conventional mechanism, the fixed-length exhaustive transmission can effectively reduce downlink delay to less than 1 ms by increasing downlink transmission duration and improved energy performance can potentially be achieved thanks to the fixed MAC configuration. Further, with our model, suitable downlink durations can be determined by considering delay constraints in practical applications.

The Smart Body Area Network (SmartBAN) is a recently proposed system for wireless body area networks (WBANs). Unlike widely-used WBANs that employ contention-based medium access control (MAC) protocols, the SmartBAN MAC specifies a joint time division multiple access (TDMA) and slotted ALOHA access framework. To date, a criterion for determining the time duration of different channel access phases remains unaddressed in current WBAN MAC designs. There is also a lack of understanding of how the duration of these timing periods impacts the SmartBAN's performance. In this paper, we derive closed-form analytical models for the uplink transmission delay, which is defined as the duration a data packet generated by a sensor has to wait prior to its uplink transmission. We adopt a flexible channel access mechanism considering the characteristics of two major traffic patterns in medical applications: periodic monitoring traffic and Poisson-distributed emergency traffic. Then, based on both analytical models and simulation of the MAC timing parameters, and in conjunction with the behavior of a typical SmartBAN in terms of delay and energy consumption with aggregated traffic load, we present our solution to determine different access periods of the SmartBAN MAC. With extensive simulations, the accuracy of the delay model is validated and a blocking state of the SmartBAN is discussed, where the delay and energy performance is degraded significantly due to the queuing at each sensor. The results obtained in this paper provide a first in understanding how timing parameters and the traffic features impact SmartBANs' delay and energy performance.