The openness of underwater wireless sensor networks (UWSNs) exposes them to potential eavesdropping attacks, enabling attackers to trace back and identify the source nodes of packet flows. This poses a significant threat to the confidentiality of sensitive applications, known as the Source Location Privacy (SLP) problem. Conventional packet encryption methods are ineffective in defending against SLP attacks since attackers do not need to know the content of the packets. A commonly used method to address the SLP problem is to establish fake transmission paths, making attackers follow fake paths and thus extending the time required to trace back to the source node. Existing SLP solutions that use fake transmission paths only consider individual source nodes, where the fake paths constructed for different source nodes are independent and cannot cooperate to resist attacks. In this paper, a Fake Path Co-Construction source location privacy protection protocol (FPCC) suitable for UWSNs is proposed. FPCC combines the existing transmission paths and creates co-constructed fake paths to simultaneously protect two source nodes. Simulation results confirm that FPCC, when compared with existing well-performed SLP protection protocols, extends safety time without increasing the number of nodes involved in transmitting fake packets.

Wi-Fi fingerprint-based indoor localization methods rely on establishing a complete fingerprint database of the target area. However, fingerprint collection usually requires a significant amount of labor and time. We propose that a target area with only partially labeled fingerprints can be expanded to become completely labeled through human walking trajectories. Data collectors (trainers) are requested to move within the field to obtain unlabeled fingerprint trajectories. To enable label expansion, a series of corrections is made to ensure that the walking trajectories of trainers are correctly predicted and, consequently, labeled. This work greatly reduces the manual cost of fingerprint labeling.

This study proposes an Internet of Things (IoT)-based smart home system that sends and receives packets through the RS232 protocol and processes them using deep learning. A Field-Programmable Gate Array (FPGA) development board serves as a transceiver, operating a universal serial bus (USB) interface and a Wi-Fi module. The proposed system comprises a built-in wireless transceiver, a set of sensors, a development board running a deep learning algorithm, an MQTT communication protocol, and a terminal device controller. The objective is to implement an IoT -based smart home system with wireless data transmission. Node-RED is used to develop a comprehensive smart home system on the server side for IoT applications, facilitating data access, data processing, and terminal device control. The aim is to achieve automatic regulation and ensure comfortable indoor temperatures. In experiments, the root mean squared error difference between actual and predicted temperatures was approximately 0.4426 °C. After evaluation experiments with the FPGA development board, an application-specific integrated circuit (ASIC) based on the TSMC 0.18-μm CMOS process was used. Simulation results indicate that the chip area is approximately 1.186 × 1.188 mm2, and the dynamic power consumption is approximately 8.1674 mW at a power supply of 1.8 V and operating frequencies of 50 and 5 MHz.

In this paper, we explore the problem of minimizing the uplink and downlink age of information (AoI) in non-orthogonal multiple access (NOMA)-based low Earth orbit (LEO) networks for Internet of Things (loT) devices. The objective is to effectively manage the freshness of information across loT devices while maintaining a fair power allocation over time. We propose two novel AoI models that accurately capture the freshness of information for devices during both uplink and downlink transmissions, accounting for factors such as propagation delay, satellite handover delay, and inter-satellite link transmission time. To minimize the time-average AoI, we propose an energy-aware AoI (EA-AoI) algorithm that combines a deep reinforcement learning (DRL)-based scheduling approach with a low-complexity power allocation scheme. Simulation results demonstrate that the proposed EA-AoI scheduling algorithm outperforms successive interference cancellation (SIC)-based scheduling, AoI-greedy scheduling, and random scheduling approaches in terms of lower AoI values in the uplink and downlink directions through both simulated and real-world satellite trajectories. Furthermore, the algorithm ensures that the long-term average allocated power remains within a predefined threshold, thus striking an optimal balance between AoI and power consumption for NOMA-based LEO satellite networks.

Age of Information (AoI) is a critical performance metric for measuring information freshness in task-oriented Internet of Things (IoT) applications. In this work, we investigate how AoI impacts task-oriented multicasting in multi-cell Non-orthogonal Multiple Access (NOMA) networks, while considering the effects of inter-cell interference and multicast group size. Our findings reveal that larger group sizes lead to increased instability in AoI, but lower inter-cell interference; while smaller group sizes result in higher inter-cell interference but a smaller increase in AoI when transmission fails. To balance inter-cell interference and group size, we propose a novel AoI-aware inter-cell interference mitigation (AIM) solution, which dynamically adjusts multicast group size, schedules multicast groups for transmission, and allocates transmit power to multicast groups. We derive the optimality for achieving the lowest AoI of the AIM scheduling strategy and the upper bound of the number of scheduling groups. Our simulation results show that AIM ensures more successful transmissions and low weighted sum AoI through grouping and AoI-aware inter-cell interference mitigation. We also show that our AoI-aware grouping scheme outperforms random grouping and fixed grouping, as well as evolutionary-based genetic algorithm grouping, by modeling the grouping problem as a multi-armed bandit problem and solving it using the ϵ-greedy method. To the best of our knowledge, this is the first study to examine the AoI problem of task-oriented IoT applications in multi-cell multicast NOMA networks.

With the rapid development of smart devices and wireless communication technologies, IEEE 802.11ah (WiFi HaLow) is designed to solve one of the major problems of Internet of Things (IoT): high collision probability in dense networks. It proposes the Restricted Access Window (RAW) mechanism, where stations (sensors) are partitioned into groups for time-division channel access. The grouping strategy, which highly influences network performance, needs to consider factors including the number of stations per group, and stations’ data rates, and locations. With the advance of artificial intelligence technologies, we ponder whether deep learning can help solving this station grouping problem. In this paper, we propose a multilayer perceptron (MLP) model to predict RAW performance. More precisely, the model predicts the corresponding throughputs and packet loss rates of a given set of RAW configurations. Thus, based on the predicted results, we can determine proper RAW parameters. We have validated the proposed method by ns-3 simulations.

Collecting surrounding vehicles' motion information is one of the key issues for accident prevention and autonomous driving. Although multi-vehicle simulation frameworks are widely provided, We need a platform that enable inter-vehicle V2X communications. In this work, based on the open source simulation platform, CARLA, we extend and implement several modules to build a V2X simulation framework. In the proposed framework, vehicles are allowed to share their profiles and sensory data through V2X communications. With the motion information of other vehicles, a car can thus make more intelligent decisions. To validate the effectiveness of the framework, we run simulations in variose scenarios. Each time, a primary vehicle is selected and then both its sensory data and received surrounding vehicles' information are output and recorded in a simulated dataset. It is shown that with the dataset and our multi-vehicle data fusion algorithm, the primary vehicle can visually see the driving status of surrounding cars, which can greatly help a vehicle to choose a better driving strategy. This work not only proposes a V2X communication-enabled multi-vehicle simulation framework based on CARLA, but also provides a low cost way to generate simulated V2X datasets.

Evolutionary algorithms (EAs) have shown their great capability of handling optimization problems. In the domain of large scale global optimization, many distributed EAs (dEAs) have been proposed for maintaining population diversity so as to enhance their efficacy. One well-known variant of dEA is the island model EA, in which several populations form islands communicating through migration. There are several key factors that affect the performance and search behavior of island model EA, such as population size of each island and migration topology. While most studies of dEA focus on low or medium dimensional problems, an investigation into the effects of these factors on high dimensional problems is greatly needed. This study presents an empirical analysis of island model EA on large scale global optimization problems. The analysis examines the solution quality, convergence speed, and population diversity of island model EA with different migration topologies, population sizes, migration rates, and migration frequencies on four benchmark function of 1,000 dimensions. The results render guidelines for using island model EA to solve large scale global optimization problems.

Despite the recent success of multi-modal action recognition in videos, in reality, we usually confront the situation that some data are not available beforehand, especially for multi-modal data. For example, while vision and audio data are required to address the multi-modal action recognition, audio tracks in videos are easily lost due to the broken files or the limitation of devices. To cope with this sound-missing problem, we present an approach to simulating deep audio feature from merely spatial-temporal vision data. We demonstrate that adding the simulating sound feature can significantly assist the multi-modal action recognition task. Evaluating our method on the Moments in Time (MIT) Dataset , we show that our proposed method performs favorably against the two-stream architecture, enabling a richer understanding of multi-modal action recognition in video.

It is crucial for future 5G networks to intelligently understand how users move so that the networks can allocate different resources efficiently. In this paper, we try to find practical features to identify four common types of motorized transportations, including High-Speed Rail (HSR), subway, railway, and highway. We propose a system architecture that can provide accurate, real-time, and adaptive solution by using cellular information only. Because we do not use GPS as that in most of the prior studies, we can reduce energy consumption, size of log data, and computational time. Around 500-hour data are collected for performance evaluation. Experimental results confirm the effectiveness of the proposed algorithm, which can improve well-known machine learning algorithms to approximately 98% classification accuracy. The results also show that battery consumption can be reduced about 37%.

Collecting vehicle surrounding information is a key issue for accident prevention and autonomous driving applications. Although GPS and 4G/LTE are widely accepted, it is still a challenge for a vehicle to get complete information of its surrounding vehicles. In this work, we consider the integration of multi-sensory data through V2V communications to help a vehicle to understand its complex surroundings. We propose a fusion algorithm that can integrate four types of sensory inputs: V2V communications, GPS, camera, and inertial data. We show that through such fusion, it is possible for a vehicle to visually see the driving states of its surrounding vehicles.

Mobile Crowd Sensing (MCS) is a promising paradigm to collect large-scale network data globally. However, how to motivate people to collect and share data is a challenge. We believe the major reason why many MCS systems are not pervasive is because there are no incentives for people to use them. In this paper, we present SensingGO, a system which encourages people to keep sensing data by integrating incentive mechanisms. The sensed data are then transmitted to our backend server. The data we collected and the source code of SensingGO are open to anyone freely. We also demonstrate the analysis of real mobile data collected from SensingGO.

Mobile Augmented Reality (MAR) is generally defined as the service that is capable of enhancing the real-world camera views of a smartphone with extra information on top of the screen. Thus, MAR has triggered strong interests in mobile e-commerce, location-based service, etc. However, MAR is usually constrained by the local storage and computing power of the device as well as the latency and bandwidth of the underlying wireless channel. The computation demand is high because a targeted object needs to be compared to lots of images in the database. The communication demand is high because potential objects need to be continuously transmitted to the server and the augmented information needs to be downloaded from the server. This paper investigates the Local Cache Management (LCM) problem to manage the computation and communication work of a mobile device and harmonize the local and remote workloads. We present strategies to pre-fetch higher-priority objects from the server and replace lower-priority objects in the local cache based on temporal and spatial access locality. We verify the effectiveness of our strategies in term of cache hit ratio, response latency, and remote requests via simulations.

In this paper, the target patrolling problem was considered, in which a set of mobile data collectors, known as data mules (DMs), must efficiently patrol a given set of targets. Because the time interval (or visiting interval) between consecutive visits to each target reflects the degree to which that target is monitored, the goal of this paper was to balance the visiting interval of each target. This paper first presents the basic target points patrolling algorithm, which enables an efficient patrolling route to be constructed for numerous DMs, such that the visiting intervals of all target points are stable. For scenarios containing weighted target points, a weighted target points patrolling (W-TPP) algorithm is presented, which ensures that targets with higher weights have higher data collection frequencies. The energy constraint of each DM was also considered, and this paper presents a W-TPP with recharge (RW-TPP) algorithm, which treats the energy recharge station as a weighted target and arranges for DMs to visit the recharge station before running out of energy. The performance results demonstrated that the proposed algorithms outperformed existing approaches in average visiting frequency, DM movement distance, average quality of monitoring satisfaction rate, and efficiency index.

ABSTRACT In wireless sensor networks (WSNs), coverage of the monitoring area represents the surveillance quality. Since sensor nodes are battery powered and placed outdoor, there will be failures due to energy exhaustion or environmental influence, resulting in coverage‐loss. In literature, a number of studies developed robot repairing algorithms that aim at maintaining full coverage. However, they did not consider the time constraint for network maintenance. Furthermore, they did not consider the existence of obstacles and the constraint of limited energy of the robot. This paper presents a novel tracking mechanism and robot repairing algorithm for maintaining the coverage quality of the given WSN. Without support of location information, the tracking mechanism leaves robot's footmark on sensors so that they can learn better routes for sending repairing requests to the robot. Upon receiving several repairing request messages, the robot applies the proposed repairing algorithm to establish an efficient route that passes through all failure regions with low overhead in terms of the required time and the power consumption. In addition, the proposed repairing algorithm also considers the remaining energy of the robot so that the robot can move back to home for recharging energy and overcome the unpredicted obstacles. Performance results reveal that the developed protocol can efficiently maintain the coverage quality while the required time and energy consumption are significantly reduced. Copyright © 2011 John Wiley & Sons, Ltd.

This paper considers the target patrolling problem which asks a set of mobile data mules to efficiently patrol a set of given targets. Since the time interval (also referred to visiting interval) for consecutively visiting to each target reflects the monitoring quality of this target, the goal of this research is to minimize the maximal visiting interval. This paper firstly proposes a basic algorithm, called Basic (B-TCTP), which aims at constructing an efficient patrolling route for a number of given data mules such that the visiting intervals of all target points can be minimized. For the scenario containing weighted target points, a Weighted-TCTP (W-TCTP) algorithm is further proposed to satisfy the demand that targets with higher weights have higher data collection frequencies. By considering the energy constraint of each data mule, this paper additionally proposes a RW-TCTP algorithm which treats energy recharge station as a weighted target and arranges the data mules visiting the recharge station before exhausting their energies. Performance study demonstrates that the proposed algorithms outperform existing approaches in terms of visiting intervals of the given targets and length of patrolling path.

Bounding-box mechanism is a well known low-cost localization approach for wireless sensor networks. However, the bounding-box location information can not distinguish the relative locations of neighboring sensors, hence leading to a poor performance for some applications such as location-aware routing. This paper proposes a Distinguishing Relative Locations (DRL) mechanism which uses a mobile anchor to broadcast tones and beacons aiming at distinguishing the relative locations of any two neighboring nodes. Experimental study reveals that the proposed DRL mechanism effectively distinguishes relative locations of any two neighboring nodes and hence significantly improves the performance of location-aware routing in wireless sensor networks (WSNs).

In wireless mobile sensor networks, spatial full-coverage only can be achieved when the surplus mobile sensors contribute a larger coverage area than the hole size. The temporal full coverage problem asks to monitor every point of a given monitoring region within a specific time interval. This paper considers a mobile WSN that contains holes but exists no redundant mobile sensor to heal the hole. To achieve the temporal full-coverage purpose, a distributed hole-movement mechanism is proposed to balance the energy consumptions of mobile sensors. Simulation study reveals that the proposed hole-movement mechanism enhances the coverage ratio of WSN and balance the energy consumption of mobile sensor nodes.

Object tracking is one of important applications in wireless sensor networks. The key issues involved in object tracking are object detection, target classification, location estimation, and location management. The main theme of this paper is location management. Tree-based location management schemes have been studied extensively. However, the traditional tree-based schemes have two major drawbacks. First, the structure maintenance cost is high. Second, taking statistics is required. This paper proposes a quadtree-based location management scheme in which the maintenance cost is negligible, and no statistics has to be taken. Performance evaluations are also conducted through simulations to demonstrate the efficiency of the proposed scheme.

An important issue of wireless sensor networks is object tracking, where the key steps include event detection, target classification, location estimation and location management. The main theme of this paper is location management. Because imprecision is an inherent property in object-tracking sensor networks, this paper focuses on the scenarios where users can tolerate a certain degree of imprecision in their query results. We intend to develop a location management scheme that can achieve two goals. First, multiple precision levels are provided. Second, the query cost is proportional to the precision level. To achieve these two goals, we propose a tree-based imprecision-tolerant location management scheme that includes three major components: (1) update and query mechanisms that can support imprecision-tolerant queries, (2) the approach to taking the statistics of imprecision-tolerant queries and (3) a tree construction algorithm that can reduce the query cost and minimize the increment of update cost. Performance evaluations are conducted through simulations to verify the proposed scheme.

Block Truncation Coding (BTC) is an efficient compression technique for its inherent simple coding strategy. However, the annoying blocking effect and false contour accompanied in high coding gain configurations make the applications relatively limited compares to some up-to-date compression schemes. For this, Error-Diffused Block Truncation Coding (EDBTC) is proposed to solve these problems and obtain satisfactory results. Unfortunately, the EDBTC sacrifices the parallel advantage of traditional BTC. Moreover, the number of diffused directions of EDBTC can be reduced to obtain higher efficiency. For these, the Interlaced Error-Diffused Block Truncation Coding (IEDBTC) is proposed in this work to claim back the parallel advantage. In addition, the diffused elements are also reduced from four to two with the proposed optimization procedure while preserving the image quality.

Block Truncation Coding (BTC) is an efficient compression technique for its inherent simple coding strategy. However, the annoying blocking effect and false contour accompanied in high coding gain configurations make the applications relatively limited compares to some up-to-date compression schemes. For this, Error-Diffused Block Truncation Coding (EDBTC) is proposed to solve these problems and obtain satisfactory results. Unfortunately, the EDBTC sacrifices the parallel advantage of traditional BTC. Moreover, the number of diffused directions of EDBTC can be reduced to obtain higher efficiency. For these, the Interlaced Error-Diffused Block Truncation Coding (IEDBTC) is proposed in this work to claim back the parallel advantage. In addition, the diffused elements are also reduced from four to two with the proposed optimization procedure while preserving the image quality.

The rapid progress of wireless communication and embedded microsensing MEMS technologies has made wireless sensor networks possible. In light of storage in sensors, a sensor network can be considered as a distributed database, in which one can conduct in-network data processing. An important issue of wireless sensor networks is object tracking, which typically involves two basic operations: update and query. This issue has been intensively studied in other areas, such as cellular networks. However, the in-network processing characteristic of sensor networks has posed new challenges to this issue. In this paper, we develop several tree structures for in-network object tracking which take the physical topology of the sensor network into consideration. The optimization process has two stages. The first stage tries to reduce the location update cost based on a deviation-avoidance principle and a highest-weight-first principle. The second stage further adjusts the tree obtained in the first stage to reduce the query cost. The way we model this problem allows us to analytically formulate the cost of object tracking given the update and query rates of objects. Extensive simulations are conducted, which show a significant improvement over existing solutions.

One important application of wireless sensor networks is the tracking of moving objects. The recent progress has made it possible for tiny sensors to have more computing power and storage space. Therefore, a sensor network can be considered as a distributed database, on which one can conduct in-network data processing. This paper considers in-network moving object tracking in a sensor network. This typically consists of two operations: location update and query. We propose a message-pruning tree structure that is an extension of the earlier work (H.T. Kung and D. Vlah, March 2003), which assumes the existence of a logical structure to connect sensors in the network. We formulate this problem as an optimization problem. The formulation allows us to take into account the physical structure of the sensor network, thus leading to more efficient solutions than in the previous paper of H.T. Kung and D. Vlah (March 2003) in terms of communication costs. We evaluate updating and querying costs through simulations.

In a mobile ad hoc network (MANET), one essential issue is Medium Access Control (MAC), which addresses how to utilize the radio spectrum efficiently and to resolve potential contention and collision among mobile hosts on using the medium. Existing works have been dedicated to using multiple channels and power control to improve the performance of MANET. In this paper, we investigate the possibility of bringing the concepts of power control and multi-channel medium access together in the MAC design problem in a MANET. Existing protocols only address one of these issues independently. The proposed protocol is characterized by the following features: (i) it follows an ‘on-demand’ style to assign channels to mobile hosts, (ii) the number of channels required is independent of the network topology and degree, (iii) it flexibly adapts to host mobility, (iv) no form of clock synchronization is required and (v) power control is used to exploit frequency reuse. Power control may also extend battery life and reduce signal interference, both of which are important in wireless communication. Through simulations, we demonstrate the advantage of our new protocol.