Time-Sensitive Networking (TSN) enables deterministic communication over standard Ethernet, critical for Industrial Internet of Things (IIoT) applications. However, scheduling traffic flows with non-harmonic periods poses a considerable challenge for existing online methods. This paper proposes an integrated framework consisting of conflict-aware pre-routing with a recovery scheduling strategy to enhance schedulability for non-harmonic traffic. In this paper, we first propose a non-harmonic-aware scheduling conflict index (KC) that quantifies scheduling difficulty via incorporating packet sizes and the LCM/GCD relationship between periods, and integrate it into a conflict-aware pre-routing algorithm. Then, we present a two-phase scheduling strategy in which the recovery phase combines sparsification—achieved by maximizing transmission start times—with a novel jitter-decreasing control that first applies the maximum allowable jitter and progressively reduces it. Finally, we carry out thorough experimental evaluations on a realistic topology. Results show that for non-harmonic periods, our method achieves approximately 10% better schedulability than IRAS. The recovery mechanism achieves an over 95% success rate and sub-second execution time, meeting online industrial configuration requirements.

Time-Sensitive Networking (TSN) is essential for Industrial Internet of Things (IIoT) applications demanding deterministic communication with zero jitter. Existing approaches are limited: Time-Aware Shaper (TAS) exhibits high computational complexity, while Cyclic Queuing and Forwarding (CQF) cannot guarantee zero jitter. Current integrated solutions fail to address scalability challenges for isochronous flows. We propose Zero-Jitter Cyclic Queuing and Forwarding (ZCQF). This novel architecture uses CQF for initial transmission and implements a reordering mechanism with TAS at the final hop to achieve zero-jitter transmission. Experimental results demonstrate that ZCQF outperforms existing methods across multiple network topologies, achieving higher schedulability and shorter runtime, validating its effectiveness for time-critical IIoT environments.

Wireless Local Area Networks (WLANs) are pivotal in facilitating wireless data transmission services in modern society. Furthermore, the proliferation of diverse audio and video platforms and the increasing prevalence of augmented reality (AR) and virtual reality (VR) devices has led to a significant surge in network traffic. This surge has heightened consumers' demands for improved network speed, reduced latency, and enhanced reliability. The recently updated IEEE 802.11be standard, also known as Wi-Fi 7, introduces an innovative approach to data transmission called Multi-Link Operation (MLO). Within MLO, the enhanced Multi-Link Single Radio (eMLSR) mode stands out prominently. This technique em-powers Wi-Fi access points (APs) and user devices (STAs) to utilize multiple antennas for establishing simultaneous connections across different channels. Via enabling fast channel switching, it allows Wi-Fi APs and STAs to adapt swiftly to changing environmental conditions, enhancing transmission throughput and dependability. However, the effectiveness of STAs can influence channel-switching delays, and excessive channel-switching (transition) frequencies may degrade network performance. To address these challenges, this paper proposes an innovative channel-switching scheme that harnesses the rapid switching capabilities of eMLSR to evaluate and select the optimal transmission channel. We present a neural network model, informed by our analysis of the interplay between various environmental factors and network performance, to guide the decision-making process of this scheme. The proposed scheme excels in making precise channel-switching decisions, including selecting the best channel and dynamic-changing transition frequency. Experimental results demonstrate that the proposed scheme enhances network throughput across diverse channel interference conditions.

Various modern industrial applications necessitate deterministic communication, assuring that each packet of a data flow arrives at a specific time. Most existing Time-Sensitive Networking (TSN) scheduling mechanisms enforce temporary isolation (any two flows cannot simultaneously be in the same queue in a network device) to ensure such deterministic property. This approach has no significant side effects in a wired Ethernet environment. However, recent calls have been made for incorporating TSN and wireless technology, such as 5thgeneration (5G) network, due to the device’s mobility and desire for flexibility of many emerging industrial applications. In wireless environments, the fickle nature of wireless communication will cause the existing approach to require more network resources to maintain the deterministic property, resulting in decreased schedulability. Thus, this paper aims to propose a more efficient scheduling mechanism for deterministic communication under integrated 5G and TSN environments. First, we parameterize the residence time of the bridges and analyze the requirements of deterministic communication. Then, based on the analysis, we extend and develop new scheduling rules to improve schedulability while maintaining determinism. Finally, Our evaluation via performing a set of experiments in realistic settings showcases the improved schedulability of our method with the preservation of the deterministic property.

Addressing edge computing's economic issues is critical as we need to motivate edge devices as resource providers to devote their computing resources to the service. There are multiple users and resource providers in most edge computing service scenarios. The communication delays between users and providers are not the same since their physical distances are different. However, most existing works on edge computing regarding network economics adopt single market models that do not consider the influence of communication delay. Moreover, due to the high cost of deploying 5G ultra-dense small cells, barely a network operator can reach full network coverage. Users cannot rely on a single network, but existing multi-market models do not allow resource trading between different groups. Thus, in this paper, we propose a novel multi-market trading (MMT) framework to address these shortcomings. The framework combines the double auction at each group and a market selection game to enable the resource providers to participle in multiple auctions and analyze their behavior of choosing markets. Through extensive simulations using real-world datasets of vehicular networks, we show that the proposed framework can improve the social welfare by 14.45% and 36.74%, respectively, compared with classic multi-market and single market models.

Thanks to the advances in wireless communication technologies and the popularity of mobile devices, individuals can now record and produce personal videos. Furthermore, with the power of machine learning techniques, one can envision that a new paradigm for AIoT, a combination of AI and IoT, can accomplish many novel IoT applications. A feasible, safe, and trusty data trading environment is necessary to realize this AIoT service paradigm. The classic approach of centralized trust centers is not suitable for the targeted scenario due to privacy, vulnerability, and efficiency concerns. The emerging blockchain technology could be the perfect solution; yet, it remains a challenging issue to ensure data provider's privacy. Thus, this paper envisions a blockchain- based AIoT service paradigm and proposes solutions for preserving privacy. To preserve privacy in data storage, we adopt the approach of a private InterPlanetary File System (IPFS) network. For preserving privacy in on-chain metadata, especially the GPS records of the personal videos, we propose a privacy-preserving geometric search scheme (PPGSS). Through extensive experiments using real- world datasets and settings, we demonstrate the effectiveness of the proposed scheme.

Recently, a novel network architecture called Time-Sensitive Networking (TSN), which supports high bandwidth and deterministic communication, emerges to satisfy the requirements of safety-critical real-time applications for industrial automation. TSN can operate in centralized and distributed models, and the distributed model is more suitable for the safety-critical applications that require high network availability and reliability. However, the existing approach for the TSN's fully distributed model makes a pessimistic latency bound estimation due to the problems of lower priority streams and dependency. The pessimistic estimation will result in a decrease in the schedulability of TSN streams. This paper identifies and evaluates the impact of the two problems via a case study and then proposes a TDMA-based stream reservation approach to alleviate the two problems. The simulation results validate the proposed approach in terms of the number of accepted streams and computation time.

To meet the needs of many IoT applications with low-latency requirement, fog computing has been proposed for next-generation mobile networks to migrate the computing from the cloud to the edge of the network. In this paper, we study the fog computing service deployment problem, where the operator allocates and deploys the required computing and network resources on the edge of the network to accommodate the requests of various applications operated by the application service providers (ASPs). The operator negotiates with the ASPs to determine serving QoS of applications and how much to pay. A queuing-based latency performance model with bulk arrival is proposed for the problem to estimate the resources needed for the fog network to achieve the QoS requirements of applications. We then model and analyze the interactions between the operator and multiple ASPs as sequential one-to-many bargaining using Nash bargaining. Next, to find the optimal bargaining sequence, we propose an improved optimal algorithm, along with fast heuristic algorithms, to find the optimal sequence with low complexity. Through extensive simulations, we show that the fog service can benefit all parties, and the proposed optimal and heuristic algorithms can improve the OP's payoff by averages of 21.24 and 14.16 percent respectively.

Thanks to the advances in wireless communication and machine learning technologies, we can envision a novel AIoT (AI + IoT) service platform that collects video data from the individuals' edge devices. Then, it transforms the video data into useful information, providing services to IoT or smart city applications. However, collecting raw video data directly to the cloud server is merely possible due to network bandwidth limitations and data privacy concerns. One possible solution is to adopt federated learning, which enables edge devices to collaboratively train a shared model without sending the raw data to the cloud. Unfortunately, this scheme cannot directly be applied to the targeted scenario since it assumes labeled data for training, and only at the cloud, we have the human power and time to label the video data. Thus, to tackle those issues, we propose an edge learning system based on semisupervised learning and federated learning technologies. The system trains AI models at edge devices using an improved semisupervised learning scheme and periodically uploads the training results to the cloud server to form a single model by adapting the federated learning technology. Then, we observe that in the real world, the data on the end devices are nonindependent and identically distributed (non-IID) such that it may cause weight divergence during training and result in a considerable decrease in the model performance. Therefore, we propose a new operation called federated swapping (FedSwap) to replace partial federated learning operations based on a few shared data during federated training to alleviate the adverse impact of weight divergence. We evaluate our system on both image classification using the state-of-the-art benchmark data and object detection using real-world video data. The experimental results show that the proposed system can have up to 5.9% higher accuracy of object detection for the video analysis applications by fully utilizing unlabeled data, compared with the situation that only labeled data are used. Moreover, the proposed FedSwap can improve the accuracy of image classification by 3.8% and the object detection task by 1.1%.

Wireless body area networks (WBANs) have emerged recently to provide health monitoring for chronic patients. In a WBAN, the patient's smartphone is deemed an appropriate sink to help forward the sensing data to back-end servers. Through a real-world case study, we observe that temporary disconnection between sensors and the associated smartphone can happen frequently due to postural changes, causing a significant amount of data to be lost forever. In this paper, we propose a scheme to parasitize the data in surrounding Wi-Fi networks whenever temporary disconnection occurs. Specifically, we model data parasitizing as an optimization problem, with the objective of maximizing the system lifetime without any data loss. Then, we propose an optimal offline algorithm to solve the problem, as well as an online algorithm that allows practical implementations. We have also implemented a prototype system, where the online algorithm serves as the underlying technique, based on Arduino. To evaluate our scheme, we conduct a series of experiments with the prototype system in controlled and real-world environments. The results show that the lifetime is prolonged by 100 times, and it could be further doubled if the health monitoring application permits a few packet losses.

Emerging Internet of Things (IoT) applications share the same characteristics of involving multiple processing components (i.e., function modules) and requiring a massive amount of data to be processed with low latency. To meet these needs, edge/fog computing has been proposed for next-generation mobile networks to migrate the computing from the cloud to the edge of the network. Thanks to the development of Network Functions Virtualization (NFV), with which edge computing platform can virtualize function modules and deploy them on any edge devices to provide flexible services on the edge networks. However, such platform would need to deal with complicated function module calling relationship (i.e., call graph) of applications and user mobility, and both are not thoroughly considered by existing works of NFV and edge computing. In this paper, based on our previous idea of virtual local-hub (VLH), we propose a complete design of edge computing framework, which applies NFV technology on edge computing environment for IoT applications. To handle the complicated call graphs of IoT applications with better resource utilization, the VLH framework adapts the technologies of container-based virtualization and microservice architecture, which enables remote function module sharing on the edge computing environment. The framework includes the heuristic algorithm for function module allocation with the objective of minimizing total bandwidth consumption. We also present a design of protocols for system operations and mobility handling in the framework. Then we implement the framework on commodity hardware as a testbed. Via simulations under a large-scale environment with practical settings and experiments on the testbed under real-world scenarios, we demonstrate and verify the effectiveness and practicability of the proposed VLH framework for IoT application service provision.

Internet of Things (IoT) has recently emerged as an enabling technology for context-aware and interconnected “smart things.” Those smart things along with advanced power engineering and wireless communication technologies have realized the possibility of next generation electrical grid, smart grid, which allows users to deploy smart meters, monitoring their electric condition in real time. At the same time, increased environmental consciousness is driving electric companies to replace traditional generators with renewable energy sources which are already productive in user's homes. One of the most incentive ways is for electric companies to institute electricity buying-back schemes to encourage end users to generate more renewable energy. Different from the previous works, we consider renewable energy buying-back schemes with dynamic pricing to achieve the goal of energy efficiency for smart grids. We formulate the dynamic pricing problem as a convex optimization dual problem and propose a day-ahead time-dependent pricing scheme in a distributed manner which provides increased user privacy. The proposed framework seeks to achieve maximum benefits for both users and electric companies. To our best knowledge, this is one of the first attempts to tackle the time-dependent problem for smart grids with consideration of environmental benefits of renewable energy. Numerical results show that our proposed framework can significantly reduce peak time loading and efficiently balance system energy distribution.

Wearable devices, which are small electronic devices worn on a human body, are equipped with low level of processing and storage capacities and offer some types of integrated functionalities. Recently, wearable device is becoming increasingly popular, various kinds of wearable device are launched in the market; however, wearable devices require a powerful local-hub, most are smartphone, to replenish processing and storage capacities for advanced functionalities. Sometime it may be inconvenient to carry the local-hub (smartphone); thus, many wearable devices are equipped with Wi-Fi interface, enabling them to exchange data with local-hub though the Internet when the local-hub is not nearby. However, this results in long response time and restricted functionalities. In this paper, we present a virtual local-hub solution, which utilizes network equipment nearby (e.g., Wi-Fi APs) as the local-hub. Since migrating all applications serving the wearable devices respectively takes too much networking and storage resources, the proposed solution deploys function modules to multiple network nodes and enables remote function module sharing among different users and applications. To reduce the impact of the solution on the network bandwidth, we propose a heuristic algorithm for function module allocation with the objective of minimizing total bandwidth consumption. We conduct series of experiments, and the results show that the proposed solution can reduce the bandwidth consumption by up to half and still serve all requests given a large number of service requests.

To efficiently support and manage massive number of Internet of Things (IoT) short and bursty sessions, current long-term evolution (LTE) system needs to reduce signal load generated by IoT session setup/synchronization, while balancing the system performance, such as UE power consumption and delays to time-sensitive traffic. In LTE, radio resource control (RRC) and discontinuous reception (DRX) affect power consumption, signal load, and delay. We provide a session management methodology suitable for IoT traffic over LTE. Our analysis starts with a Markov chain analysis of the impact of DRX parameters. This is followed by an optimal uplink scheduler design and an IoT-aware adaptive DRX algorithm at the client, both of which modulate the tradeoff among signal load, delay, and power consumption. Scalability is also considered by providing a high-priority clustering-based adaptive DRX algorithm at eNB. Simulation results show that for packets with 0.1 s delay, our scheduler outperforms “Tx now” (and “Wait Till Deadline”) by 50% (and 30%) in power saving and by 60% (and 15%) in signal saving. With knowledge of the traffic pattern, IoT-aware adaptive DRX can further reduce signal load by 25%, especially for delay-sensitive traffic.

With the explosive growth in mobile data traffic, femtocell technology is regarded as the most effective way to enhance the mobile service quality and system capacity of cellular networks. However, the major problem with femtocell deployment is finding an appropriate access control mode that mobile operators and users are willing to adopt. Among the various kinds of access control modes, the hybrid access mode is considered the most promising because it allows femtocells to give preferential access to femtocell owners, while other public users can only access femtocells with certain restrictions. Because all femtocell owners are selfish, how to provide sufficient incentives so that they will share their femtocell resources is a challenging issue. To address the problem, we propose an economic framework for mobile operator and femtocell users based on game theoretical analysis. We also exploit the concept of revenue sharing, which provides a positive cycle to sustain the femtocell service. In the framework, a femtocell game is formulated where the femtocell owners determine the proportion of femtocell resources they will share with public users, while the operator maximizes its benefit by setting the ratio of the revenue distributed to femtocell owners. We analyze the existence and uniqueness of the Nash Equilibrium of the game. The results of extensive simulations show that the proposed framework maximizes the operator's benefit and satisfies the users' service requirements.

With the increasing sophistication and maturity of biomedical sensors and the significant advances on low-power circuits and wireless communications technologies, wireless body area networks (WBANs) have emerged recently to provide pervasive health monitoring for humans. In WBANs, smart phones can serve as data sinks to forward the sensing data to back-end servers. Due to the battery concern of smart phones and the postural changes of humans, temporary disconnection between sensors and their associated smart phones may frequently happen in WBANs. In this case, the sensing data would be lost when the limited memory space of sensors overflows. To prevent excessive data loss, this paper proposes a scheme to parasitize the data on existing public Wi-Fi networks, once the links from sensors to the smart phones become unavailable. Specifically, an optimization problem to maximize the time during which data loss can be avoided by exploiting the data parasitizing scheme is formulated, where a decision set of the packets' size and sending timing to public Wi-Fi networks needs to be determined. We develop an offline algorithm to obtain an optimal decision set and present an efficient online algorithm for practical implementations. The feasibility of the proposed scheme and the efficacy of the algorithms are demonstrated through prototype implementations on a WBAN testbed with biomedical sensor devices for real-world experiments.

With the explosive growth of mobile data traffic, the femtocell technology is one of the proper solutions to enhance mobile service quality and system capacity for cellular networks. However, one of the key problems for femtocell deployment is to find appropriate access control in which mobile operators and users are willing to be involved. Among all kinds of access control modes, the hybrid access mode is considered as the most promising one, which allows femtocells to provide preferential access to femtocell owners and subscribers while other public users can access femtocells with certain restriction. Since all femtocell owners are selfish, how to provide enough incentives to the owners for sharing their femtocell resources is challenging. In this paper, we construct an economic framework for mobile operator and femtocell users by a game theoretical analysis and introduce the concept of profit sharing to provide a positive cycle to sustain the femtocell service. In this framework, a femtocell game is formulated, where the femtocell owners determine the proportion of femtocell resources shared with public users while the operator can maximize its own benefit by determining the ratio of revenue distribution to femtocell owners. The existence of the Nash equilibrium of the game is analyzed. Extensive simulations are conducted to show that the profit of the operator can be maximized while the service requirements of users can be maintained by the proposed framework.

Wireless charging technology is considered as one of the promising solutions to solve the energy limitation problem for large-scale wireless sensor networks. Obviously, charger deployment is a critical issue since the number of chargers would be limited by the network construction budget, which makes the full-coverage deployment of chargers infeasible. In many of the applications targeted by large-scale wireless sensor networks, end-devices are usually equipped by the human and their movement follows some degree of regularity. Therefore in this paper, we utilize this property to deploy chargers with partial coverage, with an objective to maximize the survival rate of end-devices. We prove this problem is NP-hard, and propose an algorithm to tackle it. The simulation results show that our proposed algorithm can significantly increase the survival rate of end-devices. To our knowledge, this is one of very first works that consider charger deployment with partial coverage in wireless rechargeable sensor networks.

Zigbee, formalized by the IEEE 802.15.4 standard, is a specification for wireless personal area networks with low power, low cost, and a low data rate. In Zigbee, tree topology is commonly practiced to form wireless sensor networks and perform data delivery applications. In Zigbee wireless applications, data de livery failures occur constantly due to the node movements and topology changes of networks. To tackle the topology changes, conventional route reconstruction often involves huge resource consumption. In this paper, we utilize the regularity of mobility patterns to reduce the frequency of route reconstructions and achieve higher efficiency in sending data to mobile nodes. To increase the data delivery ratio, we introduce the metric of mobility-robustness in a tree topology, and propose tree construction with an objective to maximize the mobility-robustness of the constructed tree. We develop an efficient algorithm for effective tree construction. The effectiveness of network topologies constructed using this mobility-robustness metric is demonstrated by NS2 simulations against a real-world scenario.