The Credit-Based Shaping (CBS) algorithm is a traffic scheduling technique commonly used in Time-Sensitive Networking (TSN) to manage audio and video streams. CBS is designed to solve the problem of continuous blocking of low-priority data streams caused by high-bandwidth data streams occupying communication resources for long durations. However, in certain application scenarios with stringent jitter requirements, CBS may fall short in meeting the demands of real-time control data streams. In this paper, we improve the CBS algorithm by introducing a frame preemption mechanism and propose a time-sensitive scheduling algorithm to ensure real-time transmission of critical traffic. Additionally, we conduct worst-case response time (WCRT) analysis of the proposed scheduling scheme under both preemptive and non-preemptive scenarios. Finally, we validated our design through a vehicular network scenario. The results indicate that this method not only resolves the real-time transmission issues of critical traffic but also reduces the overall traffic transmission delay.

With the development of the marine Internet of Things (IoT), more sensors are deployed for different marine applications. Abundant data are generated that are required by terrestrial users. However, nowadays, the marine Internet cannot provide enough communication resources and high service quality for the feedback of those data. Thereby, fog computation is a potential method to solve this problem via data compression before transmission. To improve transmission efficiency while guaranteeing Quality of Service (QoS) of applications, how to efficiently schedule limited computation resources at fog nodes, such as buoys for different applications should be addressed. This article proposes an online computation scheduling algorithm to improve the communication resource efficiency with guaranteed QoS via appropriate compression. This scheme needs neither the cost function nor the constraint function at the scheduling node. What is more, the gradient information, which is usually needed in learning strategies, is not required, which is the estimation based on first-order and second-order schemes, respectively. Results show that the proposed scheme achieves the goal of improving communication efficiency while guaranteeing the data quality constraint of different applications.

In this study, a user pairing scheme in mobility‐aware device‐to‐device (D2D) communication system is proposed. One D2D user equipment (DUE) pair consists of a D2D transmitting user equipment and a D2D receiving user equipment. First of all, the network model, communication time model and moving model are presented. Then the mathematical characteristics of the capacities of DUEs in different moving mode are analysed. Based on the analysis above, a user pairing scheme in mobility‐aware D2D communication system is proposed. Finally, simulation results show the capacity performance of the proposed scheme outperforms that of the other algorithms. This result can be applied into the deployment and optimisation of D2D communication system.

Device-to-device (D2D) technique provides a feasible way for content delivery among proximal users without deploying additional infrastructures. For the realization of D2D content delivery, it is important to guarantee quality-of-service (QoS) for supporting real-time content transmissions. In this paper, we formulate the content delivery problem based on a flow model and investigate mechanisms providing delay QoS guarantee. The delay performance is analyzed for the content dissemination with the aid of D2D communication. Based on the analyses of parameters affecting the performance of content delivery, an optimal resource allocation scheme is proposed to achieve the best energy efficiency of content delivery in cellular networks with guaranteed delay constraints for hybrid content requests. Then effects of content caching on D2D content delivery is studied and a caching management scheme is proposed to avoid cache overflow of users with a certain probability and satisfy QoS requirements of content delivery. Simulation results are presented that verify the analyzed influencing factors on D2D content delivery and the proposed QoS-provisioning schemes.

Mode selection issue is studied for Device-to-Device (D2D) communication underlaying the two-tier cellular networks (including macro cells and small cells) to guarantee the Quality of Service (QoS) requirement not only for the cellular link but also for the D2D link. Firstly the link performances are investigated for UE under D2D transmission mode and cellular transmission mode, respectively. In addition, distance bounds are derived for potential D2D links as references to select the UE transmission mode. Lastly, a mode selection scheme is proposed based on these analyses which makes the base station determine an appropriate transmission mode for transmitter with the link performance satisfying QoS requirement. Simulation results show that link performances of cellular link and D2D link can be guaranteed by the proposed scheme. In addition, compared with the existing mode selection scheme, the proposed scheme improves the link performance significantly. The average outage probability can be reduced at most 92% for communication links by the proposed scheme.

Resource management issue is studied for device‐to‐device (D2D) communication in cellular networks to improve the overall system capacity and spectrum efficiency. Firstly, a lower bound of the interference distance is derived for a D2D user equipment's according to the transmitter density and the quality of servicerequirement of a D2D pair. By using this distance, an interference set can be chosen adaptively based on the interference condition to guarantee the signal‐to‐interference‐plus‐noise ratio of the D2D pair at a desired probability. Then, the authors propose two resource allocation schemes, the dual metric scheme and the TID (tolerant interference degree) scheme, to mitigate interference among coexistent D2D pairs. The dual metric scheme aims for uniform resource allocation to avoid excess interference in some resources. The TID scheme is designed to guarantee the interference at each D2D pair tolerable. The simulation results show that both the dual metric scheme and the TID scheme achieve more uniform resource allocation at the cost of a negligible increase of resource usage. Compared with the existing schemes such as the random scheme and the freedom degree scheme, the system performance in terms of resource efficiency is significantly improved by two proposed schemes.

D2D communication in LTE-A cellular networks can improve overall system performance such as throughput and power consumption substantially and can enable and enhance many proximity based applications. This research topic gains great interest in both academic and industrial fields in recent years. The relevant 3GPP standardization on D2D communication has already begun in several discussion groups. We introduce standardization progress in this area, and describe relevant architecture design to enable D2D communication. We also discuss the challenges to incorporate D2D communication into existing cellular networks.

Interference mitigation scheme is studied for Device-to-Device (D2D) communication in cellular networks to improve the overall system capacity and spectrum efficiency. Firstly a lower bound of the neighbor distance is derived for a D2D UE according to the transmitter density and the QoS requirement of a D2D pair. By using this distance, a neighbor set can be chosen adaptively based on the interference condition so that the SINR of the D2D pair can be guaranteed at a desired probability. In addition, a metric, tolerant interference degree (TID) is defined to evaluate the permissible interference scope for a successful D2D transmission. Finally, an interference mitigation scheme based on greedy coloring is designed according to the derived neighbor distance and TID to mitigate interference among D2D pairs. The simulation results show that the throughput is significantly improved by proposed scheme compared with the existing freedom degree scheme (from 135% to 220%). And the resource efficiency is up to 4 times higher than that of the freedom degree scheme.