Search

The rapid expansion of AI-generated content (AIGC) reflects the iteration from assistive AI towards generative AI (GAI). Meanwhile, the 6G networks will also evolve from the Internet-of-Everything to the Internet-of-Intelligence. However, they seem to be an odd couple, due to the contradiction of data and resources. To achieve a better-coordinated interplay between GAI and 6G, the GAI-native Networks (GainNet), a GAI-oriented collaborative cloud-edge-end intelligence framework, is proposed in this article. By deeply integrating GAI with 6G network design, GainNet realizes the positive closed-loop knowledge flow and sustainable-evolution GAI model optimization. On this basis, the GAI-oriented generic Resource Orchestration Mechanism with Integrated Sensing, Communication, and Computing (GaiRomISCC) is proposed to guarantee the efficient operation of GainNet. Two simple case studies demonstrate the effectiveness and robustness of the proposed schemes. Finally, we envision the key challenges and future directions concerning the interplay between GAI models and 6G networks.

Vehicular Clouds (VCs) have attracted tremendous attention for offering commendable computing services to vehicles with computation-intensive tasks. Such tasks are often represented as Directed Acyclic Graphs (DAGs) consisting of several interdependent subtasks and directed edges. Processing of DAG tasks often needs pooling the computation resources of vehicles. However, the selfishness of vehicles prevents them from sharing their resources. To this end, we propose a Truthful Dynamic Reverse Auction (TDRA) mechanism to motivate vehicles to participate in service provisioning. To realize TDRA, we first propose an enumeration-based allocation strategy to optimally allocate subtasks among vehicles and obtain a Vickrey-Clarke-Groves (VCG)-based pricing strategy that can ensure the economic properties of individual rationality and truthfulness. Then, to deal with the high computational complexity of obtaining the optimal solution, we develop a near-optimal Dynamic Bilateral Ranking (DBR) allocation strategy to allocate subtasks within polynomial time and design a critical value-based pricing strategy that can also guarantee the two above-mentioned economic properties. Through simulating real-world movement traces of vehicles, we demonstrate that DBR outperforms the existing benchmarks, and verify our theoretical analysis on the economic properties of our developed pricing strategy.

Implementing either Federated learning (FL) or split learning (SL) over clients with limited computation/communication resources faces challenges on achieving delay-efficient model training. To overcome such challenges, we investigate a novel distributed Clustering-based Hybrid fEdErated Split lEarning (CHEESE) framework, consolidating distributed resources among clients by device-to-device (D2D) communications, working in an intra-serial inter-parallel manner. In CHEESE, each learning client can form a cluster with its neighboring helping clients via D2D communications to train an FL model collaboratively. Inside each cluster, the model is split into multiple segments via a model splitting and allocation (MSA) strategy, while each cluster member trains one segment. After completing intra-cluster training, a transmission client (TC) is determined from each cluster to upload a complete model to the base station for global model aggregation under allocated bandwidth. Accordingly, an overall training delay cost minimization problem is formulated, involving the following subproblems: client clustering, MSA, TC selection, and bandwidth allocation. Due to its NP-Hardness, the problem is decoupled and solved iteratively. The client clustering problem is first transformed into a distributed clustering game based on potential game theory, where each cluster further investigates the remaining three subproblems to evaluate the utility of each clustering strategy. Specifically, a heuristic algorithm is proposed to solve the MSA problem under a given clustering strategy, while a greedy-based convex optimization approach is introduced to solve the joint TC selection and bandwidth allocation problem. Extensive experiments on practical models and datasets demonstrate that CHEESE can significantly reduce training delay costs.

Vehicular cloud (VC) platforms integrate heterogeneous and distributed resources of moving vehicles to offer timely and cost-effective computing services. However, the dynamic nature of VCs (i.e., limited contact duration among vehicles), caused by vehicles' mobility, poses unique challenges to the execution of computation-intensive applications/tasks with a directed acyclic graph (DAG) structure, where each task consists of multiple interdependent components (subtasks). In this paper, we study the scheduling of DAG tasks over dynamic VCs, where multiple subtasks of a DAG task are dispersed across vehicles and processed by vehicles cooperatively. We formulate DAG task scheduling as a 0-1 integer programming problem, aiming to minimize the overall task completion time while ensuring a high execution success rate, which turns out to be NP-hard. To tackle the problem, we develop a ranking and foresight-integrated dynamic scheduling scheme (RFID). RFID consists of i) a dynamic downward ranking mechanism that sorts the scheduling priority of different subtasks, while explicitly taking into account the sequential execution nature of DAG; ii) a resource scarcity-based priority changing mechanism that overcomes possible performance degradations caused by the volatility of VC resources; and iii) a degree-based weighted earliest finish time mechanism that assigns the subtask with the highest scheduling priority to the vehicle which offers rapid task execution along with reliable transmission links. Simulation results reveal the effectiveness of our proposed scheme in comparison to benchmark methods.

The development of the Internet of Vehicles has attracted a large number of innovative vehicular applications. Specifically, vehicular crowdsensing represents an emerging paradigm that utilizes on-board sensing and computing capacity to provide location-based services. The conventional vehicular crowdsensing method aims to improve the quality of service or reduce the cost while guaranteeing coverage and data quality. Inappropriate sensing vehicle selection may cause high execution latency, which, however, has rarely been considered. In this article, we propose a fog-based vehicular crowdsensing scheme, where on-board sensors can act as sensing participants and vehicles can act as fog nodes. It unifies crowdsensing participant selection and computing task allocation to jointly optimize execution latency and cost. Furthermore, we present a heuristic algorithm named as random chemical reaction optimization (RCRO) to solve the proposed problem. Finally, we evaluate the proposed scheme and algorithm by simulating real-world traffic. Numerical results show the performance of the RCRO algorithm and the superiority of the proposed method in comparison with the conventional method.

With the rapid development of the Internet technology, millimeter wave (mmWave) will be used as a supplement to 5G low frequency bands to meet the extremely high system capacity requirements of 5G in hot spots. Although 5G mmWave communication can adapt to the needs of 5G network and carry a large amount of transmitted data, transmission stability has become one of the key technical issues of 5G network mmWave communication due to problems such as strong attenuation and poor penetration of mmWave. In order to improve the efficiency of the mmWave multi-hop transmission, we propose a 5G mmWave multi-hop transmission method based on network coding, which can adapt to the current wireless network environment, improve spectrum efficiency and increase network throughput. Based on MATLAB simulation experiments, it is verified that the proposed method can greatly improve the transmission efficiency and reduce the signal loss under the premise of ensuring the accurate signal transmission.

As a kind of biocompatible material with long history, silk fibroin is one of the ideal platforms for on-skin and implantable electronic devices, especially for self-powered systems. In this work, to solve the intrinsic brittleness as well as poor chemical stability of pure silk fibroin film, mesoscopic doping of regenerated silk fibroin is introduced to promote the secondary structure transformation, resulting in huge improvement in mechanical flexibility (∼250% stretchable and 1000 bending cycles) and chemical stability (endure 100 °C and 3–11 pH). Based on such doped silk film (SF), a flexible, stretchable and fully bioabsorbable triboelectric nanogenerator (TENG) is developed to harvest biomechanical energy in vitro or in vivo for intelligent wireless communication, for example, such TENG can be attached on the fingers to intelligently control the electrochromic function of rearview mirrors, in which the transmittance can be easily adjusted by changing contact force or area. This robust TENG shows great potential application in intelligent vehicle, smart home and health care systems.

Metabolomics can indicate the physiological and biochemical responses of mosquitoes to different stimulants, including insecticides, which allow them to adapt to different inhospitable environments. Though metabolic differences between insecticide-resistant and -susceptible strains have been established for other mosquito species, such as Anophelesand Culex, it is yet to be done for Aedes albopictus(Skuse). In this study, nuclear magnetic resonance (NMR) spectroscopy–based metabolomic analysis performed on Ae. albopictusdeltamethrin-resistant and -susceptible strains showed significant differences in amino acid, organic acid, and sugar metabolism. Concentrations of neutral amino acids and sugars tended to be lower in the deltamethrin-resistant strain than in the deltamethrin-suceptible strain, but the concentration of basic and acidic amino acids and organic acids increased. All these changes might accommodate biochemical and physiological needs in deltamethrin-resistant mosquitoes, such as enzyme synthesis and detoxification. This was further confirmed by the predictable draft metabolic map. This is the first report using NMR spectroscopy to investigate the metabolic differences between deltamethrin-resistant and -susceptible strains of Ae. albopictus. To a certain degree, this demonstrates how Ae. albopictusdevelop insecticide resistance by metabolic reprograming to survive under the insecticide pressure.

In 5G era, it is expected to achieve wireless network coverage including offshore areas. Modeling of marine wireless channels is the basis of constructing a marine communication system. In this paper, a communication scene between an unmanned aerial vehicle (UAV) and a boat is simulated to study the marine wireless channel. Firstly, an improved spatial partitioning ray tracing algorithm is proposed to track the propagation path of electromagnetic waves at sea surface. Secondly, a mobile channel is simulated and modeled based on the track results. Finally, a loss measurement is carried out in the coastal waters based on the simple wireless channel loss measuring platform, and a path loss propagation model is built. Then we compare the actual measurement data with the simulation results and find that the two are have good consistency, which further verifies the reliability of the simulation.

Small cells are beneficial for both users and operators to improve communication quality. The femtocell is one type of small cell used for indoor coverage. Enterprise femtocell installation inevitably produces intralayer and interlayer interference management issues. Enterprise femtocell optimization and indoor propagation modeling are built in this work based on the communications environment characteristics. Q-learning-based distributed and hybrid power control strategies are then proposed. The throughput, energy efficiency, and user experience satisfaction with conventional scheduling methods are compared. The simulation results show that distributed Q-learning performs better local optimization while hybrid Q-learning enhances global performance.

Machine-type communication is undergoing ubiquitous development by connecting with trillions of intelligent devices and providing a broad range of applications through all of our lives. MTC can access different networks, which is similar to a fusion of heterogeneous networks, including LTE, WiFi, ZigBee, and so on. However, security and privacy issues still bring about a great challenge for the development of MTC, particularly secure transmission under heterogeneous networks. This article first presents an overview of the MTC architecture and its evolution toward LTE, security threats, and solutions. Then, as 5G systems have been designed to support a diverse set of communications, we propose a scalable cross-layer authentication solution based on MTC devices’ hardware fingerprints toward an LTE heterogeneous network.

We investigate a data quality-aware dynamic client selection problem for multiple federated learning (FL) services in a wireless network, where each client offers dynamic datasets for the simultaneous training of multiple FL services, and each FL service demander has to pay for the clients under constrained monetary budgets. The problem is formalized as a non-cooperative Markov game over the training rounds. A multi-agent hybrid deep reinforcement learning-based algorithm is proposed to optimize the joint client selection and payment actions, while avoiding action conflicts. Simulation results indicate that our proposed algorithm can significantly improve training performance.

An efficient carrier allocation algorithm is important for self-organised networks (SONs). SON can significantly improve network performance and user experience which greatly help relieve the heavy loading of the current multimedia applications for mobile environments. This paper proposes a distributed algorithm for carrier allocation using a reverse auction algorithm, wherein the original cells affected by newly added cells self-organise into auction groups to facilitate new carrier allocation. The computational overhead required to adapt to topology changes can be significantly reduced compared with the conventional topology-based approaches that ignore prior assignments. The number of original cells affected by the reallocation is as small as possible in the reverse auction algorithm while optimising carrier allocation. We investigate the algorithm performance in various scenarios. Our results indicate that the reverse auction algorithm exhibits significant advantages compared with traditional approaches, especially in highly dynamic networks where frequent resource re-allocation is needed.

Multiple antenna technology for wireless communications is becoming mature and wide ranging in the wireless broadband network. Many spectrum usage measurement reports have shown that inaccurate spectrum sensing will waste spectrum resources. A Multiple Antenna Asynchronous Cognitive MAC (MAAC-MAC) protocol is proposed through introducing a multiple antenna architecture and power control mechanism into the hardware-constrained cognitive MAC (MHC-MAC) protocol with Asynchronous-Assembly Line Mode in cognitive radio network. In the spectrum sensing stage, the protocol makes use of multiple antennas to perform spectrum sensing and positioning, aimed at improving spectrum sensing and detection performance. During the negotiation phase, the CU (Cognitive User) access power control mechanism fully utilizes the licensed band without interfering PU (Primary User) and also increases the transmission rate and system throughput. It may be used for improved positioning and location-based services both indoors and outdoors for IT convergence.

The random access channel model with retransmission packet mechanism is studied in the emergency wireless broadband access network (EWBAN). The success probability and access time delay packet request for MAC protocol access in EWBAN is analysed. The random access channel model with EWBAN retransmission packet mechanism and the MAC protocol is established using simulation results. Theoretical and engineering application guidance is provided for emergency wireless broadband communications.

The caching‐enabled mobile edge computing (MEC) system becomes increasingly crucial to the future long‐term evolution vehicle‐to‐infrastructure (LTE‐V2I) network, especially for fast delivery of popular content of delay‐sensitive applications in the backhaul capacity–limited vehicular networks. Most related studies mainly focus on computation offloading strategy design and coding caching, while the deployment strategy design of caching capacities for high‐speed vehicles has been overlooked. In this paper, we investigate the deployment problem in LTE‐V2I networks with particular consideration on high‐speed vehicles. First, in a one‐way scenario, a joint optimization framework is formulated to minimize the cache size of the MEC system, meanwhile maximizing the average downloading percentage, where the MEC system's backhaul capacities, vehicle speeds, content popularity distributions, and sides of the highway are taken into account. The impacts of LTE‐V2I system parameters, eg, vehicle speed, on the objective function are revealed through the one‐way optimization. Then, the two‐way optimization is formulated and solved by developing an iterative optimization approach based on the aforementioned one‐way scenario. Particularly, the optimal caching allocation between two ways is derived. Simulation results validate the effectiveness of the proposed approaches even with low backhaul capacities and high vehicle speed. In this paper, we investigate the deployment problem in LTE‐V2I networks with particular consideration on high‐speed vehicles. Two‐way optimization is formulated and solved by developing an iterative optimization approach based on the one‐way scenario. Particularly, the optimal caching allocation between the two ways is derived.