With the development of smart agriculture, the Agricultural Artificial Intelligence Internet of Things (Agri-AIoT) has shown great potential in fields such as farmland monitoring and precision pesticide application. However, the computationally intensive tasks generated by massive heterogeneous agricultural sensing devices pose serious challenges to real-time and sustainability. To address this issue, this paper proposes a Cloud-Fog-Edge based collaborative computing framework, decoupling the problem into unmanned aerial vehicles (UAVs) deployment and task scheduling two sequential problems. Specifically, a hierarchical optimization framework is proposed with the goal of optimizing system latency and energy consumption. In network deployment stage, a semi-supervised K-Means based UAV deployment algorithm (SKm-UD) is designed, while an improved multi-agent deterministic policy gradient (MTD 3 PG) strategy is proposed in the distributed task offloading phase, which integrates dual delay network, a lightweight local policy update mechanism, and an adaptive learning rate adjustment strategy based on zebra optimization (ZOA-L) to support efficient computation offloading decisions in partially observable environments. Simulation results show that our proposed method can quickly converge to the optimal strategy, with network energy consumption decreases by at most 30.1%. In addition, the scalability of the mechanism in large-scale Agri-AIoT scenarios is also validated.

In recent years, with the acceleration process of smart cities, the Artificial Intelligence of Things (AIoT) has become a novel approach for emergency rescue. However, due to small capacity and heterogeneity, AIoT nodes are facing various energy limitations and complex routing challenges. Therefore, emergency AIoT often have issues, such as data transmission failures, energy consumption, reduced network lifespan, and latency. To overcome these problems, this study proposes an energy efficiency emergency air–ground integrated heterogeneous AIoT (E3-HetAIoT) structure to improve the energy management efficiency of smart cities. First, we adopt a cell-based clustering mechanism, and the multiobjective zebra optimization algorithm (M-ZOA) is used to select cluster heads (CHs) and super nodes (UNs). The UN uses an urgency level-driven energy efficient sleep scheduling (U-ESS) mechanism to balance the remaining energy of sensors, especially in scheduling sleep time slots for sensors have energy below typical threshold. Second, an air–ground integrated data transmission mechanism is adopted, in which the generator de bits pseudo aleatorios (GBPA) is used to eliminate redundant data in CH and improve security. Then, the data packets are divided into normal packets and emergency ones. The normal packets wait for transmission of unmanned aerial vehicle (UAV) UAVs as intelligent mobile agents (U-iAgents), the trajectory of U-iAgents are dynamically predicted by dueling double deep Q-Network (Dueling-DDQN), meanwhile emergency data packets are immediately transmitted through inter cluster routing. Simulation results demonstrate that compared to existing algorithms, our proposed E3-HetAIoT framework achieves lower energy consumption and higher network lifetime, respectively.

Emergency Communication Network (ECN) is improving network quality of service (QoS) performance via numerous resource allocation and management technologies, such as network slicing, in order to meet the requirements of heterogeneous users in various types of emergency events. Unmanned aerial vehicles (UAVs) serving as NodeB, a.k.a., UxNB, can assist ground base stations (GBSs) to extend the coverage range and network utility of Heterogeneous ECN (HetECN), but make the resource allocation issues for different slice demands in HetECN more complex. This paper investigates the dynamic resource allocation problem of HetECN with a goal of maximizing traffic efficiency while concurrently guaranteeing the transmission rate and the latency by adopting network slicing. Firstly, in order to model the dynamic and uncertain environment of HetECN, we describe the long-term resource allocation problem as a stochastic game, which is an extension of game theory in Markov decision process-like environment. Subsequently, we develop an independent Q-learning based multi-agent reinforcement learning (IQ-MARL) framework, for which all agents execute decision algorithm independently but share a common structure. Simulation results demonstrate that our proposed IQ-MARL algorithm achieves a good balance between performance gains and information exchange overheads in HetECN, which is superior to those of other benchmark schemes.

Heterogeneous Internet of Things (HetIoT) has received widespread attention due to its provision of various convenient services in fields such as smart cities, intelligent transportation, environmental monitoring and security systems. Due to HetIoT inherently demands high data rates, bandwidth, and low latency, the application of millimeter-wave (mmWave) unmanned aerial vehicle (UAV) as emergency aerial base station (ABS) providing services to HetIoT users has become a low-cost and efficient means. However, due to the sensitivity of mmWave to obstacles and jamming, guaranteeing network performance has become a pressing issue. This paper considers a mmWave UAV-assisted HetIoT under jamming conditions, where auxiliary ABSs serve multiple ground users (GUs) who generate a large amount of sensor data. We establish a coverage maximization problem under the constraints of signal-to-interference ratio (SIR) threshold, maximum power of ABSs and maximum number of GUs that the base station can serve, and propose a novel ABS hovering deployment algorithm M-HiAPSO that combines the artificial potential field (APF) method and the improved particle swarm optimization (PSO) algorithm in a hierarchical manner. Specifically, the multi-element APF method is used to characterize the horizontal force between nodes, combined with the improved hierarchical adaptive PSO algorithm to adjust the horizontal position of the ABS to obtain the optimal UAV hovering position and power allocation strategy. Numerical results show that the coverage rate reached 96.2% when the number of iterations was 283, and it can reach up to 99.6%.

Emergency communication network (ECN) can provide fast, efficient and high-capacity communication services for specific areas by using mmWave transmission and unmanned aerial vehicles (UAVs) serving as aerial base stations (ABSs) or relay nodes. Now, in order to satisfy diverse demands, ECN should support different types of nodes, access methods, and traffic distributions, which is referred to as heterogeneous ECN (HetECN). Therefore, inappropriate trajectory planning and unbalanced traffic loading can lead to UAV flight collisions and network congestion. In this article, we jointly optimize UAV jamming avoidance trajectory and user association strategy aimed to load balancing, to maximize the utilization of HetECN. Specifically, an improved artificial potential field (APF) method along with mmWave beam forming technology is used to obtain the jamming avoidance trajectory of UAVs, and the optimal deployment location of UAVs are determined based on the distribution of ground users (GUs). Subsequently, the matching game and alliance game are comprehensively used to determine the load balancing based GU-UAV associated strategy under various GU demands, thereby ensuring traffic load balancing and resource optimization allocation. In addition, altitude fine-tuning have been made to further power consumption, thereby improving overall network efficiency. Simulation results demonstrate that the proposed method can achieve the expected performance in network utilities such as coverage rate, network capacity, load balancing effect of mmWave UAV-assisted HetECNs.