Leveraging flexible deployment, extensive coverage, and reliable communication links, unmanned aerial vehicle (UAV)-assisted mobile edge computing offers new opportunities to support mobile devices with heavy computational tasks. Considering the energy constraint of UAVs and delay requirement of tasks, many efforts should be devoted to pursuing lower service latency, which however is under explored in this innovational architecture. In this paper, with the purpose of minimizing the overall network service duration, departing from traditional serial task execution, we first design a multi-task parallel execution paradigm, and then investigate a joint optimization problem encompassing content caching, computation offloading, and channel allocation. To address this intractable problem involving large state and action spaces, we decompose it into two subproblems, i.e., an intra content caching and computation offloading optimization of each UAV, and an inter channel allocation of all UAVs. We then propose a reinforcement learning-based two-layer optimization scheme that integrates the efficient representation of DQN learning and the comprehensive exploration of regret minimization learning. Specifically, in the lower layer, a DQN-based algorithm is developed to solve the intra subproblem, and in the upper layer, a regret minimization-based algorithm is designed to tackle the inter subproblem. Through nested optimization between the two layers, optimal strategies for content caching, computation offloading and channel allocation can be achieved. Numerical results demonstrate that the proposed scheme significantly reduces service latency compared to various baseline methods.

Motion planning, considered a fuzzy sequential decision-making problem, encounters significant challenges due to inherent environmental uncertainty. Traditional planning methods that rely on single strategies often struggle in complex scenarios. While fuzzy systems excel at handling uncertainty, high-dimensional continuous spaces require a large number of fuzzy rules, which significantly increases computational complexity. In contrast, humans leverage limited and fuzzy information to address various decision-making scenarios flexibly and efficiently. The concurrent reasoning mechanism in the prefrontal cortex plays a crucial role during this process. Consequently, the brain-inspired model and the concept of multiple fuzzy rules offer a novel perspective for the above issues. Motivated by these insights, this article proposes a brain-inspired motion planning method called harmonized learning with concurrent arbitration (HLCA). Specifically, inspired by the concurrent inference model, a concurrent arbitration module is employed in the planning process to effectively manage the boundary between exploration and exploitation. Furthermore, inspired by the multistrategy processing mechanism, HLCA introduces multistrategy harmonized learning by referring to the mechanism for operating multiple fuzzy rules, allowing the dynamic selection of strategies through a reliability function to enable self-improving learning. Experimental results demonstrate that HLCA outperforms state-of-the-art benchmarks, highlighting its potential to enhance the planning performance of robots by learning from the human brain.

Cross-sensory interaction is a key aspect for multisensory recognition. Without cross-sensory interaction, artificial neural networks show inferior performance in multisensory recognition. On the contrary, the human brain has an inherently remarkable ability in multisensory recognition, which stems from the diverse neurons that exhibit distinct responses to sensory inputs, especially the multisensory neurons with multisensory responses hence enabling cross-sensory interaction. Based on this neuronal diversity, we propose a Neuronal Diversity inspired Multisensory Recognition Model (ND-MRM), which, similar to the brain, comprises unisensory neurons and multisensory neurons. To reflect the different responses characteristics of diverse neurons in the brain, special connection constraints are innovatively designed to regulate the features transmission in the ND-MRM. Leveraging this novel concept of neuronal diversity, our model is biologically plausible, enabling more effective recognition of multisensory information. To validate the performance of the proposed ND-MRM, we employ a multisensory emotion recognition task as a case study. The results demonstrate that our model surpasses state-of-the-art brain-inspired baselines on two datasets, proving the potential of brain-inspired methods for advancing multisensory interaction and recognition.

With significantly dynamic characteristics of the new aerial users, the emerging cellular-enabled unmanned aerial vehicle (UAV) communication paradigm raises great challenges to current research of UAV applications. As far as the robust channel allocation is concerned, the high mobility of UAV nodes and the unexpected disturbance of external environment would render most existing methods which rely on definite information and are vulnerable to dynamic environment, become less attractive or even invalid. In this paper, we particularly investigate a cellular-enabled mesh UAV network exploiting partially overlapping channels (POCs), and propose a distributed fuzzy space based learning scheme for POCs allocation to combat the dynamic environment. Rather than the perfect channel state information (CSI) assumption, the dynamic and uncertain CSI of UAVs is characterized by fuzzy number. On this basis, the allocation process can be implemented in a mapped fuzzy space. Integrating fuzzy-logic and game based learning, we formulate the problem of POCs assignment as a fuzzy payoffs game (FPG), and demonstrate the existence of fuzzy Nash equilibrium for our designed FPG. Then, with the derived priority vector in the fuzzy space, the equilibrium solution can be achieved by the proposed algorithm. Numerical simulations demonstrate the advantages of our new scheme.

To support large-scale unmanned aerial vehicle (UAV) networks with both payload communication (PC) packets and control and non-payload communication (CNPC) packets, millimeter wave (mmWave) communication is convinced to be a promising solution. However, efficient user association and spectrum allocation are still challenging in mmWave-enabled UAV networks considering that the network state is dynamic and the status information of each UAV is incomplete. In this paper, we investigate a joint UAV association and spectrum allocation problem under a hybrid mmWave sharing paradigm, where PC packets are transmitted over both licensed and pooled bands in a shared manner to achieve high throughput and CNPC packets are transmitted over licensed band in an exclusive manner to guarantee high reliability. To this end, we introduce a strategic form game that can characterize network stochastic states and individual partial observations to reformulate the problem, and then propose a counterfactual regret minimization scheme to achieve its correlated equilibrium (CE). Benefited from the updating mechanism on randomobservation-actionpairs, the designed scheme can converge to the corresponding CE solutions for the two types of packets with partial observation. Finally, our simulation results demonstrate the superior performance of the proposed scheme over the baseline schemes.

Emerging cellular‐enabled unmanned aerial vehicles (UAVs) communication poses both opportunities and challenges to the current research of UAV applications. On the one hand, the advanced cellular technologies and authentication mechanisms make significant performance improvements of reliability, security, coverage, and throughput of UAVs possible. On the other hand, the considerably dynamic characteristics of the new aerial users bring some complicated and open issues to the future heterogeneous network. To throw some light on this field, the authors study the joint power allocation and channel reuse problem of uplink transmission in a cellular‐enabled UAVs network with full consideration of the rapid channel variations resulting from high mobility of UAVs. Given the diverse requirements of terrestrial cellular users (CUs) and UAVs, i.e. large capacity for CU links and ultra‐reliability for UAV links, they formulate the problem as optimising the uplink capacities of CUs with reliable transmission constraint of UAVs. By decoupling the intricate problem into two sub‐problems, a joint resource allocation algorithm is proposed, which relies only on the statistical information of dynamic channels to implement, hence is robust. Finally, numerical results are provided to corroborate the anticipated performances of the proposed scheme.

Despite the great promises in next-generation wireless communications, dynamic spectrum access (DSA) remains still as a major challenge in uncertain environments, e.g., varying unknown channels. Existing popular schemes, i.e. potential game approaches, rely on the definite reward and a greedy strategy, which become unfortunately invalid in such varying and uncertain scenarios. In this paper, we propose a robust fuzzy-game approach to combat inherent channel uncertainties. Rather than the definite reward, we first project the decision space to another fuzzy-logical space and thereby characterize the varying uncertain information with the fuzzy numbers. Thus, the sensitiveness to random fluctuation in definite rewards would be alleviated. On this basis, we formulate DSA in uncertain channels as a centralized fuzzy payoffs game (FPG). We then develop a novel fuzzy-learning algorithm to achieve the optimal network throughput even in face of uncertain information, with which the network controller executes the decision making of sharing users by exploiting the fuzzy-logic method. Numerical results are finally provided to validate our new FPG scheme. Although uncertain environments render existing crisp-game approaches invalid, our new algorithm can converge after tens of iterations (even in fast-varying conditions), thereby permitting reliable shared accessing and improved throughput, which is of great significance to next- generation communications operated in dynamical and uncertain environments.

We investigate the global throughput maximization of distributed spectrum reusing (DSR) in cognitive radio (CR) network, which reaches the two-dimensional spectrum multiplexing. Most previous works only consider the temporal-domain accessing, which greatly underutilize the spectrum resources. In this paper, we propose a new temporal-spatial spectrum reusing scheme by fully exploiting the location information of devices, where multiple users can access one channel simultaneously. In distributed applications, the global information will be unavailable, and therefore a non-cooperative game is formulated. It is proved as an exact potential game (EPG), which has at least one pure strategy Nash equilibrium (NE). Then, an improved decentralized reinforcement learning (RL) algorithm is developed to achieve the NE points. The network performance is evaluated by computer simulations.