Accurate short-time traffic flow prediction has gained gradually increasing importance for plan and management with the deployment of intelligent transportation systems (ITSs). However, existing approaches short-term are unable to efficiently capture complex nonlinearity flow, which provide unsatisfactory accuracy. In this paper, we propose a deep learning based model uses hybrid multiple-layer architectures automatically extract inherent features data. Firstly, built on convolutional neural...

Evidence indicates that downloading on-demand videos accounts for a dramatic increase in data traffic over cellular networks. Caching popular the storage of small-cell base stations (SBS), namely, caching, is an efficient technology reducing transmission latency while mitigating redundant transmissions back-haul channels. In this paper, we consider commercialized caching system consisting network service provider (NSP), several video retailers (VRs), and mobile users (MUs). The NSP leases...

Heterogeneous cellular networks (HCNs) with embedded small cells are considered, where multiple mobile users wish to download network content of different popularity. By caching data into the small-cell base stations, we will design distributed optimization algorithms via belief propagation (BP) for minimizing downloading latency. First, derive delay-minimization objective function and formulate an problem. Then, develop a framework modeling underlying HCN topology aid factor graph....

Small-cell caching utilizes the embedded storage of small-cell base stations (SBSs) to store popular contents for sake reducing duplicated content transmissions in networks and offloading data traffic from macrocell SBSs. In this paper, we study a probabilistic strategy, where each SBS caches subset with specific probability. We consider two kinds network architectures: 1) The SBSs are always active, which is referred as always-on architecture; 2) activated on demand by mobile users (MUs),...

In intelligent transportation systems (ITS), incomplete traffic data due to sensor malfunctions and communication faults, seriously restricts the related applications of ITS. Recovering missing from becomes an important issue for Existing works on imputation cannot achieve satisfactory accuracy inefficiently exploiting underlying topological structure data. this paper, we model topology road network as a graph introduce Fourier transform (GFT) process Then utilize algebraic framework termed...

Due to the rapidly increasing number of deployed base stations (BSs) in current cellular networks, energy consumption has emerged as a great challenge network operation.In this paper, we propose novel data-driven intelligent BS sleeping mechanism based on wireless traffic prediction model that measures BSs' capacity different regions.Firstly, spatial-temporal is proposed, where multi-graph convolutional (MGCN) developed capture spatial features, and multi-channel long short-term memory...

In this paper, we consider the intercell interference coordination (ICIC) problem in heterogeneous cellular networks with randomly deployed small-cell base stations (BSs). Current research on ICIC mainly focuses optimizing spectrum and power allocations at BSs, whereas user-BS association is treated as a separate issue. Nevertheless, an important issue should be jointly optimized resource to achieve global optimality. objective of maximizing system sum rate distributed manner, propose novel...

In heterogeneous networks (HetNets), the load between macrocell base stations (MBSs) and small-cell (SBSs) is imbalanced due to their different transmission powers locations. This imbalance significantly impacts system performance affects experience of mobile users (MUs) with priorities. this paper, we aim distributively optimize user association in HetNets various priorities solve balancing problem. Since a binary matching problem, which NP-hard, propose distributed belief propagation (BP)...

The growing number of complex and heterogeneous Internet Things (IoT) applications has imposed a high demand for scarce communications computing resources. To meet this stringent requirement, it is desirable to develop large-scale highly adaptive online resource allocation strategies streamline existing network operations. Deep reinforcement learning (DRL), which combines the merits deep learning, capable addressing decision-making tasks, thus enabling efficient allocation. In article, we...

Cell-free networks and reconfigurable intelligent surfaces (RIS) are two promising techniques for future wireless communications. The integration of RIS into cell-free networks, termed RIS-aided offers the potential to significantly enhance network performance. However, realization this is constrained by limited capacities fronthaul links. To address challenge, we investigate joint design access point (AP) clustering, transmit passive beamforming in networks. objective maximize weighted...

Abstract The flexible control of electromagnetic (EM) waves across the entire spatial domain is a long‐standing aspiration in metasurface research, driven by its potential to enhance signal coverage and channel capacity. However, most existing full‐space metasurfaces are restricted manipulating incidence within one specific half‐space, failing exploit EM space. This paper introduces novel bidirectional transmission‐reflection‐integrated (BTRIM) for simultaneous independent incident waves. By...

The development of BACE-1 (β-site amyloid precursor protein cleaving enzyme 1) inhibitors is a crucial focus in exploring early treatments for Alzheimer's disease (AD). Recently, graph neural networks (GNNs) have demonstrated significant advantages predicting molecular activity. However, their reliance on structures alone often neglects explicit sequence-level semantic information. To address this limitation, we proposed Graph and multi-level Sequence Fusion Learning (GSFL) model the...

A tremendous amount of visual data are bing collected by the Internet Video Things (IoVT) systems in which ubiquitous cameras deployed cities enable new applications domains smart transportation and public security. However, limited resources terms communication, computing, caching (3C) conventional cellular network make it challenging to adopt centralized artificial intelligence (AI) conduct real-time video-based analytics. In this work, based on 5G architecture with edge servers, a...

The number of leaves in maize seedlings is an essential indicator their growth rate and status. However, manual counting inefficient limits the scope investigation. Deep learning has shown potential for quickly identifying seedlings, but it requires larger, labeled datasets. To address these challenges, we proposed a method from fields using combination semi-supervised learning, deep UAV digital imagery. Our approach leveraged novel methods detecting seedling accurately efficiently....

This letter considers covert communications in the context of unmanned aerial vehicle (UAV) networks, where a UAV is employed as base station to transmit data legitimate ground user, while ensuring that transmission cannot be detected by warden. Aiming at maximizing user's average effective throughput (AECT), UAV's trajectory and power are jointly optimized. Taking advantage deep reinforcement learning (DRL) on solving dynamic unpredictable problems, we develop twin-delayed deterministic...

Integrated Sensing and Communication (ISAC) Systems, which can support simultaneous information transmission target detection, have been regarded as promising solutions for various emerging wireless network applications. In this work, a covert system with the aid of ISAC technique is proposed, where radar intends to sense send message legitimate users covertly against warden's surveillance either perfect or imperfect channel state (CSI) at warden. We formulate beamforming optimization...

Failure to account for the variation of kernel growth in a cereal crop simulation model may cause serious deviations estimates yield. The goal this research was revise GREENLAB-Maize incorporate source- and sink-limited allocation approaches simulate dry matter accumulation individual kernels an ear (GREENLAB-Maize-Kernel). used potential rates characterize sink demand. remobilization non-structural carbohydrates from reserve organs also incorporated. Two years field experiments were...

Federated learning (FL) provides a novel framework to collaboratively train shared model in distribution fashion by virtue of central server. However, FL is inappropriate for serverless scenario and also suffers from some major drawbacks Industrial Internet Things (IIoT) networks, such as unresilience network failures communication bottleneck effect. In this article, we propose decentralized federated (DFL) approach IIoT devices achieve consensus exchanging parameters only with their...

The dynamic exploitation of unlicensed spectrum by mobile operators is becoming a trend future 5G networks, with several efficient solutions being standardized for enabling sharing. Solutions various design philosophies diversify the network architectures and protocols. Among them, LAA aims at physical-layer coexistence LTE WiFi within spectrum, while LWA LWIP focus on aggregating link capacity in licensed one. In this article, comprehensive survey these three sharing technologies are...

We propose a new secure transmission scheme which uses directional modulation (DM) with artificial noise and is aided by the intelligent reflecting surface (IRS). Specifically, direct path IRS-enabled reflect carry same confidential signal thus can be coherently added at desired position to maximize total received power, while signals other positions are distorted. derive closed-form expression for secrecy rate achieved proposed scheme. Using simulation results, we show that achieve...

Small-cell caching utilizes the embedded storage of small-cell base stations (SBS) to cache popular network contents, for purpose reducing duplicate transmissions in mobile networks and offloading data traffic from macro-cell stations. In this paper, we propose a random system, where each SBS randomly caches subset contents with specified probability. We particularly focus on probability that users can successfully download their requested files SBSs, namely, downloading (SDP). A...