The pinching-antenna systems (PASS), which activate small dielectric particles along a dielectric waveguide, has recently emerged as a promising paradigm for flexible antenna deployment in next-generation wireless communication networks. While most existing studies assume rectangular indoor layouts with full coverage waveguide, practical deployments may involve geometric constraints, partial coverage, and non-negligible waveguide attenuation. This paper presents the first analytical investigation of PASS in a circular indoor environment, encompassing both full coverage and partial coverage waveguide configurations with/without propagation loss. A unified geometric– propagation framework is developed that jointly captures pinching-antenna placement, Internet of Things (IoT) device location distribution, and waveguide attenuation. Closed-form expressions for the outage probability and average achievable rate are derived for four scenarios, with accuracy validated via extensive Monte-Carlo simulations. The analysis reveals that, under the partial coverage waveguide scenario with propagation loss, the system performance demonstrates a non-monotonic trend with respect to the waveguide length, and the optimal length decreases as the attenuation coefficient increases. Numerical results further quantify the interplay between deployment strategy, waveguide propagation loss, and coverage geometry, offering practical guidelines for performance-oriented PASS design.

Conventional collaborative machine learning (CML) faces significant challenges in resource-constrained environments, such as emergency scenarios with limited power, bandwidth, and computing resources, leading to increased communication delays and energy consumption. To address these issues, this paper introducesAir-CoKD, a novel CML framework designed to reduce resource consumption and training latency while preserving model performance.Air-CoKDleverages knowledge distillation (KD) to minimize data transmission by avoiding the direct sharing of model parameters. It also integrates over-the-air computation (AirComp) to aggregate local logits, optimizing bandwidth utilization. To address the dimensional differences in local logits caused by the unbalanced device data class,Air-CoKDemploys orthogonal frequency division multiplexing (OFDM) to transmitting local logits for different target classes. To handle aggregation errors introduced by AirComp, we conduct a detailed analysis of error bounds. Specifically, we convert the Kullback-Leibler (KL) divergence, used in KD loss function, into a quadratic upper bound for precise error quantification and effective optimization. Based on these insights, we propose a strategy to manage bandwidth constraints, transmission power limits, and device energy budgets withinAir-CoKD. Extensive simulations demonstrate thatAir-CoKDsurpasses state-of-the-art methods, effectively balancing training efficiency and model performance. The framework proves to be a robust solution for CML in resource-constrained networks.

Intelligent reflecting surface (IRS) can significantly enhance the performance of data and energy integrated networks (DEIN) by adjusting its amplitude and/or phase. However, there is a lack of comprehensive performance analysis model for realistic DEIN where multiple cells exist rather than only one cell as assumed by most existing work. In this paper, we consider an IRS-assisted multi-cell DEIN. Specifically, in the downlink wireless energy transfer (WET) stage, the hybrid access point (HAP) in each cell broadcasts radio frequency (RF) energy signals to edge user equipments (UEs). Subsequently, during the uplink wireless information transfer (WIT) stage, the edge UEs employ the harvested energy to send their information to the HAP. We first represent the statistical characteristics of the signal-to-interference-plus-noise ratio (SINR) at the edge UE. Then, we derive the closed-form expressions for outage probability, ergodic rate and average symbol error probability of the edge UE in the typical cell. To gain more insights, we obtain the minimum required number of reflection elements and a sub-optimal solution for time allocation coefficients. Finally, extensive numerical results are provided to validate the correctness of the theoretical results.

Reconfigurable intelligent surface (RIS) is a promising technology to improve the spectral and energy efficiency of Internet of Things (IoT) systems. In this paper, we investigate an RIS-assisted simultaneous wireless information and power transmission (SWIPT) system by utilizing stochastic geometry. Moreover, we consider not only the case of random phase shift, but also the case where the phase shift of the RIS are aligned to thek-th IoT device. We first derive the closed-form expressions of the uplink outage probability and the average uplink data size for thek-th IoT device under the Rayleigh channel. Then, we extend the performance analysis to the Rician fading channel and multi-antenna scenarios. Finally, extensive numerical results have been carried out to verify the effectiveness of our derived results.

Backscatter communication (BackCom) is an intriguing technology that enables devices to transmit information by reflecting environmental radio frequency signals while consuming ultra-low energy. Applying BackCom in the Internet of things (IoT) networks can effectively address the power-unsustainability issue of energy-constraint devices. Considering many practical IoT applications, networks are finite-scale and devices are needed to be deployed at hotspot regions organized in clusters to cooperate for specific tasks. This paper considers finite-scale clustered backscatter communication networks (F-CBackCom Nets). To ensure communications, this paper establishes a theoretic model to analyze the communication connectivity of F-CBackCom Nets. Different from prior studies analyzing the connectivity with a focus on the transmission pair located at the center of the network, this paper analyzes the connectivity of a transmission pair located in an arbitrary location, because the performance of transmission pairs potentially varies with their network location. Extensive simulations validate the accuracy of our analytical model. Our results show that the connectivity of a transmission pair can be affected by its network location. Our analytical model and results can offer beneficial implications for constructing F-CBackCom Nets.

Wearable-sensor-based Human Activity Recognition (HAR) has long been a hot topic in ubiquitous computing, which is benefit by the success of deep learning algorithms. The critical difficulties in multimodal sensing environments are how to represent the spatial-temporal dependencies while concurrently extracting features with high characterization. In this work, we propose a self-attention based deep dilated convolution network. Our method uses two channels, named temporal channel and spatial channel, respectively, to extract the readings-over-time and time-over-readings features from sensor signals. The self-attention mechanism helps directly capture the long time dependence of sensor signals. To extract local features while expanding the receptive field and avoiding information loss caused by pooling and upsampling, we use deep dilated convolution, which expanding the receptive field and avoiding information loss caused by pooling and upsampling. Extensive experiments on a self-built dataset and two available benchmark datasets (PAMAP2, OPPORTUNITY) reveal that the effectiveness of our proposed model is more competitive than the state-of-the-art methods in HAR tasks.

The unmanned aerial vehicles (UAVs) have been widely used in civilian environments, due to its high flexibility, low cost and ease of deployment. In this paper, an UAV-enabled mobile edge computing (MEC) system is studied, in which the UAV serves as an aerial mobile base station to provide services for a group of ground user equipments (UEs) with computation task requests. We jointly optimize the time allocation, resource allocation and the UAV flying trajectory to minimize the time required for the UAV to complete the task, subject to the constraints of different kinds of resources, energy and velocity. Due to the non-convexity of the formulated problem, we first transform it to a feasibility check problem and then divide it onto three convex optimization subproblems. By using the block coordinate descent method and the successive convex approximate (SCA) method, we propose an efficient iterative algorithm to solve the three subproblems alternately with ensured convergence. Extensive simulation results show that the proposed joint optimization algorithm reduces the task completion time compared with other schemes.

In this article, we investigate multiple unmanned aerial vehicles (UAVs)-enabled data collection system in Internet of Things (IoT) networks with time windows, where multiple rotary-wing UAVs are dispatched to collect data from time-constrained terrestrial IoT devices. We aim to jointly minimize the number and the total operation time of UAVs by optimizing the UAV trajectory and hovering location. To this end, an optimization problem is formulated, considering the energy budget and cache capacity of UAVs as well as the data transmission constraint of IoT devices. To tackle this mix-integer nonconvex problem, we decompose the problem into two subproblems: 1) UAV trajectory and 2) hovering location optimization problems. To solve the first subproblem, an modified ant colony optimization (MACO) algorithm is proposed. For the second subproblem, the successive convex approximation (SCA) technique is applied. Then, an overall algorithm, termed the MACO-based algorithm, is given by leveraging the MACO algorithm and SCA technique. Simulation results demonstrate the superiority of the proposed algorithm.

This article presents an analytical model to analyze the communication performance of cluster backscatter communication networks (CBackCom Nets) by considering their unique interferences. In CBackCom Nets, interferences are from both backscatter transmitters (BTs) and carrier emitters (CEs), i.e., RF signal emitters. Because BTs are distributed in clusters around CEs, interferences from BTs and interferences from CEs constitute coupled interferences. In addition, since BTs conduct backscatter communications by reflecting RF signals from CEs, interfering signals from BTs, and interfering signals from CEs are power-correlated, leading to the particularity and complexity of coupled interferences of CBackCom Nets. In contrast to previous studies that analyze the performance of CBackCom Nets ignoring coupled interferences, this article develops a novel interference analysis approach to analyze their coupled interferences, and then analyze performance, including coverage probability and spatial throughput of a cluster. Our numerical results show that our analytical model can obtain more accurate results than prior analytical models. In addition, our results reveal the relationship between the communication performance and multiple factors, such as the node density, the energy harvesting model, the interferences from CEs, and the cluster size, offering insightful implications for constructing and configuring CBackCom Nets.

This paper considers a multi-drone enabled data collection system for smart cities, where there are two kinds of drones, i.e., Low Altitude Platforms (LAPs) and a High Altitude Platform (HAP). In the proposed system, the LAPs perform data collection tasks for smart cities and the solar-powered HAP provides energy to the LAPs using wireless laser beams. We aim to minimize the total laser charging energy of the HAP, by jointly optimizing the LAPs' trajectory and the laser charging duration for each LAP, subject to the energy capacity constraints of the LAPs. This problem is formulated as a mixed-integer and non-convex Drones Traveling Problem (DTP), which is a combinatorial optimization problem and NP-hard. We propose an efficient and novel search algorithm named Drones Traveling Algorithm (DTA) to obtain a near-optimal solution. Simulation results show that DTA can deal with the large-scale DTP (i.e., more than 400 data collection points) efficiently. Moreover, the DTA only uses 5 iterations to obtain the near-optimal solution whereas the normal Genetic Algorithm needs nearly 10000 iterations and still fails to obtain an acceptable solution.

In this paper, multi-unmanned aerial vehicle (UAV) enabled mobile edge computing (MEC), i.e., UAVE is studied, where several UAVs are deployed as flying MEC platform to provide computing resource to ground user equipments (UEs). Compared to the traditional fixed location MEC, UAV enabled MEC (i.e., UAVE) is particular useful in case of temporary events, emergency situations and on-demand services, due to its high flexibility, low cost and easy deployment features. However, operation of UAVE faces several challenges, two of which are how to achieve both 1) the association between multiple UEs and UAVs and 2) the resource allocation from UAVs to UEs, while minimizing the energy consumption for all the UEs. To address this, we formulate the above problem into a mixed integer nonlinear programming (MINLP), which is difficult to be solved in general, especially in the large-scale scenario. We then propose a Reinforcement Learning (RL)-based user Association and resource Allocation (RLAA) algorithm to tackle this problem efficiently and effectively. Numerical results show that the proposed RLAA can achieve the optimal performance with comparison to the exhaustive search in small scale, and have considerable performance gain over other typical algorithms in large-scale cases.

In the past few years, person re-identification (reID) has developed rapidly due to the success of deep convolutional neural networks. The softmax loss function is an important component for learning discriminative features. However, the classifier trained by the softmax loss is difficult to distinguish the hard samples. In this work, we introduce a new auxiliary loss function, called batch-contrastive loss, for person reID to further separate the features of different identities and pulls the features of same identity closer. Furthermore, the proposed loss function does not rely on the pairwise or triplet sampling which is commonly used in the Siamese model. We test our loss function on two large-scale person reID benchmarks, Market-1501 and DukeMTMC datasets. Under the combination of the batch-contrastive loss and the softmax loss, even only employing the generic L2-distance metric, we can achieve competitive results among the state-of-the-arts.

Existing deep models using attributes usually take global features for identity classification and attribute recognition. However, some attributes exist in local position, such as a hat and shoes, therefore global feature alone is insufficient for person representation. In this work, we propose to use the attribute recognition as an auxiliary task for person re-identification. The attributes are recognised from the local regions of mid-level layers. Besides, we extract local features and global features from a high-level layer for identity classification. The mid-level attribute learning improves the discrimination of high-level features, and the local feature is complementary to the global feature. We report competitive results on two large-scale person re-identification benchmarks, Market-1501 and DukeMTMC-reID datasets, which demonstrate the effectiveness of the proposed method.

This paper considers an unmanned aerial vehicle based mobile edge computing (UAV based MEC) system, where we assume there is one UAV, acts as an edge cloud, providing data processing services to the Internet of things devices (IoTDs). We consider the UAV hovers at difference places for different time to receive and process data for IoTDs. We aim to minimize the energy consumption of the UAV, including its hovering energy and computation energy, by optimizing the hovering time, scheduling and resource allocation of the tasks received from IoTDs, subject to the quality of service (QoS) requirement of all the IoTDs and the computing resource available at UAV. This is formulated as a mixed-integer non-convex optimization problem, which is difficult to solve in general. We propose an efficient iterative algorithm to get a high-quality suboptimal solution. Simulation results show that our proposed method has a very good performance compared with the other benchmarks.

Machine-to-machine (M2M) type communications (MTCs) over cellular networks feature the large number of MTC devices (MTCDs), small and time controlled data transmissions, and rigorous energy limitation. Considering full-duplex (FD) relaying can achieve high spectrum and energy efficiency, this paper proposes an MTC-enabled cellular communication scheme, where a traditional cellular user equipment (UE) is configured as an FD relaying based gateway to assist the uplink transmissions of the served MTCDs. The designed objective is to minimize the aggregate energy consumption of a group consisting of a UE and multiple MTCDs, while fulfilling their minimum throughput requirements. To this end, a convex optimization problem is formulated and a low complexity algorithm is also developed to find the optimal power allocation strategies for the UE and the MTCDs. The simulation results show that the proposed scheme can achieve most of the channel reuse gain of the FD relaying if the self-interference at the UE is controlled below a certain level.

Sequential labeling addresses the classification of sequential data, which are widespread in fields as diverse as computer vision, finance, and genomics. The model traditionally used for sequential labeling is the hidden Markov model (HMM), where the sequence of class labels to be predicted is encoded as a Markov chain. In recent years, HMMs have benefited from minimum-loss training approaches, such as the structural support vector machine (SSVM), which, in many cases, has reported higher classification accuracy. However, the loss functions available for training are restricted to decomposable cases, such as the 0-1 loss and the Hamming loss. In many practical cases, other loss functions, such as those based on the $F_{1}$ measure, the precision/recall break-even point, and the average precision (AP), can describe desirable performance more effectively. For this reason, in this paper, we propose a training algorithm for SSVM that can minimize any loss based on the classification contingency table, and we present a training algorithm that minimizes an AP loss. Experimental results over a set of diverse and challenging data sets (TUM Kitchen, CMU Multimodal Activity, and Ozone Level Detection) show that the proposed training algorithms achieve significant improvements of the $F_{1}$ measure and AP compared with the conventional SSVM, and their performance is in line with or above that of other state-of-the-art sequential labeling approaches.

An ability to predict collisions is essential for current vehicles and autonomous robots. In this paper, an integrated collision predication system is proposed based on neural subsystems inspired from Lobula giant movement detector (LGMD) and directional selective neurons (DSNs) which focus on different part of the visual field separately. The two type of neurons found in the visual pathways of insects respond most strongly to moving objects with preferred motion patterns, i.e., the LGMD prefers looming stimuli and DSNs prefer specific lateral movements. We fuse the extracted information by each type of neurons to make final decision. By dividing the whole field of view into four regions for each subsystem to process, the proposed approaches can detect hazardous situations that had been difficult for single subsystem only. Our experiments show that the integrated system works in most of the hazardous scenarios.

Because of random deployment patterns of femtocells, interference scenarios in a heterogeneous cellular network can be very complicated because of its changing network topology. Especially when each eNodeB occupies a fixed bandwidth, interference management becomes much more difficult. The benefit of dynamic management for local resource optimation is limited. Recently, resource virtualization has been proposed as a dynamic resource management scheme to optimize network performance. In fact, resource virtualization is viewed as a more flexible model, in which mobile network service providers can control physical resources in a global scope. This paper presents a joint resource virtualization and allocation scheme for its applications in heterogeneous macro-femto-cellular networks. The proposed scheme involves two major processes. First, it virtualizes physical resources as logical resources. Second, it carries out logical resource allocation optimization globally and aggregates logical and physical resources for resource allocation. The proposed scheme takes into account spectrum reuse and frequency domain interference jointly in order to achieve a high spectral efficiency and provide rate-on-demand services to all users. Simulation results demonstrate the effectiveness of the proposed scheme. Copyright © 2016 John Wiley & Sons, Ltd.

Since wireless network virtualization (WNV) enables physical resources abstraction and sharing, the overall resources inefficiency can be reduced dramatically. This paper investigates a pricing-based energy efficient (EE) optimization problem for orthogonal frequency-division multiple Access (OFDMA) WNV. A typical WNV environment consists of an infrastructure provider (InP), virtual network operators (VNOs) and end users. The objective of this paper is to maximize VNOs' EE in bits per joule unit. This is achieved by allocating each VNO certain amount of power. The problem is formulated as a commercial market competition based on a pricing function. A non-cooperative game is applied and a power allocation algorithm is developed to search the Nash equilibrium which is the solution of this game. The Nash equilibrium indicates the best strategy that each VNO can employ. The performances of the proposed algorithm are obtained in a frequency selective fading environment. Evaluation results reveal the VNO adaptation of power sharing strategies and also shows the inefficiency of the Nash equilibrium.

Structural SVM is an extension of the support vector machine for the joint prediction of structured labels from multiple measurements. Following a large margin principle, the training of structural SVM ensures that the ground-truth labeling of each sample receives a score higher than that of any other labeling. However, no specific score ranking is imposed among the other labelings. In this letter, we extend the standard constraint set of structural SVM with constraints between “almost-correct” labelings and less desirable ones to obtain a partial-ranking structural SVM (PR-SSVM) approach. Experimental results on action segmentation and classification with two challenging datasets (the TUM Kitchen mocap dataset and the CMU-MMAC video dataset) show that the proposed method achieves better detection and false alarm rates and higher F1 scores than both the conventional structural SVM and a comparable unstructured predictor. The proposed method also achieves higher accuracy than the state of the art on these datasets in excess of 14 and 31 percentage points, respectively.

Sequential labeling addresses the classification of sequential data and is of increasing importance for the classification and segmentation of video data. The model traditionally used for sequential labeling is the hidden Markov model where the sequence of class labels to be predicted is encoded as a Markov chain. In recent years, hidden Markov models and other structural models have benefited from minimum-loss training approaches which in many cases lead to greater classification accuracy. However, the loss functions available for training are restricted to decomposable cases such as the zero-one loss and the Hamming loss. Other useful losses such as the F1loss, equal error rates and others are not available for sequential labeling. For this reason, in this paper we propose a training algorithm that can cater for the F1loss and any other loss function based on the contingency table. Experimental results over the challenging TUM Kitchen Dataset depicting human actions in a kitchen scenario show that the proposed training approach leads to significant improvement of different performance metrics such as the classification accuracy (4.3 percentage points) and the F1measure (8.9 percentage points).

This paper considers the network situation where the primary users (PUs) in a cognitive radio network have leased out the idled spectrum to the secondary users (SUs) via pricing-based dynamic spectrum allocations (DSAs). We take into account that a SU can serve as a cooperative relay for a PU, and, then, stimulate the PU to split more spectrum while maintaining the minimum transmission rate of the PU. Without using pricing-based mechanisms again, a resource-exchange based bargaining game is proposed to develop the incentive mechanism. Considering multiple SUs should compete with each other for the newly-obtained spectrum from a PU, the novelty of the game scheme is in taking explicitly account of that each PU and SU have their own minimum rate demands. Simulation results show the game guarantees the minimum rate requirement for the PU, and, at the same time, ensures each SU can get a fair rate-reward from the PU according to the level of contribution that it can make to compensate the PU's rate-loss.

In this paper, a multi-user cooperative game is proposed to stimulate selfish user nodes to participate in cooperative relaying in wireless ad hoc networks. Without using the traditional reputation-mechanisms or pricing-mechanisms, we resort to the resource-exchange mechanism by assuming a source node could reward the relaying nodes by, in return, forwarding data that are originated from these relaying nodes. Then the cooperation stimulating problem can be formulated as a multi-player cooperative bargaining game. We prove that there exists a unique Nash bargaining solution (NBS) of the game and propose a fast Particle Swarm Optimizer (PSO) algorithm to solve the NBS. Simulation results show that the NBS-based incentive strategy achieves social optimality, i.e., all cooperative nodes could achieve significant rate-gains in comparison with direct transmission. Moreover, the relaying nodes could also get fairness rewards by the source node according to the level of contribution that they have made to improve the performance of the source node.

In this paper, a novel modulation strategy based on two dimensional modulation is proposed, which can not only achieve fast balancing DC-link capacitor voltages for each cell in the cascaded H-Bridges Rectifier (CHBR), but can also extend the stable operating range to the maximum with heavy unbalanced loads under unity power factor. The control method has been fully investigated with detailed theoretical analysis, the performance of the proposed modulation strategy is evaluated with two H-bridges in cascade using the simulation results obtained from the Matlab/simulink. It is shown that the dynamics of the system for balancing the DC link capacitor voltages of each cell is considerably improved with very fast response time even at very heavy unbalanced load, and in whole the process, the input of the CHBR maintains in unity power factor.

In this paper, a cooperative game is proposed to perform a fair and efficient resource allocation for the time division multiple access (TDMA) based cooperative communication networks. In the considered system, two selfish user nodes can act as a source as well as a potential relay for each other. A transmission node with energy limitation is willing to seek cooperative relaying only if the data-rate achieved through cooperation is not lower than that achieved without cooperation by consuming the same amount of energy. The cooperative strategy of a node can be defined as the number of data-symbols and power that it is willing to contribute for relaying purpose. We formulate this two-node fair and efficient resource sharing problem as a bargaining game. Since the Nash bargaining solution (NBS) to the game is computationally complex to obtain, a low-complexity algorithm to search the suboptimal NBS is proposed. Simulation results show that the NBS results are fair in that both nodes could experience better performance than if they work independently. And the NBS results are efficient in that the performance loss of the game to that of the maximal overall rate scheme is small while the maximal-rate scheme is unfair.

A pricing-based non-cooperative game is proposed to stimulate cooperation and perform spectrum allocation in multi-relay cooperative transmission networks. The authors construct a buyers' market competition model to consider that multiple relays are willing to share their spectrum resources with a single user. Both the benefits of the relays and the user are concerned in the game. First, according to the current user's demand, the relays as sellers compete with each other to determine the price of relaying that can maximise their profits. Then to maximise its utility, the user purchases the optimal amount of spectrum resources from each relay. The existence of the Nash equilibrium (NE), that is, the solution of the game, is proved. Even though the NE can be obtained in a centralised manner, a distributed algorithm to search for the NE is developed, which is more applicable in practical systems. Also, the convergence conditions of the algorithm are also analysed. Furthermore, the authors have also proved that the NE is not efficient when considering the total relays' profits. Thus, a general method to find the global optimal solution that maximises the total relays' profits is given. Simulation results show, by using the game, that a reasonable spectrum allocation can be performed between the relays and the user.

Abstract In this paper, we consider the problem of stimulating cooperation and resource allocation in cooperative transmission networks. We formulate this problem as a sellers' market competition where a relay is willing to share its resource with multiple users. We use a Stackelberg game to jointly consider the benefits of the relay and the users. Firstly, the relay determines the price of relaying according to the user demand. Secondly, the users purchase the optimal amount of resources to maximize their utilities. Although the Nash equilibrium, i.e., the solution of the game, can be obtained in a centralized manner, we develop a distributed algorithm to search the Nash equilibrium, which is more applicable in practical systems. Also, the convergence conditions of the algorithm are analyzed. Simulation results show, by using the distributed algorithm, the relay and the users could determine what price should ask for and how much bandwidth should buy, respectively. Copyright © 2010 John Wiley & Sons, Ltd.