A maximum-independent-set-based channel allocation algorithm for multi-channel wireless networks

A channel allocation algorithm based on the maximum independent set is proposed to decrease network conflict and improve network performance. First, a channel allocation model is formulated and a series of the maximum independent sets （MISs） are obtained from a contention graph by the proposed approximation algorithm with low complexity. Then, a weighted contention graph is obtained using the number of contention vertices between two MISs as a weighted value. Links are allocated to channels by the weighted contention graph to minimize conflicts between independent sets. Finally, after channel allocation, each node allocates network interface cards （NICs） to links that are allocated channels according to the queue lengths of NICs. Simulations are conducted to evaluate the proposed algorithm. The results show that the proposed algorithm significantly improves the network throughput and decreases the end to end delay.

基于GhostNet的改进模型轻量化方法

为了降低深度卷积神经网络的部署成本,优化模型的检测性能,提出一种改进的轻量化主干网络算法S-GhostNet.该算法通过引入特征图生成优化的Ghost Module结构降低卷积操作的计算量,并结合改进类残差模块提升模型的精确度.S-GhostNet具有较强的即插即用性,可以应用于多数卷积神经网络模型.实验结果表明:在目标分类以及目标检测任务中,S-GhostNet相较于MobileNetV2、ShuffleNetV2以及GhostNet,模型计算量更小,模型的精确度持平,甚至更高.

Joint channel assignment and cross-layer routing protocol for multi-radio multi-channel Ad Hoc networks

To study multi-radio multi-channel （MR-MC） Ad Hoc networks based on 802.11, an efficient cross-layer routing protocol with the function of joint channel assignment, called joint channel assignment and cross-layer routing （JCACR）, is presented. Firstly, this paper introduces a new concept called channel utilization percentage （CUP）, which is for measuring the contention level of different channels in a node’s neighborhood, and deduces its optimal value for determining whether a channel is overloaded or not. Then, a metric parameter named channel selection metric （CSM） is designed, which actually reffects not only the channel status but also corresponding node’s capacity to seize it. JCACR evaluates channel assignment by CSM, performs a local optimization by assigning each node a channel with the smaller CSM value, and changes the working channel dynamically when the channel is overloaded. Therefore, the network load balancing can be achieved. In addition, simulation shows that, when compared with the protocol of weighted cumulative expected transfer time （WCETT）, the new protocol can improve the network throughput and reduce the end-to-end average delay with fewer overheads.

Joint Link Allocation and Rate Assignment Algorithm for Multi-Channel Wireless Networks

This paper presents a link allocation and rate assignment algorithm for multi-channel wireless networks. The objective is to reduce network con-flicts and guarantee the fairness among links. We first design a new network model. With this net-work model, the multi-channel wireless network is divided into several subnets according to the num-ber of channels. Based on this, we present a link allocation algorithm with time complexity O（l^2）to al-locate all links to subnets. This link allocation algo-rithm adopts conflict matrix to minimize the network contention factor. After all links are allocated to subnets, the rate assignment algorithm to maximize a fairness utility in each subnet is presented. The rate assignment algorithm adopts a near-optirml al-gorithm based on dual decomposition and realizes in a distributed way. Simulation results demonstrate that, compared with IEEE 802.11b and slotted see-ded channel hopping algorithm, our algorithm de-creases network conflicts and improves the net-work throughput significantly.

Performance Analysis of Multi-Channel CR Enabled IoT Network with Better Energy Harvesting

Wireless Sensor Networks(WSNs)can be termed as an autoconfigured and infrastructure-less wireless networks to monitor physical or environmental conditions,such as temperature,sound,vibration,pressure and motion etc.WSNs may comprise thousands of Internet of Things(IoT)devices to sense and collect data from its surrounding,process the data and take an automated and mechanized decision.On the other side the proliferation of these devices will soon cause radio spectrum shortage.So,to facilitate these networks,we integrate Cognitive Radio(CR)functionality in these networks.CR can sense the unutilized spectrum of licensed users and then use these empty bands when required.In order to keep the IoT nodes functional all time,continuous energy is required.For this reason the energy harvested techniques are preferred in IoT networks.Mainly it is preferred to harvest Radio Frequency(RF)energy in the network.In this paper a region based multi-channel architecture is proposed.In which the coverage area of primary node is divided as Energy Harvesting Region and Communication Region.The Secondary User(SU)that are the licensed user is IoT enabled with Cognitive Radio(CR)techniques so we call it CR-enabled IoT node/device and is encouraged to harvest energy by utilizing radio frequency energy.To harvest energy efficiently and to reduce the energy consumption during sensing,the concept of overlapping region is given that supports to sense multiple channels simultaneously and help the SU to find best channel for transmitting data or to harvest energy from the ideal channel.From the experimental analysis,it is proved that SU can harvest more energy in overlapping region and this architecture proves to consume less energy during data transmission as compared to single channel.We also show that channel load can be highly reduced and channel utilization is proved to be more proficient.Thus,this proves the proposed architecture cost-effective and energy-efficient.

CAPACITY EVALUATION OF MULTI-CHANNEL WIRELESS AD HOC NETWORKS

In this paper, the capacity of multi-channel, multi-hop ad hoc network is evaluated.In particular, the performance of multi-hop ad hoc network with single channel IEEE 802.11MAC utilizing different topologies is shown. Also the scaling laws of throughputs for large-scale ad hoc networks and the theoretical guaranteed throughput bounds for multi-channel gridtopology systems are proposed. The results presented in this work will help researchers to choosethe proper parameter settings in evaluation of protocols for multi-hop ad hoc networks.

Network-Wide Time Synchronization in Multi-Channel Wireless Sensor Networks

Recent advances in wireless sensor technology have enabled simultaneous exploitation of multiple channels in wireless sensor systems. In this paper, a novel time synchronization algorithm is proposed for multi- channel Wireless Sensor Networks (WSNs) called Multi-Channel Time Synchronization (MCTS) protocol. Time synchronization is critical for many WSN applications and enables efficient communications between sensor nodes along with intelligent spectrum access. Contrary to many existing protocols that do not exploit multi-channel communications, the protocol takes advantage of potential multiple channels and distributes the synchronization of different nodes to distinct channels and thus, reduces the convergence time of synchronization processes significantly.

Monitoring Sea Fog over the Yellow Sea and Bohai Bay Based on Deep Convolutional Neural Network

In this paper,we utilized the deep convolutional neural network D-LinkNet,a model for semantic segmentation,to analyze the Himawari-8 satellite data captured from 16 channels at a spatial resolution of 0.5 km,with a focus on the area over the Yellow Sea and the Bohai Sea(32°-42°N,117°-127°E).The objective was to develop an algorithm for fusing and segmenting multi-channel images from geostationary meteorological satellites,specifically for monitoring sea fog in this region.Firstly,the extreme gradient boosting algorithm was adopted to evaluate the data from the 16 channels of the Himawari-8 satellite for sea fog detection,and we found that the top three channels in order of importance were channels 3,4,and 14,which were fused into false color daytime images,while channels 7,13,and 15 were fused into false color nighttime images.Secondly,the simple linear iterative super-pixel clustering algorithm was used for the pixel-level segmentation of false color images,and based on super-pixel blocks,manual sea-fog annotation was performed to obtain fine-grained annotation labels.The deep convolutional neural network D-LinkNet was built on the ResNet backbone and the dilated convolutional layers with direct connections were added in the central part to form a string-and-combine structure with five branches having different depths and receptive fields.Results show that the accuracy rate of fog area(proportion of detected real fog to detected fog)was 66.5%,the recognition rate of fog zone(proportion of detected real fog to real fog or cloud cover)was 51.9%,and the detection accuracy rate(proportion of samples detected correctly to total samples)was 93.2%.

Traffic-aware static channel assignment algorithm in wireless mesh networks

A channel assignment algorithm with awareness of link traffic is proposed in multi-radio multi-channel wireless mesh networks. First, the physical interference model based on the signal-to-interference-plus-noise ratio and successful transmission condition is described. The model is more suitable for a wireless communication environment than other existing models. Secondly, a pure integer quadratic programming （PIQP） model is used to solve the channel assignment problem and improve the capacity of wireless mesh networks. Consequently, a traffic- aware static channel assignment algorithm（TASC） is designed. The algorithm adopts some network parameters, including the network connectivity, the limitation of the number of radios and the successful transmission conditions in wireless communications. The TASC algorithm can diminish network interference and increase the efficiency of channel assignment while keeping the connectivity of the network. Finally, the feasibility and effectivity of the channel assignment solution are illustrated by the simulation results. Compared witb similar algorithms, the proposed algorithm can increase the capacity of WMNs.

基于改进的YOLOv4-GhostNet水稻病虫害识别方法

针对水稻病虫害检测精度低、速度慢、模型复杂度高、部署困难等问题,改进了YOLOv4目标检测算法,结合轻量化GhostNet网络,提出了一种基于改进的YOLOv4-GhostNet水稻病虫害识别方法:1)利用幻象模块代替普通卷积结构,替换主干特征提取网络CSPDarkNet53,构建GhostNet模块进行图像的特征提取;2)改进YOLOv4网络的加强特征提取部分PANet结构;3)结合迁移学习与YOLOv4网络训练技巧。通过试验将YOLOv4及其MobileNet系列轻量化网络与Faster-RCNN系列网络和SSD系列网络进行对比,结果表明,改进的YOLOv4-GhostNet模型平均准确率达到79.38%,检测速度可达1 s 34.51帧,模型权重大小缩减为42.45 MB,在保持检测精度达到较高水平的同时模型参数量大幅度降低,适用于部署在计算能力不足的嵌入式设备上。

A Hybrid Deep Learning Scheme for Multi-Channel Sleep Stage Classification

Sleep stage classification plays a significant role in the accurate diagnosis and treatment of sleep-related diseases.This study aims to develop an efficient deep learning based scheme for correctly identifying sleep stages using multi-biological signals such as electroencephalography(EEG),electrocardiogram(ECG),electromyogram(EMG),and electrooculogram(EOG).Most of the prior studies in sleep stage classification focus on hand-crafted feature extraction methods.Traditional hand-crafted feature extraction methods choose features manually from raw data,which is tedious,and these features are limited in their ability to balance efficiency and accuracy.Moreover,most of the existing works on sleep staging are either single channel(a single-lead EEG may not contain enough information)or only EEG signal based which can not reveal more complicated physical features for reliable classification of various sleep stages.This study proposes an approach to combine Convolutional Neural Networks(CNNs)and Gated Recurrent Units(GRUs)that can discover hidden features from multi-biological signal data to recognize the different sleep stages efficiently.In the proposed scheme,the CNN is designed to extract concealed features from the multi-biological signals,and the GRU is employed to automatically learn the transition rules among different sleep stages.After that,the softmax layers are used to classify various sleep stages.The proposed method was tested on two publicly available databases:Sleep Heart Health Study(SHHS)and St.Vincent’s University Hospital/University College Dublin Sleep Apnoea(UCDDB).The experimental results reveal that the proposed model yields better performance compared to state-of-the-art works.Our proposed scheme will assist in building a new system to deal with multi-channel or multi-modal signal processing tasks in various applications.

Traffic-Aware Channelization Medium Access Control for Wireless Ad Hoc Networks

Existing multi-channel Medium Access Control (MAC) protocols have been demonstrated to significantly increase wireless network performance compared to single channel MAC protocols. Traditionally, the channelization structure in IEEE 802.11 based wireless networks is pre-configured, and the entire available spectrum is divided into subchannels and equal channel widths. In contrast, this paper presents a Traffic-Aware Channelization MAC (TAC-MAC) protocol for wireless ad hoc networks, where each node is equipped with a single half duplex transceiver. TAC-MAC works in a distributed, fine-grai-ned manner, which dynamically divides variable-width subchannels and allocates subchannel width based on the Orthogonal Frequency Division Multiplexing (OFDM) technique according to the traffic demands of nodes. Simulations show that the TAC-MAC can significantly improve network throughput and reduce packet delay compared with both fixed-width multi-channel MAC and single channel 802.11 protocols, which illustrates a new paradigm for high-efficient multi-channel MAC design in wireless ad hoc networks.

基于迁移学习与GhostNet模型的农业害虫图像识别研究

针对农作物害虫图像识别,选取10种常见农作物害虫作为检测目标,样本通过IP102数据集筛选并进行数据增强扩充后,构建一个含3 413张图像的农作物害虫数据集Cron Insect。使用轻量级卷积神经网络GhostNet构建识别模型,并运用迁移学习方法训练GhostNet模型,构建一个轻量级的农作物害虫检测网络。基于迁移学习训练的GhostNet网络识别精度为93.64%,未使用迁移学习的GhostNet网络识别精度为92.33%。实验结果表明,基于迁移学习的GhostNet模型更适合于小样本农作物害虫识别,可进一步应用于农业害虫防治中。

A Routing Metric Based on Available Bandwidth Routing in Wireless Mesh Networks

Wireless Mesh Network has drawn much attention due to wide area service coverage with low system cost and being easy to install.However,WMN suffers from high bit error rate,which provides different link capacity among wireless mesh routers.The conventional routing metrics select the path based on link quality.The link with the best quality is preferred as the data transmission path,and thus all nodes likely select the same link,which leads to network performance degradation.This paper proposes a routing metric that considers the available bandwidth and the number of nodes suffering congestion in the path.It is confirmed that the proposed method provides higher network performance of reduced delay,reduced packet loss and increased throughput than conventional routing metrics.

Opportunistic Routing with Multi-Channel Cooperative Neighbour Discovery

Due to the scattered nature of the network,data transmission in a dis-tributed Mobile Ad-hoc Network(MANET)consumes more energy resources(ER)than in a centralized network,resulting in a shorter network lifespan(NL).As a result,we build an Enhanced Opportunistic Routing(EORP)protocol architecture in order to address the issues raised before.This proposed routing protocol goal is to manage the routing cost by employing power,load,and delay to manage the routing energy consumption based on theflooding of control pack-ets from the target node.According to the goal of the proposed protocol techni-que,it is possible to manage the routing cost by applying power,load,and delay.The proposed technique also manage the routing energy consumption based on theflooding of control packets from the destination node in order to reduce the routing cost.Control packet exchange between the target and all the nodes,on the other hand,is capable of having an influence on the overall efficiency of the system.The EORP protocol and the Multi-channel Cooperative Neighbour Discovery(MCCND)protocol have been designed to detect the cooperative adja-cent nodes for each node in the routing route as part of the routing path discovery process,which occurs during control packet transmission.While control packet transmission is taking place during the routing path discovery process,the EORP protocol and the Multi-channel Cooperative Neighbour Discovery(MCCND)protocol have been designed to detect the cooperative adjacent nodes for each node in the routing.Also included is a simulation of these protocols in order to evaluate their performance across a wide range of packet speeds using Constant Bit Rate(CBR).When the packet rate of the CBR is 20 packets per second,the results reveal that the EORP-MCCND is 0.6 s quicker than the state-of-the-art protocols,according to thefindings.Assuming that the CBR packet rate is 20 packets per second,the EORP-MCCND achieves 0.6 s of End 2 End Delay,0.05 s of Routing Overhead Delay,120 s of Network Lifetime,and 20 J of Energy Consumption efficiency,which is much better than that of the state-of-the-art protocols.

Research of Tile Type Transceiver Module Integrating with Two-Dimensional Sum Difference Network

Transceiver module and two-dimensional sum difference network are important components of phased array antenna. In this paper, multilayer printed board is used to integrate millimeter wave multi-channel transceiver circuit and sum difference network. The interconnection between them is realized through RF coaxial vertical transition. At the same time, the heat dissipation design and inter channel shielding design of the module are carried out. The RF and low frequency required by the module are completed through the wiring between and within the dielectric plate layers. Finally, 128 arrays are fabricated and verified by multi-channel passive test. The results show that the type transceiver module integrating with two-dimensional sum difference network has good performance, and 128 channels have excellent amplitude and phase characteristics. The integration technology has the characteristics of lightweight, miniaturization, high integration and low manufacturing cost. It can be widely used in miniaturized phased array antennas.

基于改进GhostNet模型的表情识别研究

针对目前卷积神经网络应用到人脸表情识别任务上时,计算复杂、输入尺度过大、类间差异小和类内差异大等问题,提出了一种基于改进GhostNet模型的解决方法。首先,结合GhostNet模型思想,设计了改进GhostNet模型来提取表情特征,比原网络模型拥有更好的轻量级特性,并解决了Ghost瓶颈层中可能导致信息丢失的问题。其次,结合Island损失函数和Circle损失函数设计思想,设计并采用了基于余弦相似性的损失函数来指导神经网络的学习。该方法可以在特征空间中减小类内差异,增大类间差异,从而提升特征判别能力。在FERplus数据集上进行实验验证,基于改进GhostNet模型方法在模型参数量和计算量更少的前提下,依旧有着更高的识别准确率和更快的识别速度,效果优于Softmax交叉熵损失函数和基于余弦距离的损失函数等,更适合移动端和嵌入式设备的使用场景。

基于改进DeepLabv3+算法的起重机锈迹检测

室外工作的起重机金属结构易产生锈蚀现象,严重的锈蚀会导致结构承载能力显著降低,从而引发灾难性事故。文中针对当前起重机人工锈迹巡检中存在的漏检、误检和费时等问题,提出一种基于改进DeepLabv3+算法的自动化锈迹检测方法。该方法依托于机器视觉,将原始DeepLabv3+的骨干网络替换为幽灵网络(GhostNet)以提升网络的轻量化程度;使用特征金字塔网络(FPN)进行特征提取,用于抑制噪声和背景对锈迹提取的不良干扰;引入空间感知独立自注意机制(SSA)来提高网络区域感知性能;最后使用特征融合(Add)代替原始网络的特征堆叠来降低算法参数量。将所提方法应用于室外起重机锈迹检测,结果表明所提算法的检测性能优于原始算法和其他经典语义分割算法,具有重要的工程应用价值。

基于电子舌结合VMD-IMG组合模型的清香型白酒掺假检测

针对传统白酒掺假检测方法分析流程长、仪器操作复杂等问题,提出一种基于电子舌结合变分模态分解和改进多通道GhostNet网络组合模型,实现对白酒掺假快速检测的方法。采用贝叶斯优化后的VMD对信号进行自适应模态分解,将原始信号分解为多个本征模态函数。在传统GhostNet网络的基础上,引入瓶颈注意力模块和空洞卷积,并扩展为多通道GhostNet,进而构建为VMD-IMG组合模型,以实现电子舌信号特征的有效提取和分类识别。试验结果表明,VMD-IMG组合模型具有较好的白酒掺假分辨性能,其测试的准确率、精确率、F1分数分别为98.33%,98.43%,98.34%。研究为白酒掺假检测提供一种快速、低成本的检测方法,并可为其他酒类或饮品掺假检测提供参考。

基于改进YOLOX-tiny算法的交警手势识别

为了在城市中实现无人驾驶,需要能够高效检测交警的现场指挥手势。针对现有手势识别算法识别精度低、检测速度慢、难以应对复杂道路环境等问题,提出一种改进的YOLOX-tiny交警手势识别算法。首先,使用改进后的GhostNet网络替换原主干网络,并且插入坐标注意力机制,全面提取输入图像特征,提高了网络的检测精度,同时提升了对中小型目标的检测效果;其次,改进解耦头部分,设计了SCDE Head结构,在减少计算量的同时过滤冗余信息,使得解耦头更有效率,并且解耦头融合了多尺度的特征,提升了目标检测准确率;最后,将SIoU应用到定位损失中,加快网络收敛的速度,提升回归精度。在自制的交警指挥手势数据集上进行测试,实验结果表明,与YOLOX-tiny模型对比,改进后算法参数量减少了27.97%,模型计算量减少了33.31%,且平均检测精度提高了2.31%,检测速度提升了45%,更适合汽车无人驾驶以及硬件部署方面的实际需求。