小波分解和组合模型相融合的网络流量网络预测

为了提高网络流量预测的准确性,提出一种小波分解和组合模型相融合的网络流量预测预测模型。首先采用小波分析对网络流量进行分解,得到网络流量的趋势序列和波动序列,然后分别采用自回归差分滑动平均模型和极限学习机对它们进行建模和预测,最后采用仿真实验测试组合模型的性能。仿真结果表明,相对于其它网络流量预测模型,组合预测模型提高了网络流量预测精度,降低了预测误差更小。     

基于小波分解和支持向量机的网络流量组合预测

研究网络流量预测问题,网络流量具有突发性、周期性、非线性特点,传统网络流量预测模型无法建立准确预测模型,导致预测误差大,预测精度低.为了提高网络流量的预测精度,提出一种小波分解和支持向量机的网络流量预测模型.首先采用小波变换对网络流量进行分解,把网络流量不同特性成分分离出来,然后采用支持向量机对各分量进行预测,最后采用小波变换对各分量预测结果进行重构,得到网络流量的最终预测结果.仿真实验结果表明,相对其它预测模型,提高了网络流量的预测精度,为网络流量预测优化提供了可靠依据.     

组合神经网络的网络流量预测研究

提出一种组合神经网络的网络流量预测模型.首先采用SMOF网络对网络流量数据进行聚类,然后采用Elman网络对聚类后的流量数据进行训练并预测,同时采用遗传算法对Elman网络的网络结构进行优化,提高网络流量预测精度.仿真结果表明,组合神经网络加快了网络流量预测速度,提高了网络流量预测精度,克服了单一预测模型不足,为网络流量预测提供了新的思路,具有很好的应用前景.     

小波消噪和蚁群优化支持向量机的网络流量预测模型

针对当前网络流量预测模型精度低的缺点,本文提出了一种新型的小波消噪和蚁群算法优化支持向量机的网络流量预测模型。首先采用小波阈值法对网络流量进行消噪处理;然后将网络流量输入到支持向量机中学习,并采用蚁群算法对支持向量机的参数进行优化,建立网络流量预测模型,最后采用实际网络流量数据进行仿真实验,结果表明,相对于其它网络流量预测模型,本文模型提高了网络流量的预测精度,具有更好的鲁棒性。     

基于包容性检验和神经网络的网络流量预测

网络流量具有复杂多变特征,为了获得理想的预测效果,提出一种包容性检验和BP神经网络相融合的网络流量预测模型(ET-BPNN)。首先采用多个单一模型对网络流量进行预测,然后通过包容性检验选择最合适的基本模型,最后采用BP神经网络确定基本模型权重,建立网络流量预测模型。结果表明,ET-BPNN更加准确地刻画了网络流量变化趋势,各项评价指标均达到更优,为实现网络流量准确预测提供了更为科学的方法。     

基于组合核函数高斯过程的网络流量预测

为了提高网络流量的预测精度,提出一种组合核函数高斯过程的网络流量预测模型。首先采用混沌理论构建网络流量的学习样本,然后将网络流量的训练样本输入组合函数高斯回归模型进行训练,建立网络流量预测模型,最后采用多个网络流量数据进行单步和多步预测对比测试。结果表明,相对于对比模型,本文模型准确描述了复杂多变的网络流量变化趋势,提高了网络流量的预测精度。     

基于增量优化极端学习机的网络流量预测模型

为了获得更加理想的网络流量预测结果,针对极端学习机人工设置隐层节点数目的不足,提出一种增量优化极端学习机的网络流量预测模型。首先对极端学习机工作原理和不足进行分析,然后采用增量优化方式提高极端学习机的性能,最后采用具体网络流量时间序列对增量优化极端学习机的性能进行仿真试验。结果表明,相对于其它网络流量预测模型,增量优化极端学习机不仅加快了网络流量建模速度,可以适合于网络流量的长期和在线预测,而且提高了网络流量的预测精度。     

相空间重构和正则极限学习机的网络流量预测

网络流量预测一直是网络研究技术中的热点,针对网络流量变化的时变性、混沌性,提出一种相空间重构和正则极限学习机的网络流量预测模型。首先收集大量的网络流量历史样本,并进行相应的预处理,然后根据混沌理论确定最优延迟时间和嵌入维数,并重构网络流量学习样本,最后采用正则极限学习机建立网络流量预测模型,并进行仿真对比实验。结果表明,相对于其它网络流量预测模型,本文模型可以更加准确描述网络流量的非线性变化特点,提高网络流量预测精度,预测结果具有一定实用价值。     

基于包容性检验和神经网络的网络流量预测模型

网络流量具有高度复杂的非线性特征,采用单一预测模型往往难以达到理想的预测效果,为此,提出一种包容性检验和BP神经网络相融合的网络流量预测模型（ET-BPNN）。首先采用多个单一模型对网络流量进行预测,然后通过包容性检验,根据t统计量检验选择最合适的基本模型,最后采用BP神经网络对基本模型预测结果进行组合得到最终预测结果。实验结果表明,相对于单一模型以及传统组合模型,ET-BPNN更加准确刻画了网络流量变化趋势,各项评价指标均达到更优,为实现网络流量准确预测提供了更为科学的方法。     

遗传优化神经网络的网络流量预测

为了提高网络流量预测精度,利用相空间重构和预测模型参数间的相互联系,提出一种遗传算法优化神经网络的网络流量预测模型.首先将相空间重构和神经网络参数进行编码,网络流量预测精度作为目标函数,然后通过遗传算法选择模型最优参数,最后进行网络流量仿真实验.实验结果表明相对传统预测模型,遗传优化神经网络模型具有更高预测精度及稳定性更好.     

Power Efficient Video Multipath Transmission over Wireless Multimedia Sensor Networks

This paper proposes a power efficient multipath video packet scheduling scheme for minimum video distortion transmission (optimised Video QoS) over wireless multimedia sensor networks. The transmission of video packets over multiple paths in a wireless sensor network improves the aggregate data rate of the network and minimizes the traffic load handled by each node. However, due to the lossy behavior of the wireless channel the aggregate transmission rate cannot always support the requested video source data rate. In such cases a packet scheduling algorithm is applied that can selectively drop combinations of video packets prior to transmission to adapt the source requirements to the channel capacity. The scheduling algorithm selects the less important video packets to drop using a recursive distortion prediction model. This model predicts accurately the resulting video distortion in case of isolated errors, burst of errors and errors separated by a lag. Two scheduling algorithms are proposed in this paper. The Baseline scheme is a simplified scheduler that can only decide upon which packet can be dropped prior to transmission based on the packet’s impact on the video distortion. This algorithm is compared against the Power aware packet scheduling that is an extension of the Baseline capable of estimating the power that will be consumed by each node in every available path depending on its traffic load, during the transmission. The proposed Power aware packet scheduling is able to identify the available paths connecting the video source to the receiver and schedule the packet transmission among the selected paths according to the perceived video QoS (Peak Signal to Noise Ratio—PSNR) and the energy efficiency of the participating wireless video sensor nodes, by dropping packets if necessary based on the distortion prediction model. The simulation results indicate that the proposed Power aware video packet scheduling can achieve energy efficiency in the wireless multimedia sensor network by minimizing the power dissipation across all nodes, while the perceived video quality is kept to very high levels even at extreme network conditions (many sensor nodes dropped due to power consumption and high background noise in the channel).     

Queue Analysis and Multiplexing of Heavy-tailed Traffic in Wireless Packet Data Networks

Recent research based on traffic measurements shows that Internet traffic flows have a fractal nature (i.e., self-similarity property), which causes an underestimation of network engineering parameters when using the conventional Poisson model. Preliminary field measurements demonstrate that packet data traffic in wireless communications also exhibits self-similarity. In this paper, we investigate the queuing behavior of self-similar traffic flows for data applications in a packet-switching single-server wireless network. The traffic is generated by an on–off source with heavy-tailed on periods and exponentially distributed off periods. We extend previous analysis of a relation among the asymptotic distribution of loss probability, traffic specifications, and transmission rate for a wireline system to a wireless system, taking into account wireless propagation channel characteristics. We also investigate the multiplexing of heavy-tailed traffic flows with a finite buffer for the downlink transmission of a wireless network. Computer simulation results demonstrate that assumptions made in the theoretical analysis are reasonable and the derived relationships are accurate.     

Optimal choice of Reed-Solomon codes to protect against queuing losses in wireless networks

This article proposes algorithms to determine an optimal choice of the Reed-Solomon forward error correction (FEC) code parameters (n,k) to mitigate the effects of packet loss on multimedia traffic caused by buffer overflow at a wireless base station. A network model is developed that takes into account traffic arrival rates, channel loss characteristics, the capacity of the buffer at the base station, and FEC parameters. For Poisson distributed traffic, the theory of recurrent linear equations is applied to develop a new closed form solution of low complexity of the Markov model for the buffer occupancy. For constant bit rate (CBR) traffic,an iterative procedure is developed to compute the packet loss probabilities after FEC recovery.     

Adaptive Rate Optimal Design for Wireless Heterogeneous Networks: Jointly Congestion Discrimination and Queue Updating

In this paper, we propose a novel rate adaptive optimization scheme for streaming media transmission over wireless heterogeneous IP networks. In the proposed adaptive scheme, through the analysis of the packet loss characteristics in wireless channel, we develop the relationship between the packet loss rates and the packet sizes. Furthermore, the scheme detects the nature of packet losses by sending large and small packets alternately, and then adopts an adaptive rate optimization strategy to decrease the network congestion and increase the network throughput. Using congestion discrimination and updating factor, the scheme can adapt to the changes of network states quickly and improve delivery quality of wireless multimedia streaming. Simulation results show that, in comparisons to the existing rate optimization algorithms, our proposed scheme offers significantly improved performance in terms of throughput and network congestion, especially when the channel quality is poor in different network topology environments.     

传感器网络变权组合预测流量整形算法

针对现有流量整形算法在传感器网络应用上的不足,提出了一种新的流量整形算法。分析了传感器网络流量具有突发随机性以及时变不均衡性的原因,根据传感器网络流量的模糊性、随机性以及时变性统一建模,提出了变权组合预测流量整形算法（TSAV,Traffic Shaping Algorithm with Variable weight combination forecast）,该算法通过逼近最优组合理论分配模糊AR预测与Kalman预测的组合权重,得到更为精确的预估流量值,提前规划整形速率从而平滑的输出分组流。实验表明,TSAV算法应用到传感器网络时能够准确预测流量,减少分组丢弃率的同时增大网络吞吐量,改善了传感器网络信息传输的QOS性能。      

一种环境感知的无线Mesh网络自适应QoS路径选择算法

本文针对如何改善无线多跳Mesh网络的服务质量,满足无线多媒体业务对数据传输的带宽、时延、抖动的要求等问题,研究了一种基于无线信道状态和链路质量统计的MAC层最大重传次数的自适应调整算法。该算法通过对无线Mesh网络的无线信道环境的动态感知,利用分层判断法区分无线分组丢失的主要原因是无线差错还是网络拥塞导致,实时调整MAC层的最佳重传次数,降低无线网络中的分组冲突概率。基于链路状态信息的统计和最大重传策略,提出了一种启发式的基于环境感知的QoS路由优化机制HEAOR。该算法通过动态感知底层链路状态信息,利用灰色关联分析法自适应选择最优路径,在不增加系统复杂度的基础上,减少链路误判概率,提高传输效率。NS2仿真结果表明,HEAOR算法能有效减少重路由次数,降低链路失效概率,提高网络的平均吞吐率。本文提出的方法不仅能够优化MAC层的重传,而且通过发现跨层设计的优化参数实现对路径的优化选择。      

Congestion avoidance in cognitive wireless sensor networks using TOPSIS and response surface methodology

Congestion in wireless sensor networks degrades the quality of the channel and network throughput. This leads to packet loss and energy dissipation. To cope with this problem, a two-stage cognitive network congestion control approach is presented in this paper. In the first stage of the proposed strategy, initially downstream nodes calculate their buffer occupancy ratio and estimate congestion degree in the MAC layer. Then, they send the estimated value to both network and transport layers of their upstream nodes. The network layer of the upstream node uses TOPSIS in order to rank all neighbors to select the best one as the next relay node. In the second stage, transport layer of the given node adjusts the transmission rate using an optimized regression analysis by RSM. Extensive simulations demonstrated that the proposed method not only decreases packet loss, but also significantly improves throughput and energy efficiency under different traffic conditions, especially in heavy traffic areas. Also, Tukey test is used to compare performance of algorithms as well as to demonstrate that the proposed method is significantly better than other methods.     

Performance Analysis of Network Coding Based Two-Way Relay Wireless Networks Deploying IEEE 802.11

In this paper, we investigate the performance analysis of the IEEE 802.11 DCF protocol at the data link layer. We analyze the impact of network coding in saturated and non-saturated traffic conditions. The cross-layer analytical framework is presented in analyzing the performance of the encode-and-forward (EF) relaying wireless networks. This situation is employed at the physical layer under the conditions of non-saturated traffic and finite-length queue at the data link layer. First, a model of a two-hop EF relaying wireless channel is proposed as an equivalent extend multi-dimensional Markovian state transition model in queuing analysis. Then, the performance in terms of queuing delay, throughput and packet loss rate are derived. We provide closed-form expressions for the delay and throughput of two-hop unbalanced bidirectional traffic cases both with and without network coding. We consider the buffers on nodes are unsaturated. The analytical results are mainly derived by solving queuing systems for the buffer behavior at the relay node. To overcome the hidden node problem in multi hop wireless networks, we develop a useful mathematical model. Both models have been evaluated through simulations and simulation results show good agreement with the analytical results.     

基于机器学习和卫星图像的路径损耗预测

基于反向传播神经网络（back propagation neural network,BPNN）构建了一种路径损耗预测模型. 通过卫星图像的红、绿、蓝（red, green and blue,RGB）通道的颜色信息来表征无线通信电波传播路径的环境特征,结合路测点与基站的距离特征构建数据集,迭代训练网络参数,以预测传播路径损耗. 结果表明,对跨基站路测点的预测结果与实测数据之间的相关系数达到0.83,绝对平均误差控制在0.66 dB,标准差控制在6.65 dB,说明在缺乏某一场景的详细模型和材质参数时,本文模型也能可靠预测无线通信电波的传播路径损耗. 此外,本文信道模型与传统信道建模方法多方面的对比与分析表明,本文模型在相同计算资源下可以提供和传统信道建模方法相差很小的预测结果,同时大大缩短预测所需的时间,说明本文模型对传播路径损耗做出快速预测的能力可以用于无线通信网络系统的优化.