传感器网络中一种基于估计代价的数据聚合树生成算法

无线传感器网络是一种全新的技术,能够广泛应用于恶劣环境和军事领域.传感器网络在数据收集中,为减少冗余数据的传输耗能,降低延迟,需要采用数据聚合技术.本文采用定向传输方式,在消息路由机制基础上提出了一种基于估计代价的数据聚合树生成算法.该算法主要思想在于将节点能耗、传输距离与聚合收益三方面作为估计代价,优化聚合路径,实现数据聚合在能量与时延上的折中.     

基于信誉的延迟容忍网络激励方案

由于长延迟、频繁中断等特性,延迟容忍网络采用“存储-携带-转发”的路由方式实现报文的传递.这种传输方式建立在网络节点互相信任,诚实合作的假设之上.然而,由于受到自身资源的限制,网络中存在自私节点拒绝参与网络合作的行为,导致网络传输性能下降,甚至无法正常运行.针对此类问题,本文提出一种基于信誉的延迟容忍网络激励方案,鼓励节点共享资源,并约束自私行为.设计了一种观测协议,以解决延迟容忍网络环境下对网络节点行为的观测问题;提出了一种信誉模型,通过信任值评价节点的可信程度,从而优化路由决策,惩罚自私节点.仿真结果表明了本文提出的激励方案的有效性和可行性.     

WDM网络中的一种基于BFP-Tree方法的负载平衡路由机制

文章提出一种基于数据包频率模式-树结构(BFP-Tree)方法的负载平衡路由机制,该机制首先统计所有路径节点的使用频度;然后选取使用频度最高的节点路径,按照网络中各链路使用波长数的统计方差最小的原则,对业务请求进行路由选择.仿真结果表明:该方法在阻塞概率和传输延迟性能方面都优于BFP-Tree路由机制.     

无线Ad Hoc网络容量及延迟性能的研究

网络容量与传输延迟是评估无线Ad Hoc网络性能的两个重要参数。对传输延迟 保持在一定门限下如何使Ad Hoc网络获得较高容量进行了研究。对Ad Hoc网络建模,将模 型添加不同的假设条件简化分析复杂度,通过具体公式说明网络增加移动节点能够提高网络 容量。基于延迟条件提出一种路由算法,它支持网络动态变化,使吞吐量接近理论优化值成为 可能。     

一种适用于虚拟机网络的数据包高效传输方法

为了提高数据包在云计算数据中心中基于虚拟机构成网络中的传输性能,提出了一种基于网络编码的高效数据包传输方法.基于网络编码机制,采用对传输过程中丢失数据包高效的编码组合策略,多个虚拟机终端可以在一次多播或广播传输中获取多个从交换机优先传输的数据包,因此,提出的方法可以提高基于虚拟机网络的多播及广播业务的数据包传输延迟,并提高多播及广播业务的网络吞吐量.仿真结果表明提出的方法在典型信道条件下均获得了较好的数据包传输时延及网络吞吐量性能.     

延迟容忍网络自适应差错控制方法

根据延迟容忍网络的时延较大、误码率较高等特点,在现有的传输控制层和MAC层差错控制的基础上改进,结合IEEE 802.11 DCF机制的特点,提出一种基于MAC层丢包率的自适应差错控制方案.该方案能有效改善延迟容忍网络的传输时延,满足对时延要求比较敏感的业务的要求.     

无线传感器网络多频率查询的节能优化

在多频率查询的无线传感器网络中,多个接收节点以不同的查询频率请求同一个数据源节点的数据.由于查询频率的不同,如何共享传输链路以降低能耗成为一个新的研究问题.考虑数据序列之间的数据相关性,本文提出了一种节能优化方法,通过在共享链路上广播整合频率的数据序列,再根据需要重构出与查询请求相对应的数据序列.理论分析和模拟实验均表明,在可容忍的平均相对误差下,该方法能够节省能量的消耗.     

机会网络中基于摆渡节点与簇节点相互协作的路由机制

机会网络是一种不需要在源节点和目的节点之间存在完整路径,利用节点移动带来的相遇机会实现网络通信的延迟容忍自组织网络,它以“存储-携带-处理-转发”的模式进行.为实现互不相交簇间的信息传输,本文设计了一种带阈值的簇移动模型CMMT,并提出了一种基于摆渡（Ferry）节点与簇节点协作的路由算法（CBSW）.该算法减少了冗余的通信和存储开销,以及在Spray阶段簇节点没有遇到目的节点或摆渡节点,进入Wait阶段携带消息的节点采用直接分发方式只向目的节点传输等问题.仿真实验表明,CBSW算法能够增加传输成功率,减少网络开销和传输延迟.     

片上网络GALS接口设计

本文提出了一种基于握手协议的GALS接口设计方法。该接口采用异步FIFO作为输入缓冲区,有效降低了数据传输延迟;采用环形缓冲的概念来管理缓冲区,使接口具有了可扩展性。FPGA验证结果表明,该接口保证了适配单元与网络路由之间完成准确的异步传输,4通道的接口共占用了405个ALUT（Adaptive Look-Up Table）和支持211 MHz的时钟频率。     

基于运动状态的延迟容忍移动传感器网络数据传输策略

针对延迟容忍移动传感器网络(DTMSN, delay tolerant mobile sensor networks)提出一种基于节点运动状态的数据传输(MSD, motion state-based delivery)策略.MSD使用2个通信频率,传感器节点利用基站在频率f1上的广播信号完成自身运动状态和传输概率的计算,并以此作为利用频率f2进行消息转发的依据.为了提高性能,MSD中利用消息生存时间和自身消息优先传输机制来完成消息队列的管理.仿真结果表明,与现有的几种DTMSN数据传输策略相比,MSD能够以较低的通信开销和传输延迟达到较高的数据传输成功率.     

Performance improvements of communication-based train control (CBTC) systems with unreliable wireless networks

Communication-based train control (CBTC) uses wireless networks, such as wireless local area networks (WLANs), to transmit trains status and control commands. CBTC systems have stringent requirements for train-ground communications, as the train control system depends on the accurate, timely and reliable data over the communication links. However, since WLANs are not originally designed for high mobility environment, random packet delay and losses are inevitable in WLAN-based CBTC systems, which could result in unnecessary traction, brake or even emergency brake of trains, losses of line capacity and passenger satisfaction. In this paper, we study the impacts of random packet delay on the performance of CBTC systems, and propose a novel scheme to improve the CBTC performance by mitigating the impacts of random packet delay. Unlike the existing works that only consider a single train, we consider a group of trains in CBTC systems to improve the CBTC performance. Extensive field test and simulation results are presented. We show that the current adopted control scheme in CBTC systems needs to keep stability and performance under transmission delays through increasing the distance between trains in steady state which means cost of line capacity. By contrast, our proposed scheme can significantly improve the CBTC performance.     

Blind adaptive multiuser detection for DS/CDMA over multipathfading channels

Using a hard null scheme, multipath fading and multiple access interference suppression can be realised for a multiple constrained minimum variance (MCMV) detector at the same time. A modified version of the MCMV detector is also presented, which utilises the eigenstructure of the correlation matrix to enhance the performance of the MCMV detector. Numerical results demonstrate the effectiveness of the proposed detectors     

无线传感器网络自适应功率控制策略

无线传感器网络功率控制技术对于网络的拓扑连通、能量效率、网络容量、吞吐量、实时性等性能均有显著影响,是其实用化的重要支撑技术。该文提出了一种适用于无线传感器网络的自适应功率控制策略APCS(Adaptive Power Control Strategy),该策略是只需要局部信息的分布式算法,通过调整路径损耗指数和功率控制参数可以获得性能极佳的目标拓扑,并能满足实时性和容错能力要求较高的应用场景。另外,该算法还采用了动态功率调整以保持网络的连通性,延长网络的生命周期。仿真结果证实了所提方法的有效性。     

Mobility management approaches for SDN-enabled mobile networks

The evolving network technologies aim at meeting the envisioned communication demands of future smart cities and applications. Although software-defined networking (SDN) enables flexible network control, its applicability to mobile networks is still in its infancy. When it comes to introducing the SDN vision to mobile networks, handling of wireless events and mobility management operations stand out as major challenges. In this paper, we study the scalability issues of SDNized wireless networks, specifically those relevant to mobility management. We design and implement different mobility management approaches in SDNized wireless networks and investigate the impact of various system variables on the overall handover delays. We also study the improvements in handover delays: (i) when a proposed proactive mobility management algorithm is implemented; (ii) when the controller delegates partial control of mobility management to the forwarding entities. For the implementation of the proposed approaches on the OpenFlow network, the paper also suggests potential extensions to the OpenFlow protocol. The contributed approaches are validated on a full-scale demonstrator, with results showing that proactive outperforms reactive and that the delegated control approach performs better than proactive for smaller topology sizes. Furthermore, a proposal for LTE X2-specific control delegation is discussed as a use case.     

Multilevel fault tolerance in infrastructure‐oriented wireless networks: framework and performance evaluation

Significant advances have been made in the design and implementation of dependable systems and networks over the last several years. However, many wireless networks have not been designed for highly dependable operation owing to network cost and complexity, and a lack of regulatory requirements on wireless service quality. One way to address this significant challenge is by introducing fault tolerance; however, very limited work has been done so far in the fault‐tolerant design of wireless networks. In this paper, we address how to utilize fault tolerance in the design of infrastructure‐oriented wireless networks. More specifically, an architectural design scheme is presented for multilevel fault tolerance using adaptable building blocks. The scheme utilizes ‘selective’ redundancy at component, link and block levels and a fault‐tolerant architecture for interconnection of building blocks. The design scheme has been implemented in both analytical and simulation models. The detailed performance results show that fault tolerance at component, link, block, and interconnection levels can significantly improve the overall dependability performance. One interesting observation is that, to achieve highest dependability, fault tolerance at link, component or block level is not sufficient and must be combined with the interconnection level fault‐tolerance. Copyright © 2006 John Wiley & Sons, Ltd.     

Autonomous time synchronization among radio ports in wirelesspersonal communications

In digital wireless access communications systems, a large number of radio ports are deployed to provide wide-area coverage. Achieving time synchronization among ports is beneficial to these systems, especially for systems providing access to the infrastructure networks such as local exchange networks. This leads to better cochannel interference management and easier control for automatic link transfer. This paper describes a practical “over-the-air” algorithm which is autonomously performed by individual ports and hierarchically controlled by the ports having the most reliable timing. This process involves only small overhead for demodulating special timing bits transmitted by other ports and adjusting timing accordingly. Computer simulations based on a time division multiple access (TDMA) system with port spacing of 2000-3000 feet and 500 kb/s transmission rate are used as an example to evaluate possible impacts on wireless access     

WSNs中基于数据匹配的分布式故障节点检测算法

针对无线传感网络的节点故障问题,提出一种新的分布式故障节点检测算法(DFDA).DFDA算法利用节点度信息估计节点对网络的重要性,并尽可能将节点度高的节点保存到网络中.通过比较节点间感测的数据,检测故障节点.为了增强检测的准确性,采用双重测定策略.仿真结果表明,相比于同类算法,DFDA算法提高了检测故障节点的精确度,并...     

Fast Handoff in SIP-Based Next Generation Mobile Networks

It is commonly held that next generation mobile systems will be developed on the Internet in combination with diverse access technologies, as the future network architecture will be the coming together of various overlapping wireless access networks. Integrating various wireless networks in future heterogeneous networking environments poses many difficulties, the most critical challenge of which is efficient support for seamless mobility. SIP is a promising nominee for managing mobility in heterogeneous networks as it provides mobility within the application layer and the characteristics of the lower layer protocols are invisible to it. However, the performance of SIP-based mobility management is downgraded, resulting from its adoption of TCP/UDP for signaling and its strict separation between the lower layers and the application layer of the protocol stack. In this paper, a SIP-based cross-layer design for fast handoffs is proposed to shorten the service interruption time when a mobile node crosses the overlapped area of a WLAN/3G cellular system. As will be shown by the simulation results, the SIP-based solution proposed in this paper effectively lessens the handoff delays caused by either the horizontal handoff or vertical handoff in future all-IP heterogeneous wireless networks.     

IPv6-based dynamic coordinated call admission control mechanism over integrated wireless networks

Many wireless access systems have been developed recently to support users mobility and ubiquitous communication. Nevertheless, these systems always work independently and cannot simultaneously serve users properly. In this paper, we aim to integrate IPv6-based wireless access systems and propose a coordinated call admission control mechanism to utilize the total bandwidth of these systems to minimize the call blocking probabilities, especially the handoff call dropping probabilities. First, we propose an integrated hierarchical wireless architecture over IPv6-based networks to combine the wireless access systems including cellular systems (second-generation, General Packet Radio Service, or third-generation), IEEE 802.11 a/b/g WLAN, and Bluetooth. In the proposed architecture, mobile user can request a call with quality-of-service (QoS) requirements by any wireless network interfaces that can be accessed. When the proposed coordinated call admission control (CCAC) mechanism receives a request, it takes the QoS requirements of the incoming call and the available and reserved bandwidth of this wireless system into consideration to accept or reject this request. Besides, the mechanism can coordinate with other wireless systems dynamically to adjust the bandwidth reserved for handoff calls at each wireless system in this architecture so as to reduce the call blocking probabilities. Once the call is admitted, the mobile user is able to access heterogeneous wireless access networks via multiple interfaces simultaneously. Finally, we evaluate this system to show that the CCAC on the proposed architecture outperforms other mechanisms proposed before.     

Soft capacity analysis of TDMA systems with slow-frequency hoppingand multiple-beam smart antennas

Smart antenna is considered as one of the most effective means for enhancing wireless system capacity. When fractional loading is accompanied with slow-frequency hopping (SFH), soft capacity can be realized in time-division multiple access (TDMA) wireless networks. Then, the interference reduction due to smart antennas, power control, and discontinuous transmission can be directly translated into capacity gain. This paper addresses the capacity gain due to multiple-beam (MB) smart antennas in TDMA wireless systems with soft capacity. The system capacity is determined analytically and by simulation. MB smart antennas with practical antenna pattern are used in this study. Perfect power control and discontinuous transmission are assumed in the simulation and the theoretical analysis. A novel call admission control algorithm is proposed to enhance the system capacity without degrading the signal quality. The TDMA system is assumed to be global system for mobile communications (GSM)-like, however, the analysis can be extended and applied to other TDMA systems