COPS：802.16e中一种协作节能算法

 该文首先对多业务场景下移动站休眠模式的能量消耗情况进行了理论分析,发现由于不同功率节省类之间侦听窗口的不同步,移动站的节能效率很低;然后提出了一种多功率节省类的协作节能算法（COoperative Power Saving Mechanism, COPS）,通过调整功率节省类I的休眠参数,同步功率节省类Ⅰ和Ⅱ的侦听窗口,从而提高节能效率. 理论分析和仿真实验结果显示COPS不仅提高了节能效率还降低了MAC帧的反应时间.     

基于区分业务的OFDM跨层动态资源分配算法

跨层设计是目前比较流行的、有效的无线系统设计方法.本文针对多业务(实时业务和非实时业务)正交频分复用(OFDM)系统,提出了一种跨层结构的动态资源分配算法.该算法不仅在物理层上考虑信道状态信息和功率限制条件,而且在MAC层上针对不同类型业务采取不同的调度策略:(1)对于高优先级的实时业务,采用最大化容量的OFDMA策略,在频域上最大限度地利用多用户分集提高频谱效率,在时域上通过时隙的分配来保证时延边界;(2)对于低优先级的非实时业务,在频域上采用基于比例公平的OFDMA策略,在时域上尽力而为地分配时隙.仿真结果表明,与传统的单层资源分配算法相比,所提算法能够在保证不同业务服务质量(QoS)的前提下,大幅度地提高系统性能.     

LR-PON中改进的节能动态带宽分配算法

长距离无源光网络(Long-reach Passive Optical Networks,LR-PON)的覆盖范围超过100 km,光网络单元(Optical Network Units,ONU)在一个轮询周期中处于空闲状态的时间较长.如果使处于空闲状态的ONU进入低功耗(休眠或假寐)模式,可以降低ONU的能耗.基于此,文章提出一种改进的节能动态带宽分配算法.该算法通过改变ONU的调度顺序,使得ONU在上下行传输结束后能够立刻进入休眠状态,而无需等待授权帧的到达,从而延长ONU的休眠时间.仿真结果表明:所提出算法能够降低ONU能耗和下行数据包平均时延.     

面向混合业务的无线传感器网络能量有效接入策略

研究了在实时业务和非实时业务同时存在的混合背景下,非实时业务的无线传感器节点自适应侦听和睡眠的动态接入机制。网络节点处于睡眠状态时所需的能量很低,节约了无线传感器网络节点的平均能量消耗;但是,过长的睡眠时间可能使得网络节点错失传输机会。因此,根据信道的使用情况,合理地设定无线传感器网络节点的睡眠时间,能够在网络能量消耗和传输效率之间进行调整,从而最大化无线传感器网络的能量传输效率。首先,利用连续时间 Markov 方法对问题进行建模,并利用基于摄动分析理论对系统模型进行分析,获得求解无线传感器网络能量效率最大化的最优睡眠时间梯度算法。最后通过理论结果和计算机仿真模拟的对比,验证了推荐方法的可行性。     

带有卸载压缩激励的云增强FiWi网络节能机制

针对云增强型光纤-无线(FiWi)网络能耗以及卸载的通信开销过大问题,该文提出一种自适应卸载压缩节能机制(ESAOC),针对不同类型的业务属性和最大的容忍时延,结合光网络单元的负载变化和无线网状网的流量情况,通过统计的方式获得不同优先级卸载数据的平均到达率,再结合各个节点的压缩时延,动态调整业务的卸载压缩比,以降低卸载的通信开销;同时,建立排队模型分析卸载业务在MEC服务器的排队时延,协同调度无线侧中继节点,进而对光网络单元和终端设备进行协同休眠调度,最大化休眠时长,提高系统能源效率。结果表明,所提方法在有效降低整个网络能耗的同时能够保证卸载业务的时延性能。     

基于模糊流感知的动态优先公平调度算法

为适应网络的动态性,提高调度公平性和资源效率,流感知优先公平调度机制需要动态的多业务区分转发。与链路负载状态相关的模糊流感知能够实现路径上的一致性业务区分,而基于模糊流感知的动态优先公平调度算法通过调整优先队列负载门限在流式流和弹性流之间实现转发优先权的动态交替,在链路轻载时实现不同流间的相对公平调度,在链路重载时则强调实时业务的绝对优先权以保证其时延要求。算法公平性分析和仿真计算显示提出算法的动态区分转发通过适度增加优先业务队长能够大幅度提高弹性流的接纳率,具有较高的链路平均吞吐量和资源效率。     

多业务DS-CDMA系统的资源分配算法研究

针对提出的业务模型，分析了多业务DS-CDMA系统中各类业务的QoS性能如平均延时、吞吐量及比特差错率等与各类业务的处理增益和功率分配的关系，提出了一种适用于两类业务（一类实时、一类可靠）的优化处理增益和功率分配算法，数据结果表明优化算法可以在保证两类业务QOS的同时提高系统容量及可靠业务的吞吐量。     

IP over WDM网络中能耗自感知的混合疏导专有保护算法

针对绿色生存性IP over WDM网络中资源开销大、网络阻塞率高的问题,提出一种能耗自感知的混合疏导专有保护(HG-PA-DPP)算法。首先在IP层的核心路由器对低粒度业务请求集中疏导;然后通过实时感知WDM层的链路负载和双层器件带来的链路能耗状态定义链路权值,基于分层图在不同波长平面上为各业务请求寻找权值小且链路分离的工作路径和保护路径,同时在业务的中间节点处对不需要光-电-光转换的光路进行光旁路处理;最后,将空闲或保护资源设置为休眠模式以实现节能。仿真结果表明,所提算法在高负载时有着更好的节能效果,且在大网络拓扑下资源开销(RO)为传统算法的76.5%,阻塞率(BP)仅为传统算法的61.1%。     

非实时业务下认知网络的频谱共享

根据某种特定业务的QoS要求制定相应的频谱共享策略,可以在这种情况下提高系统的频谱利用率,增加用户平均收益。对非实时业务下的认知网络的频谱共享策略进行了研究,考虑各非授权用户业务量和信道质量对系统频谱效率的影响,提出了一种非实时业务专用的NRT-dedicate频谱共享算法。仿真结果表明,在非实时业务情况下,该算法可以获得比通用业务的频谱共享算法的频谱利用率。     

EPON中节能动态带宽分配算法

结合光网络单元(ONU)休眠控制和动态带宽分配机制,提出一种节能的双向DBA算法。该算法根据上/下行业务量为每个ONU分配合适的上/下行授权带宽,保证每个ONU在上/下行方向都具有较低的传输时延,同时尽可能增加ONU的休眠时间来降低网络能耗。仿真结果表明,与已有算法相比,该算法具有较低的上/下行平均时延,并能减少ONU的能耗。     

Sleep assistive dynamic bandwidth assignment scheme for passive optical network (PON)

In passive optical network (PON), in addition to efficient bandwidth management, a dynamic bandwidth assignment (DBA) scheme can also enhance the energy efficiency performance of the optical networks units (ONUs) during sleep mode. A few such green DBA schemes have been proposed in literature for EPON, however, ITU compliant PONs have not got attention. In this study, the role of a DBA scheme during the cyclic sleep mode for XGPON has been investigated. A sleep assistive (SA)-DBA scheme is proposed that not only improves the energy saving performance of cyclic sleep mode but also reduces the upstream delays and variance for all the type-2 (T2), type-3 (T3) and type-4 (T4) traffic classes. Although, the upstream delay of type-1 (T1) traffic class slightly increases, the average upstream delay of all the traffic classes remains below the set target delay limit of 56 ms.     

Sleep mode operation for IEEE 802.16e UGS based WiMAX

To solve the problem of energy efficiency drop in wireless metropolitan area network communication due to frequently sleep-awake transitions, the energy-saving mechanism of the sleep mode operation was researched in WMAN. In this paper we propose a dynamic algorithm to tune the ratio of the sleep windows and receive windows according to the real time load. Then, a Markov chain model was set up to analyze the energy efficiency and mean access delay. NS2 simulation results show that the proposed algorithm can achieve marked gain in energy efficiency compared to the traditional energy saving mechanism.     

Statistical leakage current reduction in high-leakage environments using locality of block activation in time domain

This paper describes a new leakage current reduction methodology that can give a statistical leakage current reduction even if the chip is in active mode, as well as in sleep mode. The proposed scheme utilizes a time locality of activation probability of a given circuit block like cache memory characteristics. The leakage cut-off switch is operated by a self-timed sleep timer, which puts the block into sleep mode. By waiting for a certain number of cycles before entering sleep mode, power overhead associated with the sleep and wake-up process is optimized, and its conditional probability is also analyzed. The effectiveness of the proposed scheme is verified by an 8-bit RISC microprocessor using Verilog HDL with real firmware, and demonstrated by a 64-bit carry-look-ahead adder with the self-cut-off switch fabricated with dual-threshold voltage SOI technology. The criterion of the effectiveness of the proposed scheme is also discussed.     

Design of Predictive and Dynamic Energy-Efficient Mechanisms for IEEE 802.16e

Three predictive and dynamic sleep time planning (PDSTP) energy-efficient mechanisms are proposed in this paper to simultaneously improve energy efficiency and packet delay for IEEE 802.16e. To estimate the time instant when a mobile station (MS) should wake up for receiving downlink packets, a prediction method is designed. With the predicted time instant, an MS is then allowed to sleep as much as possible using multiple maximum sleep intervals followed by a smaller sleep interval before the predicted time instant. After the predicted time instant, a few smaller sleep intervals with a trend of constant level (CL), exponential decrease (ED), or linear decrease (LD) can be further arranged. To react to the outlier of prediction, exponential increase for sleep intervals can be extended. The combination of the aforementioned designs then forms our three proposed mechanisms, namely, PDSTP-CL, PDSTP-ED, and PDSTP-LD. Via simulations, we show that PDSTP-CL not only performs better than PDSTP-ED and PDSTP-LD under general situations but also outperforms the standard sleep mode operation of the type-I power saving class (PSC-I) in IEEE 802.16e and the exponential sleep time backoff mechanism (ESTBM) in the literature in terms of energy efficiency and packet delay.     

一种能效优先的认知无线电模仿主用户攻击防御策略设计与性能分析

研究了一种能效优先的认知无线电PUEA防御策略并对其进行了性能分析.以PUEA存在时CSS能效作为目标函数,以CSS安全检测性能作为约束条件,从而构建一个安全的CSS系统模型进而求解该优化问题,实现保障能量有效性的安全频谱检测.在CSS中,融合中心以K秩准则(K-out-of-N准则)作为全局判决准则,保证能效最大化时得到最佳门限K和最佳认知用户数N.通过与最大比合并(MRC)准则下的安全CSS进行比较,所提PUEA防御策略在保证系统能效的同时提高了系统的检测性能,即所提PUEA防御策略可以有效折中能效与检测性能.数值仿真结果表明,所提策略不受PUE干扰功率影响,对于PUEA攻击具有较高的顽健性.     

Novel Automotive High Intensity Discharge(HID) Lamps with Electronic Ballasts

A new scheme of automotive high intensity discharge(HID) lamps with electronic ballasts is proposed. The design of the proposed ballast and some experimental results are presented. The proposed scheme is consisted of the high frequency DC-DC converter and the low frequency DC-AC inverter. This system separates the input voltage of the ignitor from DC link voltage using auxiliary winding, then it could use the lower voltage rating power devices for HID lamp ballast system and reduce the size of HID lamp ballast. The proposed ballast controller using micro-controller unit(MCU) controls the frequency to operate the DC-DC converter in critical conduction mode, which reduces the noise of the circuit and improves the efficiency by 2%~4%.     

基于EDA的机会监听中继选择的改进机制

机会监听中继选择机制结合了中继间监听与机会选择,可以充分利用重传机制进一步降低误包率.然而其忽视监听信道和低适应性的不足,使得该机制在某些场景中不能发挥长处.因此,提出了基于分布估计算法的改进方案,使得改进后的机制不仅考虑了监听信道对于中继选择的影响,还具备了不断自我优化的学习性能.通过仿真验证后表明,提出的方案可以进一步提升系统的性能,还具有很强的自适应能力.     

单级与双级Boost变换器微能量收集电路的对比研究

随着物联网技术和无线传感技术的快速发展,低功耗设备的自供能技术得到了广泛研究和关注。针对传统微能量收集系统存在的电路结构复杂、效率低、电压传输比受限等问题,提出了双级Boost变换器微能量存储方案,并与传统单级Boost升压方案进行了对比分析。通过Matlab仿真进行理论验证,结果表明,双级Boost变换器微能量收集方式极大地提高了同等电压高变比传输下的效率,使设备的组网供能更易于实现,为低电压微能量收集系统的设计提供了一种新方案。