DAG任务同步中无锁机制实现方法研究

随着多核嵌入式实时系统的发展,DAG任务同步问题得到了广泛的关注.目前的任务同步方法大都采用锁机制,但锁机制存在许多问题,如自旋锁存在任务忙等状态,浪费CPU资源;使用互斥锁的任务若获取不到共享资源会被阻塞,产生上下文切换开销;顺序锁允许写任务有更高的优先级,但写任务不能频繁更新数据,否则读任务会产生饿死现象.上述锁机制如果应用于多核平台下的DAG任务同步,不仅会影响系统整体执行效率,导致后继任务无法执行,严重时会引发死锁现象导致系统崩溃.因此,提出了在DAG任务同步过程中使用DCAS无锁机制,有效避免了锁机制存在的问题.在LITMUSRT多核平台下,以多任务同时申请、填充和释放Vxworks网络缓冲区为例,对缓冲池中的三元组m Blk,clBlk,cluster分别使用DCAS无锁机制.实验结果表明,相比传统锁机制,DCAS无锁机制在DAG任务同步方面有较好的效果,响应时间减少了10.4%,系统的整体执行效率提高了4.2%.     

DSSS系统中伪码防错锁技术分析

直接序列扩展频谱技术正被越来越多地应用于各种测控系统中。讨论了发生伪码错锁原因,并在此基础上针对伪码的自相关特性和互相关特性,分析了产生伪码错锁可能产生的影响。根据伪码错锁的原因以及伪码的捕获过程提出了3种伪码错锁的措施,包括选用改进伪码的自相关特性、互相关特性、功率控制技术以及改进伪码的捕获跟踪策略,并对3种伪码错锁措施进行了详细论述。     

关于JVSTM无锁提交算法的研究

软件事务存储（software Transaction Memory）思想提出的初衷是用来作为并发控制的一个无锁机制。由于早期的STM实施面临着效率的诸多限制,并且不久之后一种无阻塞的思想出现了,它能够有效解决效率问题,简化STM的实施,因此,现在大多数活跃的STM都是采用了阻塞的设计方法,利用锁机制来保证事务提交操作的原子性。该方法在实际应用中有着更加优越的性能表现,当然部分由于它更加简单。然而当我们将阻塞方法应用到多核系统中,特别当事务写操作频繁的时候,该方法将会暴露出可拓展方面的天生缺陷。     

光纤激光器锁模技术研究进展

锁模光纤激光器能够产生超短脉冲,在众多领域中有着广泛的应用前景。概述了光纤激光器锁模实现技术的基本原理、典型结构和最新研究进展,分析了主动锁模、被动锁模和混合锁模技术的优缺点,并对多波长锁模、拉伸脉冲锁模和基于光子晶体的锁模光纤激光器的最新研究情况进行了介绍和分析。展望了锁模光纤激光器研究的发展方向和急需解决的问题。     

光纤激光器技术

在概述光纤激光器基本原理、主要特性及其多种分类方法的基础上,首先介绍了双包层光纤激光器的工作原理和最新进展。通过对几种典型实验装置的分析阐述了光纤激光器的多波长产生机制。并详细讨论了光纤激光器产生锁模脉冲的三种技术：采用振幅或相位调制器的主动锁模技术、利用非线性效应的被动锁模技术以及混合锁模技术,最后比较了这三种技术和实验装置的优缺点并介绍了它们的最新发展情况。     

基于多级电源系统中闩锁的研究

对在多级电源系统可靠性设计中遇到的几个关于闩锁的问题进行了研究,针对闩锁产生的原因进行分析,并提出解决方法。     

机抖陀螺锁区补偿的角度随机游走改善分析EI北大核心CSCD

二频机抖陀螺每个抖动周期要两次经过锁区,每次过锁区时的随机误差会使激光陀螺产生随机游走。在工程上实现了二频机抖陀螺的锁区补偿,并采用Allan方差方法分析了锁区补偿前后输出数据的角度随机游走,实验结果表明,锁区补偿后随机游走具有大幅度的改善。首次报道了机抖激光陀螺中锁区补偿对角度随机游走的改善。     

自锁模Ti:Al2O3激光器自启动条件分析

本文分析了Ti:Al2O3激光器的自启动条件,讨论了不同腔参数对锁模产生的影响,计算了自锁模及自启动的调节范围,并在实验中实现了自启动锁模,计算结果与实验相吻合.     

基于图论理论解决关系数据库死锁方法的研究

根据关系数据库的基本理论，关系型数据库对数据的存储使用了标准的共享锁、修改锁、独占锁等机制，提供数据库、表、页面、记录等多种不同的资源粒度以及锁升级机制。因此所有的关系型数据库，都存在着并发访问的锁资源争夺风险，如何能在确保数据一致性的前提下避免因锁资源争夺而导致数据库系统性能下降甚至崩溃，一直是数据库系统及其相关应用等领域内关注的焦点。文章将基于图论理论的基础，分析和探讨解决关系数据库死锁的思路和方法，通过构造一些基础的数学模型来分解和本质化死锁现象，提出一些解决方法和理论思路。     

光纤锁模激光器：从单模时域耗散孤子锁模到多模时空耗散孤子锁模（特邀）

在光纤锁模激光器中,模式相位锁定产生周期短脉冲的过程称为锁模过程,产生的脉冲在广义上被称为“光耗散孤子”。光纤锁模激光器从传统的单模光纤锁模激光器发展到了多模光纤锁模激光器,锁模机理从一维（1D）时域耗散孤子锁模发展到了（3+1）维时空耗散孤子锁模。通过深入理解耗散孤子的产生机理,有望进一步推动光纤锁模激光器在科学和应用领域的发展,为更多领域带来更多创新和可能性。首先介绍单模光纤锁模激光器中的一维时域耗散孤子锁模,探讨不同色散区域中时域耗散孤子的产生机理;随后介绍多模光纤锁模激光器中时空耗散孤子的最新研究成果,讨论模间色散的补偿方法,揭示其丰富的时空锁模机理和潜在的应用场景;最后对光纤锁模激光器的发展前景进行展望。     

Partial CDR database dimensioning

Novel services in the mobile telecommunication era are far more longer then a few years ago when the price of the call was calculated purely on the length of the session. To overcome possible fraud issues, the corresponding standards are defining partial CDRs that are sent to the billing system periodically during the sessions. These partial CDRs has to be analyzed and stored in a database for correlation. The size of this database, and thus the required processing time is heavily depending on the distributions of the calls. In this paper we gave a method to calculate the expected number of partial CDRs and the required database size if these distributions are at least partially known.     

A fractals-inspired approach to binary image database indexing and retrieval

This paper applies ideas from fractal compression and optimization theory to attack the problem of efficient content-based image indexing and retrieval. Similarity of images is measured by block matching after optimal (geometric, photometric, etc.) transformation. Such block matching which, by definition, consists of localized optimization, is further governed by a global dynamic programming technique (Viterbi algorithm) that ensures continuity and coherence of the localized block matching results. Thus, the overall optimal transformation relating two images is determined by a combination of local block-transformation operations subject to a regularization constraint. Experimental results on some limited subsets of still binary images from the mpeg-7 database demonstrate the power and potential of the proposed approach.     

Provable security against a differential attack

The purpose of this paper is to show that DES-like iterated ciphers that are provably resistant against differential attacks exist. The main result on the security of a DES-like cipher with independent round keys is Theorem 1, which gives an upper bound to the probability of s-round differentials, as defined in [4], and this upper bound depends only on the round function of the iterated cipher. Moreover, it is shown that functions exist such that the probabilities of differentials are less than or equal to 2^3–n , where n is the length of the plaintext block. We also show a prototype of an iterated block cipher, which is compatible with DES and has proven security against differential attack.A preliminary version of this paper was presented in the rump session at Crypto '92. The work of Kaisa Nyberg on this project was supported by MATINE Board, Finland.     

Session integration service over multiple devices

One of the most important research subjects in session initiation protocol (SIP) is mobility management. Session mobility is a key issue in SIP mobility management. Session mobility is an advanced feature that maintains an ongoing media session from one device to another. In this paper, we consider the issue of session mobility and propose a complete integrated solution, referred to as ‘session integration service’, to transfer an ongoing media session over multiple devices. To provide flexibility in session mobility, three new agent states are introduced based on the user agent definition. They consist of session manager, session user, and free node. This paper also proposes two split session integration methods for session integration service based on these three agents. The proposed method provides session mobility flexibility and allows the user the ability to transfer, split, and retrieve a session over multiple devices. Moreover, the split session can be integrated in any devices. To ensure session mobility over multiple devices, the session integration service will be implemented through a modification of the open source project ‘SIP‐communicator’. Copyright © 2010 John Wiley & Sons, Ltd.     

Dynamic Multicast Session Provisioning in WDM Optical Networks with Sparse Splitting Capability

In this work, we study dynamic provisioning of multicast sessions in a wavelength-routed sparse splitting capable WDM network with an arbitrary mesh topology where the network consists of nodes with full, partial, or no wavelength conversion capabilities and a node can be a tap-and-continue (TaC) node or a splitting and delivery (SaD) node. The objectives are to minimize the network resources in terms of wavelength-links used by each session and to reduce the multicast session blocking probability. The problem is to route the multicast session from each source to the members of every multicast session, and to assign an appropriate wavelength to each link used by the session. We propose an efficient online algorithm for dynamic multicast session provisioning. To evaluate the proposed algorithm, we apply the integer linear programming (ILP) optimization tool on a per multicast session basis to solve off-line the optimal routing and wavelength assignment given a multicast session and the current network topology as well as its residual network resource information. We formulate the per session multicast routing and wavelength assignment problem as an ILP. With this ILP formulation, the multicast session blocking probability or success probability can then be estimated based on solving a series of ILPs off-line. We have evaluated the effectiveness of the proposed online algorithm via simulation in terms of session blocking probability and network resources used by a session. Simulation results indicate that our proposed computationally efficient online algorithm performs well even when a fraction of the nodes are SaD nodes.     

Optimization of SIP Session Setup Delay for VoIP in 3G Wireless Networks

Wireless networks beyond 2G aim at supporting real-time applications such as VoIP. Before a user can start a VoIP session, the end-user terminal has to establish the session using signaling protocols such as H.323 and session initiation protocol (SIP) in order to negotiate media parameters. The time interval to perform the session setup is called the session setup time. It can be affected by the quality of the wireless link, measured in terms of frame error rate (FER), which can result in retransmissions of packets lost and can lengthen the session setup time. Therefore, such protocols should have a session setup time optimized against loss. One way to do so is by choosing the appropriate retransmission timer and the underlying protocols. In this paper, we focus on SIP session setup delay and propose optimizing it using an adaptive retransmission timer. We also evaluate SIP session setup performances with various underlying protocols (transport control protocol (TCP), user datagram protocol (UDP), radio link protocols (RLPs)) as a function of the FER. For 19.2 Kbps channel, the SIP session setup time can be up to 6.12s with UDP and 7s with TCP when the FER is up to 10 percent. The use of RLP (1, 2, 3) and RLP (1, 1, 1, 1, 1, 1) puts the session setup time down to 3.4s under UDP and 4s under TCP for the same FER and the same channel bandwidth. We also compare SIP and H.323 performances using an adaptive retransmission timer: SIP outperforms H.323, especially for a FER higher than 2 percent.     

Strategies for management of connection oriented data sessionsemploying the idle times in cellular telephony

We consider strategies for and performance of a class of connection-oriented data session services that might be supported using only the idle times between telephone calls on the channels of a cellular radio system. Two types of service are envisaged: a short transaction service and a suspendable data session service. We give expressions for the mean active and inactive times for data sessions and show how session assignment and displacement strategies can shape the distributions of session active and inactive times. We find useful amounts of capacity in the background of telephony systems even at high load, but in its raw form the opportunities for data transmission, and the data session suspension times implied, are rather large-grained for interactive data sessions. However, enforced time-sharing can improve these statistics for interactive data uses. With prioritized session assignment, a number of logical data sessions can be supported providing a range of service characteristics onto which interactive users, mobile fax, paging, navigation, and fleet tracking applications can be mapped in decreasing order of required session availability     

Designing VoIP session management over interworked WLAN-3G networks

In an all-IP internetworked heterogeneous environment, ongoing VoIP sessions from roaming users will be subject to frequent vertical handoffs across network boundaries. Ensuring uninterrupted service continuity for these handoff calls requires successful session management among the participating access networks. As such, a mobility-aware novel interworking network design (interconnecting UMTS and WLAN over an IP-based common platform) [1] is presented in this article that facilitates VoIP session management, including session establishment and seamless session handoff across different networks. For comparison purposes, VoIP session management is evaluated in terms of session establishment, handoff delays, transient packet loss, end-to-end traffic delays, and jitter value for different voice codecs, which demonstrate satisfactory and feasible results. In the event (e.g., network congestion, buffer overflow) that session continuity cannot be guaranteed (also known as outage) across network boundaries, this article proposes an algorithm that compensates the user by reducing the unit service charge of future sessions (governed by the outage period) through a noncooperative game-theory-based pricing mechanism.     

CDMA2000中连接建立的分组延时分析

分组的RLP(无线链路协议)传送延时是连接建立延时的一个重要组成部分.已有的研究一般假设满负荷的条件计算该部分延时,但该假设在连接建立的过程中不成立.本文分析连接建立过程中分组的RLP传送延时,通过考察组成一个分组的所有数据帧在不同轮次的传送状态,提出了一种迭代计算方法.仿真结果表明,提出的方法可以准确地计算连接建立条件下的分组延时.满负荷条件下的延时和连接建立条件下的分组延时相比较,前者的延时可比后者的延时多达到100ms.     

DRA在VoLTE中的会话绑定策略研究

详细分析了DRA在VoLTE中的会话绑定原理，探讨了DRA在VoLTE的会话绑定中所面临的关键问题，并针对这一问题提出了相应的解决方案，最后给出了DRA在VoLTE中的会话绑定策略建议。