A distributed algorithm called P-REMiT is proposed for building an energy-efficient multicast tree in ad hoc networks. The P-REMiT uses the probability method to balance the Total Energy Consumption (TEC) and System Lifetime (SL) of multicast tree. It gets the better performance than SREMiT on metrics about system life and also obtains the better performance than L-REMiT on metrics about TEC. It improves SL of multicast tree efficiently with little sacrifice on TEC and has good convergence.

IP anycast allows a source node to transmit IP packets to a single destination node, out of a group of destination nodes. It can be an important paradigm for an ad-hoc network in terms of resource, robustness and efficiency for replicated service applications. DSR is a simple and efficient routing protocol designed specifically for use in multi-hop wireless ad hoc networks of mobile nodes. This paper proposes a novel anycast protocol for Ipv6 flow in mobile ad hoc networks, which is based on the DSR protocol. We also test its performance with different network parameters and the simulation results show that the anycast protocol can balance the network load efficiently and reduce the delay of packet and improve the network throughput. Anycast service can also improve performance of ad hoc network when mobility is high and a link may get disconnected frequently without the servers of repair/re-discovery due to ad-hoc network environment.

In this paper, we propose a novel active queue management algorithm PFED, which is based on network traffic prediction. The main properties of PFED are: (1) stabilizing queue length at a desirable level with consideration of future traffic, and using a MMSE (Minimum Mean Square Error) predictor to predict future network traffic; (2) imposing effective punishment upon misbehaving flow with a full stateless method; (3) maintaining queue arrival rate bounded by queue service rate through more reasonable calculation of packet drop probability. To verify the performance of PFED, PFED is implemented in NS2 and is compared with RED and CHOKe with respect to different performance metrics. Simulation results show that PFED outperforms RED and CHOKe in stabling instantaneous queue length and in fairness. It is also shown that PFED enables the link capacity to be fully utilized by stabilizing the queue length at a desirable level, while not incurring excessive packet loss ratio.

This paper proposes a new architecture for web-based virtual laboratory with CORBA technology. In the architecture, the Java Applet acts as the client tool and CORBA acts as the communication bridge between different objects. The integration of JavaBean, Matlab and COM/DCOM is implemented as computing tools on the server side. By using the architecture, the efficiency of developing virtual laboratories can be improved remarkably. This paper also introduces the implementation of virtual laboratory for communication principle based on the architecture. In the virtual laboratory, the instruments are developed as components which improves the development efficiency and reuse of the components. Matlab is also used for computing in the server, which enhances the simulation capability.

The mechanism of forwarding request plays the most important role in Crowds anonymous communication protocol. On one hand, it hides the identity of the request initiator against the responder, the participants in protocol, and eavesdroppers. On the other hand, it causes additional latency on communication and payload on participants in the protocol. In this paper, we investigate the influence of the forwarding mechanism with respect to the performance and the security in Crowds. Different from the previous approaches, our analysis focuses on the length of forwarding paths, and is independent of the underlying length control strategy. In the study of system performance, we consider the participant payload in Crowds and prove that the expected participant payload is equal to the expected length of forwarding paths. Applying this result to the currently used length control strategy in Crowds, we derive that the expected participant payload in Crowds is 1/(1-Pf)+1, where Pf is the forwarding probability in Crowds. This improves Reiter and Rubi's original result and demonstrates that the participant payload in Crowds is entirely independent of the size of Crowds protocol. Thus, Crowds has a very nice scalability property. In the study of security, we investigate the number of rounds of attack to identify the request initiator with high probability, and show that it is not affected significantly by the maximum length of forwarding paths. Therefore, in order to improve the system performance (such as participant payload and communication latency) without significantly affecting the system security, length control strategies that shorten the forwarding paths can be considered.

This paper develops an effective randomized on-demand QoS routing algorithm on networks with inaccurate link-state information. Several new techniques are proposed in the algorithm. First, the maximum safety rate and the minimum delay for each node in the network are pre-computed, which simplify the network complexity and provide the routing process with useful information. The routing process is dynamically directed by the safety rate and delay of the partial routing path developed so far and by the maximum safety rate and the minimurn delay of the next node. Randomness is used at the link level and depends dynamically on the routing configuration. This provides great flexibility for the routing process, prevents the routing process from overusing certain fixed routing paths, and adequately balances the safety rate and delay of the routing path. A network testing environment has been established and five parameters are introduced to measure the performance of QoS routing algorithm.q. Experimental results demonstrate that in terms of the proposed parameters, the algorithm outperforms existing QoS algorithms appearing in the literature.

Crowds is a recently proposed protocol for anonymous communication, which is based on the idea of ”blending into a crowd". Thus, each participant of the crowd gets the benefits of anonymity but also serves as a proxy for other participants. An important measure of the protocol is the participant payload in the system, which is measured by the amount of work a participant needs to pay on serving as a proxy for any communication requested by the participants in the system. In this paper, we derive a precise formula for the participant payload in Crowds, which improves the previous results. Moreover, our result shows the first time that the participant payload in Crowds is entirely independent of the size of the crowd. In consequence, Crowds protocol has a very nice scalability property.

提出了一种用于网络测量的动态时间驱动分组采样技术，根据网络流量的变化动态调整采样次数，从而能将最后的测量误差率控制在一个给定的范围内，在保证精度的同时极大地降低了测量的代价。并且该方法的实现复杂度与“1 out of N”采样技术近似，文章中给出了有关的理论分析和实验研究。实验结果表明该方法达到了设计的目标。根据网络运行状态动态调整网络测量方案的思想可以广泛用于网络测量当中。

及时传输有服务质量（QoS, Quality of Service）要求的多媒体信息对于集成服务的宽带网络发展提出了新的挑战。QoS 路由算法的优劣直接影响网络服务质量，而由于链路信息的不及时更新必将造成网络链路信息的非精确性。本文提出了一种在非精确网络状态信息下基于改进的Bellman-Ford算法的随机QoS路由算法Random_link_S0N0，该算法采用随机策略来平衡网络负载和改善网络性能。通过网络模拟器的测试，该算法具有良好的性能指标，同时减少了处理和协议的额外开销。

提出了一种以组件技术构架的基于Internet的数字图像处理仿真系统模型。该模型采用基于Java的组件对象技术，提高了开发效率，实现了软件重用，使得系统易于维护和扩充，并通过CORBA技术来实现Java与Matlab之间的无缝联接，大大提高了系统的仿真能力。该系统客户端用Java语言实现，具有与平台无关，安全，健壮等内在特点，能够实现仿真流程的动态定制、仿真试验的保存，以及用户添加新的算法组件等功能。服务器端以Matlab为计算后台，提高了算法组件开发的效率和正确性。