Baoping Yan

An application layer multicast approach based on topology-aware clustering

Application layer multicast accelerates ubiquitous deployment of the multicast communication, as well as brings unavoidable performance penalties because the group members are dynamic and lack direct knowledge about the underlying topology. A promising approach to improve the multicasting performance is to cluster the nearby nodes in groups. However, it confronts some practical challenges. A challenge is that it is difficult to assign a proximity bound to determine whether some nodes should be clustered or not. Another practical problem is how to organize the corresponding multicast structure. In this paper, we propose a new topology-aware hierarchical clustering model, which implements clustering in different grain sizes. Based on the model, we propose an application layer multicast solution named HCcast, especially for large-scale group applications. HCcast employs a topology-aware approach to choose candidate parents at different levels, and uses distributed depth first searching (DFS) approach to position a host at the same level. The clusters of HCcast are topology-based, therefore cluster split and merge operations are unnecessary, which reduces the maintenance overhead. The results of our simulation experiments show that HCcast can build multicast trees with desirable delivery performance, and the performance keep stable in different join sequences.

Fast Inter-MAP Handover in HMIPv6

A scheme that supports a fast handover effectively in hierarchical mobile IPv6 networks (HMIPv6) is presented. In HMIPv6 when a mobile node (MN) moves from a MAP domain to another, it can experience disruption of connection as well as packet loss due to a long handover latency. To cope with these problems, the fast handover scheme from FMIPv6 is adopted to optimize the performance of the inter-MAP handover. The handover latency for various protocols is compared to the proposed scheme with analytical model. The results show that the proposed scheme is better approach to handle the mobile node’s movement than the other protocols.

An Enhanced Service Differentiation Mechanism for QoS Provisioning in IEEE 802.11e Wireless Networks

There are a number of works trying to adjust the contention window (CW) in order to provide differentiated service in IEEE 802.11-based wireless networks. However, these protocols cannot deliver firm service guarantees while maintaining high network utilization, particularly under congested network conditions. In this paper, we propose an adaptive enhanced service differentiation mechanism for QoS provisioning in IEEE 802.11e wireless networks. The mechanism includes traffic conditioning and adaptive MID CW updating rule with service differentiation based on per queue. Relative priorities are provisioned by adjusting the size of the CW of each traffic class taking into account both applications requirements and network conditions. Simulation results show that the proposed mechanism outperforms the basic EDCA scheme of 802.11e and another proposed scheme AEDCF in terms of network utilization, goodput and delay.

An IPv6 DNS Conformance Testing Framework

The domain name system (DNS) is one of the most basic Internet infrastructures. Therefore, the IPv6 conformance testing is very necessary for accelerating the deployment of IPv6. This paper proposes a conformance testing framework for IPv6 DNS, in terms of the features of IPv6 DNS specifications. This framework employs the extended finite state machine (EFSM) model to describe the control and data flows of dynamic behaviors of IPv6 DNS, and configures the relevant variables and data files by a configuration module which can reduce the test cost.

A Delivery Tree Building Approach for Application Layer Multicast

Many proposed solutions employ distributed depth-first-searching (DFS) approach to implement positioning in application layer multicast. In DFS-based delivery tree building approaches, the newcomer need not know the information of other members except the root, and positioning can be easily implemented. However, the DFS approach cannot effectively position a host in some cases because it usually selects a branch of active candidate parent to continue the searching procedure. This paper proposes a closest-first-searching (CFS) positioning means, and a CFS-based tree building approach name CFTB. In CFTB, the new active candidate parent is selected among all the candidate parents in terms of the known distance information. The simulation results show that CFTB can position a newcomer accurately, and build a multicast tree with good delivery performance.

Network-Processor-Based IPv4/IPv6 Translator: Implementation and Fault Tolarence

IPv6 is a new version protocol for next generation Internet, which has the advantages of supporting scalability, mobility and security better than current IPv4 Internet. However, IPv4 and IPv6 are not directly compatible. IPv4/IPv6 translator with high performance and reliability can facilitate the seamless coexistence of IPv4 and IPv6 networks during the transition period. In this paper, we proposed a design and implementation of IPv4/IPv6 translator using network processors. The challenges of this work lie in three areas: (1) effective use of network processor resources, (2) design and implementation of an advanced control plane on a commodity OS, and (3) fault tolerance to improve reliability of the whole system. We discuss our solutions to these challenges and presented analytic models to investigate the fault tolerance issue of our approaches. Our work provided guidelines and insights into design, implementation and configuration of the IPv4/IPv6 translator.

The Application of the Sensor Model Language in Sensor Observation Service of the HeiHe Experimental Research

The Heihe Watershed Allied Telemetry Experimental Research (HiWATER) is a significant research project that performs airborne, satellite-borne and ground based remote sensing experiments at various scales by coordinating interdisciplinary researchers from different institutions and projects. Therefore, in order to discover, access and process sensors within the heterogeneous systems, we use the Sensor Model Language (Sensor ML) to model these sensors and sensor systems in the stations, and use Sensor Observation Service (SOS) to display the sensor information and observation data information. In this paper, we firstly introduce the Sensor ML and SOS, then mainly present a model of sensor station and the application of this model in SOS.

Trustworthy communication mechanism based on IPv6 real-name address resource PKI

Identity management and secure end to end communication are the core topics of trustworthy network research. This paper proposes a hierarchical IPv6 real-name address space as network identity. We also built an IPv6 real-name address resource PKI according to current IP address allocation structure for real-name address allocation, management and authentication work. We designed and implemented a trustworthy communication system supported by third partys IPv6 real-name address resource PKI. This method can overcome the inherent defects of IPsec that may be circumvented in the identifier conversion process and mobile environment. It also provides end to end authentication service on the Internet. We believe this is an effective solution for Internet real-name access and privacy protection problem.

On Improving the Multicast Performance of Application Layer Multicast

In the group application, poor multicast trees might be built in some join sequences. Additionally, group members are dynamic, which sometimes brings negative influence on the multicast tree. This paper analyzes the poor shapes of multicast trees, and proposes an additional improvement scheme to enhance the multicast performance. The proposed scheme rearranges the locations of members in terms of a distance-heuristics, which makes the improved multicast tree be coincident with the underlying network topology in some degree. The simulation results show that our proposed improvement scheme can effectively enhance the multicast performance.

A New IPv6 Tunneling Protocol: Escort

Teredo is a service that enables nodes located behind one or more IPv4 Network Address Translations (NATs) to obtain IPv6 connectivity by tunneling packets over IPv4 UDP. Although Teredo is a useful protocol, it also has some weaknesses. It can not work with symmetric NAT, and it raises some security concerns, such as bypassing security controls, reducing defense in depth, and allowing unsolicited traffic. In this paper, we propose a new IPv6-in-IPv4-UDP tunneling protocol—Escort, which adopts ID/Locator split thought, and we explain how an Escort client configures its Escort host identity address, learns its global Escort IPv6 address and obtains IPv6 connectivity with the help of Escort servers and Escort relays. We also describe in details our strategies for implementing Escort server and Escort relay under linux, and make a performance measurement of our Escort relay implementation and compare the packet processing latency of the Escort relay with that of Miredo-relay.