Special Session on Grid and System Software -- TeraGrid: A Foundation for US Cyberinfrastructure -- Globus Toolkit Version 4: Software for Service-Oriented Systems -- System Software for China National Grid -- Session 1: Grid Computing -- CGSV: An Adaptable Stream-Integrated Grid Monitoring System -- Performance Modeling and Analysis for Resource Scheduling in Data Grids -- Study on ?-Calculus Based Equipment Grid Service Chain Model -- A Performance-Based Parallel Loop Self-scheduling on Grid Computing Environments -- A Resource-Based Server Performance Control for Grid Computing Systems -- IBP: An Index-Based XML Parser Model -- A Stochastic Control Model for Hierarchical Grid Service -- Service-Based Grid Resource Monitoring with Common Information Model -- Distributed Gridflow Model and Implementation -- Session 2: Peer-to-Peer Computing -- A Secure P2P Video Conference System for Enterprise Environments -- Adaptive Query-Caching in Peer-to-Peer Systems -- Design and Deployment of Locality-Aware Overlay Multicast Protocol for Live Streaming Services -- Session 3: Web Techniques -- Dynamic Thread Management in Kernel Pipeline Web Server -- QoS Aware Service Composition with Multiple Quality Constraints -- Session 4: Cluster Computing -- Performance Modelling and Optimization of Memory Access on Cellular Computer Architecture Cyclops64 -- TCP-ABC: From Multiple TCP Connections to Atomic Broadcasting -- A Parallel File System Based on Spatial Information Object -- Topology-Aware Multi-cluster Architecture Based on Efficient Index Techniques -- A Parallel Routing Algorithm on Circulant Networks Employing the Hamiltonian Circuit Latin Square -- An Efficient Load Balancing Algorithm for Cluster System -- Session 5: Parallel Programming and Environment -- A Greedy Algorithm for Capacity-Constrained Surrogate Placement in CDNs -- An Improved Scheme of Wavelength Assignment for Parallel FFT Communication Pattern on a Class of Regular Optical Networks -- A Parallel O(n27n/8) Time-Memory-Processor Tradeoff for Knapsack-Like Problems -- Improving Parallelism of Nested Loops with Non-uniform Dependences -- A Static Data Dependence Analysis Approach for Software Pipelining -- A Dynamic Data Dependence Analysis Approach for Software Pipelining -- A Parallel and Distributed Method for Computing High Dimensional MOLAP -- An Improved ACO Algorithm for Multicast Routing -- Performance Modelling of Pipelined Circuit Switching in Torus with Hot Spot Traffic -- An Incremental Compilation Approach for OpenMP Applications -- Enhanced Congestion Control Algorithm for High-Speed TCP -- Advanced Software On-Demand Based on Functional Streaming -- Can Out-of-Order Instruction Execution in Multiprocessors Be Made Sequentially Consistent? -- Efficiently Passive Monitoring Flow Bandwidth -- A Heuristic for Scheduling Parallel Programs with Synchronous Communication Model in the Network Computing Environments -- A Formal Model for Network Processor Workload -- Coping with Data Dependencies of Multi-dimensional Array References -- Session 6: Network Architecture -- QoS-Based Dynamic Channel Allocation for GSM/GPRS Networks -- Distributed Active Measuring Link Bandwidth in IP Networks -- Preferential Bandwidth Allocation for Short Flows with Active Queue Management -- A New Self-tuning Active Queue Management Algorithm Based on Adaptive Control -- Research on Multi-agent System Automated Negotiation Theory and Model -- Adaptive Congestion Control in ATM Networks -- Session 7: Network Security -- Secure Password Pocket for Distributed Web Services -- The Modified DTW Method for On-Line Automatic Signature Verification -- A Secure On-Demand Routing with Distributed Authentication for Trust-Based Ad Hoc Networks -- Probabilistic Packet Filtering Model to Protect Web Server from DDoS Attacks -- An Identity Authentication Protocol for Acknowledgment in IEEE 802.15.4 Network -- A Design of the Digital Content Distribution System Based on the Public Key and the Hierarchical Web Caching Structure -- Session 8: Network Storage -- Cluster-Aware Cache for Network Attached Storage -- Design and Implementation of a SAN Agent for Windows NT Architecture -- MagicStore: A New Out-of-Band Virtualization System in SAN Environments -- A Content Delivery Accelerator in Data-Intensive Servers -- A Systematic Scheme to Resolve QoS Dissatisfaction for Storage Cluster -- Secure Anonymous Communication with Conditional Traceability -- Session 9: Multimedia Service -- Real-Time Video over Programmable Networked Devices -- A New Raid-Disk Placement Method for Interactive Media Server with an Accurate Bit Count Control -- A New Region of Interest Image Coding for Narrowband Network: Partial Bitplane Alternating Shift -- Using Route Probing to Derive Link Traffic Load with Edge-Based Measurements -- Scheduling Multicast Traffic in a Combined Input Separate Output Queued Switch -- A QoS-Based Scheduling Mechanism for Overlay Aggregate Traffics -- Session 10: Ubiquitous Computing -- Energy Conservation by Peer-to-Peer Relaying in Quasi-Ad Hoc Networks -- Developing Energy-Efficient Topologies and Routing for Wireless Sensor Networks -- The Efficient Transmission Scheme in Wireless Crypto Communication -- Constructing k-Connected k-Cover Set in Wireless Sensor Networks Based on Self-pruning -- GCMPR: Gateway-Centric Multi-path Routing for Internet Connectivity of Wireless Mobile Ad Hoc Network -- A Semantic and Adaptive Context Model for Ubiquitous Computing -- Research of Survival-Time-Based Dynamic Adaptive Replica Allocation Algorithm in Mobile Ad Hoc Networks.These proceedings contain the papers presented at the 2005 IFIP International Conference on Network and Parallel Computing (NPC 2005), held in Beijing, China, between November 30 and December 3, 2005. The goal of the conference was to establish an international forum for engineers and scientists to present their ideas and experiences in network and parallel computing. A total of 320 submissions were received in response to our Call for Papers. These papers were from the following countries or regions: Australia, Canada, China, France, Germany, Hong Kong, India, Iran, Italy, Japan, Korea, Lux- burg, Nepal, Netherlands, Taiwan, United Arab Emirates, and United States. Each submission was sent to at least three reviewers.Each paper was judged - cording to its originality, innovation, readability, and relevance to the expected audience.Basedonthereviewsreceived,atotalof68paperswereretainedfor- clusionin theproceedings.Among the 68 papers,48 wereacceptedasfull papers for presentation at the conference. We also accepted 20 papers as short papers for a possible brief presentation at the conference, followed by discussion during a poster session. Thus, only 21% of the total submissions could be included in the ?nal program.