臺灣博碩士論文加值系統

如何建構一個符合模擬所需的大型虛擬環境，一直是研究虛擬實境學者
、專家的長期目標。由於大型虛擬環境可以用於模擬、訓練等事宜，由虛
擬實境所建構的模擬環境，最大的優點是可以減少人員傷亡及物資的耗損
，因此可大幅節省經費的支出，所以，近年來，如何透過虛擬實境與網路
技術的結合來建構一個大型的模擬環境，也逐漸受到政府與商業團體機構
的重視。對於建構一分散式的虛擬世界而言，目前所面臨的主要技術問題
是在於如何將虛擬實境與網路的技術相結合。因此我們發展一個「網路型
模擬環境基礎架構 (Networked Simulation Environment
Infrastructure」，簡稱NSEI，來解決虛擬實境與網路技術相結合的問題
。在一個網路型大型模擬環境中，由於可能有超過上萬的模擬物件
(Simulation Entities) 大量交換資料，因此在整個模擬過程中需要一個
模擬管理(Simulation Management) 系統來控管整個模擬進行，為了解決
模擬管理者的位址搜尋問題，我們在NSEI的模擬應用基礎架構
(Simulation Application Infrastructure) 中提出模擬管理者位址反查
協定 (Simulation Manager Address Resolution Protocol)，簡稱SMARP
。並且在NSEI的通訊服務基礎架構 (Communication Service
Infrastructure) 中提出傳輸服務協定 (Transmission Service
Protocol)，簡稱 TSP，來滿足NSEI中所需的通訊服務。最後，在本論文
中，我們提出一種測量軟體複雜度的方法，來確保TSP中搜尋路徑
(routing) 演算法的品質與正確性。

The construction of large scale virtual world has been a long-
stated goal of virtual environment proponents and now is a major
objective of both commercial and government organizations.
However, there exist major technical challenges that will
require new network hardware/software architectures for
distributed virtual environments. In this thesis, we have
developed a Networked Simulation Environment Infrastructure
(NSEI) for large scale distributed simulation. In a networked
virtual environment for simulation, especially those involving a
large number of interacting simulation entities, require a
simulation management to synchronize active simulation entities
and conduct exercise. In order to solve the simulation manager
addressing problem, a new protocol called "Simulation Manager
Address Resolution Protocol (SMARP)" is proposed for the
simulation entities to acquire the simulation manager address
across the network in Simulation Application Infrastructure of
NSEI. Throughout a simulation exercise, the state information
associated with the interactions that take place between
simulation entities needs to be exchanged through network. The
purpose of this thesis in Communication Service Infrastructure
of NSEI is to propose a protocol, called "Transmission Service
Protocol (TSP)," to support communication services necessary for
NSEI. Finally, a new methodology of measuring software
complexity assures the quality and reliability of TSP routing
algorithm is proposed.

COVER
Contents
1. Introduction
2. The Infrastructure of Networked Simulation Environment
2.1 Related Work
2.1.1 Based Components of DIS
2.1.2 Characteristics of DIS
2.1.3 Major Problems with the DIS Protocol
2.2 The Construction of NSEI
2.2.1 Key Concepts of NSEI
2.2.2 NSEI Architecture
2.3 Perspectives on the NSEI with OSI Model
2.3.1 Simulation Application Layer
2.3.2 Communication Service Layer
2.3.3 Network Layer
2.4 Conncetion Mode and Conncetionless Mode
2.4.1 Conncetion Mode Transmission
2.4.2 Connectionless Mode Transmission
2.5 Purpose and Scope
3. Simulation Manager Addree Resolution Protocol
3.1 Simulation Manager Address Resolution Protocol
3.1.1 The Exercise Information Server
3.1.2 Three Phases of SMARP
3.2 The Processes of SMARP
4. Transmission Service Protocol
4.1 The Transmission Service for CSI
4.2 TSP Addressing
4.2.1 TSP Address Formats
4.2.2 TSP Addresses and Simulation Entities
4.2.3 Special Purpose Addresses
4.3 PDU Encapsulation and Framing
4.3.1 TSP Header
4.3.2 Network Infrastructure Frame Creation
4.4 TSP Routing
5. The Path Complexity Metrics
5.1 Background and Meterics
5.1.1 Lines of Code
5.1.2 Halstead''s Metrics
5.1.3 McCabe''s Cyclomatic Complexity
5.1.4 Harrision and Magel''s Meterics
5.1.5 Weiser''s Data Flow based Metrics
5.1.6 Chung''s Path Complexity Hierarchy
5.2 The Path Complexity Metrics
5.2.1 Linked and Nested Types
5.3 Algorithm to Measure the Path Complexity of a Program
5.4 The Evaluation of Routin Algorithm
6. Conclusions adn Future Works
6.1 Conclusions
6.2 Future Works
Bibliography
Appendix A-1 Protocol Version Enumeration
Appendix A-2 Time Stamp
Appendix B NSEI-Based Simulator