A Distributed Management Framework for H.323-Based VoIP System

• Main Authors: Ching-Chung Huang, 黃敬中
• Other Authors: none
• Format: Others
• Language: en_US
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/49831618145116702182

碩士 === 逢甲大學 === 資訊工程所 === 92 === VoIP (Voice over Internet Protocol) has attracted tremendous attention since the late 90’s. Many call control protocols, such as H.323, SIP and MGCP, were addressed to perform integrated value added telecommunication services over IP Networks. By applying these services, a user may conveniently make a voice call through the Internet for PC-to-PC, PC-to-Phone, or Phone-to-Phone conversation. A significant advantage associated with VoIP is the tremendous savings in making long-distance calls. In order to provide advanced multimedia communication services, many industries and Internet Service Provides (ISPs) have designed and deployed IP telephony networks to provide VoIP related services. The state-of-the-art VoIP systems have the capability to largely replace the existing Plain Old Telephone Service (POTS). However, voice sessions over the best-effort IP networks are subject to users’ complaining for unavailable or unstable service. The current VoIP management systems most put their focuses on the call admission, call detail records (CDR) and call quality control, but seldom take fault management into account. It poses issues related to VoIP system’s availability and management call in question. Furthermore, as providing the same services as the traditional telephony systems, VoIP systems could have lots of branches distributed over many locations which may involve a large-scale and nationwide network architecture that presents a challenging centralized management problem. In this work, a distributed management framework is introduced for managing H.323-based VoIP systems. Within this framework, the Internet Telephony Gateway (ITG) actively sends out the SNMP-based management information as soon as the call state changes of a session are recognized by the ITG’s digital signaling processor (DSP). The call state sequences of a session can be specified by the Finite State Machine (FSM) and is used for fault detection, call quality monitoring, and accounting management. In addition, the distributed management architecture is applied to perform the ITGs management functions: The domain manager (DM), which manages the regional domain associated with ITGs and other involved network elements distributed over each VoIP branch; and the center manager (CM), which is located in NOC (Network Operating Center) manages the global domain associated with backbones and trunks of the VoIP network by gathering information from each DM.