Dynamic Optimization of Sensor Mesh for Underground Environment Monitoring Systems
Todays underground monitoring and control systems are all wired, and as a result, lack the flexibility, mobility and dynamic extensibility that the modern safety systems require. For the long term, we believe that wireless monitoring systems will be a significantly more cost-effective than current wired systems. Above ground, there are many wireless monitoring systems that operate in other industries, such as chemical processing/refining plants, and hotel security/environment monitoring systems. However, in the confined spaces of underground tunnels, microwave propagation is hindered by limiting and elongated geometries that most frequency bands cannot overcome. MIMO (with OFDM) RF technology can be introduced and modified to utilize the scattering rich environment of the confined mining environment to achieve the desired long distance propagation. With specialized antenna systems, the real maximum throughput can easily exceed the 100 Mbps range that is required by modern mining communication systems. In addition, connectivity of environmental sensors is another challenge that must be overcome. The most popular topologies of wireless network deployment are star, starmesh hybrid and full mesh, however, none of the above can solve the connectivity problem created by the confined mining environment in the event of catastrophes, such as fire or explosion. We propose a hierarchical mesh topology which can be easily optimized in normal and disastrous conditions. This new topology has the same advantages of self-healing and self-organization as a regular mesh, but also takes advantage of current (popular) WiFi standards (i. e. 802.11n) with minimum modification of the existing communication protocols. Underground mining systems are a dynamic, everchanging environment, especially during catastrophes. Any monitoring system must be able to optimize itself when part of the network may be destroyed or temporarily disconnected. We introduce the Global advised Local Optimization method (GALO) to deal with unexpected network connectivity issues. Graph theory algorithms are used to analyze the dynamic network flow issue during mining catastrophes when the connectivity of a part of or the whole monitoring sensor system is in a very unstable state. Some parts of the network may be isolated, but not totally down, and can be re-connected again when disruptions of key connections are identified and repaired. The global information of the whole network is the most important reference for optimization; however the transient changes for local network nodes may not be reflected quickly enough to be considered in the global optimization process. Therefore, local optimization, based on the available information for the global network, will be critical for maximizing the connectivity of the monitoring network. In the sample problem, we will demonstrate the detailed process of how a disrupted mesh can heal and optimize itself.
Yang Gemei Wang Jiren Zhou Xihua
Liaoning Technical University, Fuxin 123000, China Motorola Inc. Liaoning Technical University, Fuxin 123000, China
国际会议
The 3rd International Symposium on Modern Mining & Safety Technology Proceedings(第三届现代采矿与安全技术国际学术会议)
辽宁阜新
英文
919-926
2008-08-04（万方平台首次上网日期，不代表论文的发表时间）