Research Of Routing Algorithm Based On Link Quality For Industrial Wireless Networks

Recently, wireless sensor networks (WSNs) have gained great attention from the research community for various industrial applications, including industrial mobile robots, real-time inventory management and process and equipment monitoring. However, industrial wireless sensor network is operating under severe conditions of electromagnetic interference (EMI) and multi-path interferences. It will bring the following two challenges. On the one hand, since the reliable data transmission of wireless network is a big challenge, the design of routing algorithms that consider link quality to choose the best routes are enabling routing metrics to evolve. On the other hand, the existing simulation tools have not complete modeling the electromagnetic interference in industry site, and cannot meet the requirements of the WSN simulation experiment. In this paper, the industrial environment is categorized into different topographies, and the definition is based upon the specific physical characteristics of the local surroundings affecting both large-scale fading and multipath interference. The propagation model is well expressed by a one-slope path-loss model and excellent agreement with a lognormal distribution is obtained. Then the simulation environment is constructed based on the on-site data in OPNET simulator. Finally, in order to obtain the best routing metric under different industrial environments, we conduct a detailed performance analysis of four commonly used link-quality metrics in WSNs:ETX, Hop Count, PRR and WMEWMA. We performed the simulation in OPNET simulator using CTP, a tree-based WSNs routing protocol. According to the result of the simulation, we know that ETX is the most optimal routing metric on the overall performance. As to our knowledge, this paper is the first work to compare the performance between these link quality based metrics with networks of different qualities under different industrial conditions. The results show that the proposed simulation platform is suitable for industrial network simulation, and effective support for future research.The graph routing proposed by WirelessHART is one of the routing protocols with best robustness among all the wireless networks. Since there are few algorithms on graph routing at present and the routing performance is poor, a reliable multipath graph routing algorithm is presented, called LQI-based multipath graph routing (LBMGR). Instead of providing end-to-end node-disjoint multipath routing, LBMGR only ensures data forwarding for hop-by-hop multipath. LBMGR defines the network topology as a connected graph then generates the binary tree from a node to the given destination, the sub-graph that includes all the edges of the binary tree. In the packet transmission process, there are no advance paths from the source node to end for connections between nodes, this feature makes the routing easily switch to another node when the current node failure. In addition, the real-time link quality estimation is combined with periodical link quality estimation to provide reliable routing selection in face of the poor link quality in harsh industrial field. The routing generating algorithm is presented in detail, as well as the link quality index (LQI) updating mechanism. The performance of our approach is studied by OPNET simulator, simulation results show that our proposed algorithm can greatly improve the routing reliability and make contribution to reduction in the delay.WirelessHART, as one of the most popular short-range wireless application standards, is quite popular in the industry field. Due to stringent timing requirements of the WirelessHART protocol, simulating the resource allocation and the corresponding time consumption are essential for better performance. The third part of our work aims at developing a WirelessHART simulation platform based on OPNET. In order to simulate the mechanism of the TDMA resource allocation and mesh internetworking, a simple WirelessHART protocol stack is discussed. A series of experiments are conducted. The simulation result shows that the proposed simulation network is good at load balancing when the workload of the network is low. In a word, the proposed simulation platform is suitable for industrial network simulation, and effective support for future research such as routing protocol and allocation algorithm, etc.