Study Of Wireless Sensor Network For Speech Transmition

Sensor technology, micro-electromechanical systems, networks and wireless communication and other technological advances, promote the emergence and development of wireless sensor networks. Random distribution of a large number of sensor nodes in wireless ad-hoc manner can gather the data information of the external environment, such as voice, video, temperature, humidity, pressure and seismic waves etc. Then the data will be transmitted to the data center through the wireless short distance, low-loss, multi-hop communications.Wireless sensor networks have attracted the military, business, academic seriously. It will be widely used in the military, the environment, medical treatment, space exploration and disaster special areas.Wireless sensor network is a new technology. Foreign research of wireless sensors are developing speedly, but domestic wireless sensor network research is still in its infancy. However, the gap between domestic and international is not very large, so timely beginning this research will be a major strategic significance for the entire country's national defense, political, economy.Nowadays, an important aspect of sensor networks' study is how to collect, transmit, process simple environment data in mirconotes which have very limited resources. However, people dissatisfied with these environment data, they want wireless sensor nodes to transmit image, audio or video information, so that environment surveillance can be more accurate. For this reason, researchers began to study multimedia sensor networks.In this paper, we concentrate on the mac protocols of wireless sensor networks, especially the impact imposed on speech transmition. We also want to find a speech coder which is fit for wireless sensor networks. Firstly, we intorduce the existing mac protocol, such as IEEE802.11, SMAC protocol, TRAMA protocol, DMAC protocol. Moreover we analyse and make comparison of them. Secondly, we propose a new mac protocol named VR-TDMA. Simulation in NS2 proves that this protocol can have pleasing performance in real-time application. Thirdly, we make a comparison of some famous speech encoders. The result indicates that ADPCM (Adaptive Differential Pulse Code Modulation) is simpler than advanced low bit-rate voice coding techniques and doesn't require as heavy caculations. In this work, we simplify the algorithm in order to make it fit for the networks better. Actual tests prove that this simplified algorithm work well in wireless sensor networks.