| Wireless Sensor Network (WSN) is built with some mini equipment which can sense the phenomenon just like temperature,pressure and humidity. The failure of WSN nodes are always due to their limited energy and communication ability. Multi-hop routing is to gathering the data collected by all the nodes scattered in various areas, and eventually, transmit the data to sink. The process will cost maximum energy of the nodes during the lifetime of WSN. So, it's important for the nodes to transmit the minimum data.Firstly, this paper gives a data aggregation algorithm based on learning automata to solve the problem that the existing data aggregation algorithm can't solve, the uneven energy cost, and the existing algorithm can't change the gathering path dynamically existing the overhead environment. In the proposed method, nodes can change its gathering path to adjust the overhead environment. All the nodes of WSN equipped with a learning automata. These leaning automata learn all the gathering path of the nodes. In the process of transmit information two kinds of data are transmitted, including data packet, knowledge packet.When the information of the nodes changes, according to the feedback of the nods, the learning automata gives the reward or punish to the current gathering path, which help to find the best gathering path. Secondly, this paper improved the wavelet data compression algorithm, which was brought out as the correlation between different data. The algorithm do not reduce much of the data relate to the original data. After the wavelet data compression, Huffman coding compression algorithm will improve the data compression ratio.Through the simulation experiments, the gathering algorithm based on learning automata, can well solve sensor network in data collection process existing overheating node problem, and the gathering ratio of external environment is ideal improved. The low-frequency coefficients of wavelet transform can be well compressed by Huffman coding. |
PDF Full Text Request