Wireless Sensor Network Routing Protocol Based On The Improved Shuffled Frog Leaping Algorithm

With the highly development of wireless communication, micro electromechanical system(MEMS) and artificial intelligence, human beings have progressively enter the era of Internet. With the quickly application of wireless sensor network to the area of Internet of things, it has drawn much attention of the academic and scientific research institution at home and abroad and business sectors. It mainly ascribe to the self-organization ability of wireless sensor network, the topological structure of dynamic Internet, large-scale Internet and others characteristics.Wireless sensor network is self-organized network consisted of mass cheap and ubiquitous tiny sensors which equipped with wireless communication and computing power. It mainly depend on the sensors throughout the whole Internet to receive, collect, and cope with the data information of sensors in a distributed and cooperative way. However, the fatal weakness of wireless sensor node is its limited energy, and it will influence the life-cycle of the whole Internet. Therefore, a suitable wireless sensor Internet routing protocol is of great importance, and a reasonable and effective clusterrouting protocol algorithm is extremely necessary for reducing the energy consumption of wireless sensor node and promoting the energy utilization ratio.Thesis thesis gives a brief introduction for the principles of leapfrog algorithm, and analysis and summary others research efforts. The author obtains advantages of leapfrog algorithm applying to WSN routing protocol, at the same time, there are also disadvantages of leapfrog algorithm and gives corresponding improvements. In order to improving the optimal capacity of leapfrog algorithm in local and inside, this thesis introduces the Metropolis criterion of simulated annealing algorithm, and puts forward a bidirectional ring transfer mechanism to improve its leapfrog algorithm jumping out of local optimal capacity. What's more, the author also puts forward a new algorithm called SFSLA. Through a analysis of MATLAB for high-dimensional unimodal function and high-dimensional multimodal function, and the test show SFSLA is better than SFLA?ISFLA in the respect of iterations and accuracy, as well as the respect of “premature” phenomenon. It gives a verification of the superiority of SFSLA algorithm, and lays solid foundation for the further improvements of routing protocol.Finally, the thesis draws a conclusion of various cluster-routing protocol, then puts forward an improvement of wireless sensor network routing protocol for leapfrog algorithm. This algorithm combined with the residual energy of wireless sensor node and locations to established fitness function. Through the improvement of leapfrog algorithm to realize the optimal solves of fitness function. In this way to obtain a suitable cluster and adapt a new routing mode mixed of single-hop routing and multihop routing during the process of cluster head data transmission. The simulation experiment of LEACH protocol and SFLA protocol use of OPNET simulation software. The compared and analytic test result show that the improved routing protocol have apparent advantages in the distribution of cluster head, the reducing of Internet energy consumption, the extension of Internet life cycle, the improvement of energy balance degree, the diminution of network delay.