Research On Computation Offloading Technology Based On Genetic Ant Colony Algorithm In Self-organized Network

With the development of intelligent technology and global information process, intelligent terminals and various nodes in IOT play an increasingly important role in people's daily lives. They are the access for people to enjoy intelligent and informative life, the underlying tool to collect and transmit information and the specific implementer of a variety of intelligent and informatization functions. However, due to the inherent resource constraints, these devices are often unable to meet the increasing demand. Using idle resources in the network, Computing offloading technology migrates tasks on the terminal equipment to other resource-rich equipment, which effectively solves this problem, becoming a hot spot of current research.In this paper, we introduce the research status of computation offloading technology in the beginning, and emphatically study computation offloading technology in self-organizing network environment. Then the concept of self-organizing cloud is proposed and deepened. Its characteristics and classification are expounded. At present,the research of computation offloading in ad hoc network is mainly aimed at single migration target, with resource utilization rate or equipment energy consumption as the optimization target. And research now lacks consideration of multi-objective migration, overall network optimization and communication energy consumption. In this paper, we will focus on the shortcomings of existing research, and study the computation offloading techniques for asymmetrical self-organizing cloud and symmetrical self-organizing cloud scenarios respectively.This paper describes a computation offloading architecture of symmetrical self-organizing cloud. The computation offloading model of source node and processing node is designed, and the general flow of computation offloading is given. Then, we analyze the task scheduling problem in asymmetrical self-organizing cloud network environment and its optimization goals in this paper. Finally, a task scheduling algorithm based on genetic ant colony algorithm is designed and simulated, aiming at reducing the task average energy consumption and improving the task execution success rate. Compared with the probabilistic scheduling algorithm and Genetic Algorithm, the average energy consumption of the proposed algorithm is relatively stable, and the success rate can be increased by about 20%.This paper describes a typical symetrical cooperative self-organizing cloud network. Also, this paper illustrates the many-to-many computation offloading architecture of symmetrical self-organizing cloud network for the first time and gives the design of computation offloading model for node devices and central dispatching management module. Then,the task scheduling problem in symmetrical self-organizing cloud network environment is studied, and the task allocation model and task scheduling strategy are described. Finally, a task scheduling algorithm based on genetic ant colony algorithm is designed for optimization goals such as minimum task average execution time, load balancing and minimum task average execution energy consumption. The simulation analysis is given.Results show that, compared with the probabilistic scheduling algorithm and genetic algorithm, the average execution time of the proposed algorithm is reduced by 17.4% and 4.1% respectively, and the performance of the load balancing is improved a lot, which validates the effectiveness of the algorithm.