Selection Mechanism For Service Nodes In Crowd Sensing Based On Genetic Algorithm

The mobile devices are used as the basic unit in crowd sensing, and the perception is completed through the network. Nodes move randomly with the people's movement. So the perception can be completed anytime and anywhere. Moreover, the transfer mode as “store-carry-forwards” of opportunity data transfer mode is used in crowd sensing. In crowd sensing, the service nodes need to reach the target area to sense.The service nodes play a crucial role.We can say that, there is no service node there is no perception task. So, the number of the nodes and the performance of them have a direct impact on the quality of the service.Therefore, the selection of the service nodes has become a key issue to be solved. Most of the existing selection mechanisms of service nodes rely on location information or for a single attribute to select the service nodes. At present, there is no comprehensive selection mechanism of service nodes. So, this paper focuses on the selection of service nodes in crowd sensing to research, using genetic algorithm to select the service nodes. The results of research are as follows:(1) The genetic algorithm is used in crowd sensing to select the service nodes. This method optimizes the service nodes for multi-object and multi-attribute. In this mechanism, the multiple attributes are transformed into the corresponding objective function and the multi-objective function is transformed into the single objective benefit function. In the optimization process, the weights are assigned to each objective function, and the weights affect the optimization results. The assignment of weights depends on the user's prior experience. Using this mechanism to optimize the service nodes, the attributes of the selected service nodes are more comprehensive and meet the actual needs. This method realizes the leap that the service nodes are selected from a single node to multi-node.(2)Based on the single objective genetic algorithm, a multi-objective genetic algorithm is introduced to optimize the set of service nodes in crowd sensing. Multi-objective genetic algorithm is for multiple objective functions. This mechanism transforms the multi attributes into corresponding multi objective functions. And according to the characteristics of the objective functions, corresponding constraints are set. After optimization, the users choose the service nodes which they need from the set of the optimized service nodes depends on the late decision. When to select service nodes, it doesn't need to integrate multiple attributes to a target function. This method can optimize the set of service nodes directly and improve the efficiency of selection.This paper focuses on the selection of the service nodes in crowd sensing. And the genetic algorithms are applied in crowd sensing to select the service nodes. This paper realizes the optimal selection of service nodes and raises the efficiency of the perception services in crowd sensing network. The methods in this paper can give valuable references to the work of the selection of service nodes in crowd sensing.