A task allocation method based on data fusion of multimodal trajectory in mobile crowd sensing.

In the existing Mobile Crowd Sensing, the task allocation method based on sensing user trajectory only considers the sensing user's moving position, but ignores the user's other information, such as moving speed, environmental noise, and so on. To solve this problem, a task allocation method based on multimodal trajectory data fusion is proposed in this paper. Firstly, the characteristics of users' trajectory data in Mobile Crowd Sensing are analyzed, the effects of moving speed and environmental noise on users' acquisition intention are judged, and the speed grayscale image and sound grayscale image are established. Then, the velocity gray image and the sound grayscale image are fused to form the track grayscale image by using the Tow-Branch Convolution Neural Network. Finally, the sensing user selection is based on the grayscale image coverage similarity, to maximize the coverage. Compared with other methods, the results show that the method proposed in this paper is significantly better than other comparison methods in improving coverage. [ABSTRACT FROM AUTHOR]