Some Transformation Methods on Probabilistic Model for Crowdsensing Networks

The crowdsensing problem can be transformed into the coverage problem. Comparing to the traditional deterministic sensingmodel, theprobabilisticmodelismoresuitabletodescribe the characteristics of the crowdsensing problem since it has considered the errors of the sensors and the collaboration among sensors. However, to suit the probabilistic models, we need to design new algorithms which is not easy. The original algorithms cannot be used directly since we need to consider the cooperation of nearby sensors. What is worse, due to the hardness of verifying the exact probability of detection, it is difficult to design optimal algorithms on the probabilistic models directly. In this work, we propose three methods that can transform the original coverage algorithms on the disk model to ones on the probabilistic model. Our methods can preserve the characteristics of the original algorithms and the conversion process has low time complexity.