Pocket-switch-network based services optimization in crowdsourced delivery systems

There are two major issues in spatial crowdsourcing: travel route optimization and control policies. To address the two issues above, we introduce the concept of Pocket switch network (PSN) into the CD-system. First, we formulate a generalized optimization problem into three aspects of connectivity, profit and risk, motivated by the concepts in PSN. Afterward, these three aspects are mathematically described and optimized by a routing algorithm based on dynamic mobility and social graph. This algorithm consists of two parts: social graph extraction and social mobility based routing. Social graph learns the social knowledge of each patrician while social mobility based routing decides the leaving nodes of the passages according to their social graph. Finally, we evaluated the effectiveness and robustness of the proposed method on the realistic traces. The results demonstrated its superior performance in connectivity, profit, and risk.