Crowdsourcing in Cyber-Physical Systems: Stochastic Optimization With Strong Stability

Cyber-physical systems (CPSs), featuring a tight combination of computational and physical elements as well as communication networks, attracted intensive attention recently because of their wide applications in various areas. In many applications, especially those aggregating or processing a large amount of data over large spatial regions or long spans of time or both, the workload would be too heavy for any CPS element (or node) to finish on its own. How to enable the CPS nodes to efficiently collaborate with each other to accommodate more CPS services is a very challenging problem and deserves systematic research. In this paper, we present a cross-layer optimization framework for hybrid crowdsourcing in the CPSs to facilitate heavy-duty computation. Particularly, by joint computing resource management, routing, and link scheduling, we formulate an offline finite-queue-aware CPS service maximization problem to crowdsource nodes' computing tasks in a CPS. We then find both lower and upper bounds on the optimal result of the problem. In addition, the lower bound result is proved to be a feasible result that guarantees all queues in the network are finite, i.e., network strong stability. Extensive simulations have been conducted to validate the proposed algorithms' performance.