Job Scheduling Under Differential Pricing: Hardness and Approximation Algorithms

To induce a favorable energy demand pattern, generalized pricing models were proposed to achieve better aggregated energy consumption pattern. In this work we study how to schedule jobs under two differential pricing models, namely the combined pricing of day ahead pricing (DAP) and inclining block rate (IBR) both in the micro scope and macro scope. In the micro scope we study offline job scheduling with a goal to minimize the electricity cost of consumers when the electricity price and job profile are known beforehand. In the macro scope we study the aggregated effect on the cost of power generation when each entity (e.g., a household or a factory) schedules their jobs autonomously. We first prove that the job scheduling problems are either APX-hard or NP-hard under two combined price models of DAP and IBR. We then present efficient methods with bounded approximation ratio and show that our scheduling achieves comparable electricity cost saving.