In this paper, we propose a heterogeneous framework to deliver the smart grid (SG) data cost effectively. The data generated by distributed SG loads and generation units should be delivered to the utility control center within the tolerated delay, which is crucial for SG applications. To this end, a heterogeneous communication framework is proposed, where the cellular network provides ubiquitous yet expensive data transmission and vehicle-assisted device-to-device (D2D) communications are leveraged to offload the cellular network by delivering the delay-tolerant SG data in a store-carry-andforward fashion with low cost.
To improve the offloading and cost performance of the proposed framework, we put effort in the following aspects: i) optimal forwarding schemes to optimally select vehicles to carry and forward the data; ii) mode selection and dynamic resource allocation to maximize the amount of data delivered by D2D communications, reduce the cost of SG data delivery, and guarantee the fairness among SG users. Simulation results are given to validate proposed approaches and demonstrate that the proposed framework is efficient in saving cost for the utility and offloading the cellular network.