Energy-Efficient Resource Allocation for Heterogeneous Edge-Cloud Computing

With the rapid development of Internet of Things (IoT) technology, billions of mobile devices (MDs) are putting a massive burden on limited radio resources. Mobile-edge computing (MEC) can save MDs' energy consumption and relieve network pressure by offloading their tasks to edge servers. Compared with cloud servers, edge servers are closer to the users but have less storage capacity. The heterogeneous edge-cloud computing paradigm recently developed combines the advantages of both. In this architecture, edge servers provide powerful computing power, while the cloud provides sufficient storage capacity. Since many IoT devices in such a scenario are mobile, it is more practical to consider user mobility when optimizing the network. Besides, properly utilizing the mobility context can be beneficial for improving network performance as well. We focused on the edge-cloud collaborative computing scheme, as well as the joint optimization of power control, transmission scheduling, and offloading decisions among MDs and edge servers so as to minimize the total energy consumption of all MDs while considering user mobility. The problem was modeled as a mixed-integer programming (MIP) optimization problem that provided the optimal solution. We also proposed a low-complexity heuristic algorithm. Simulations showed that the proposed edge-cloud collaborative scheme could significantly reduce the energy consumption of MDs compared with other schemes and demonstrated the importance of considering mobility awareness.