The annual global expenditure of electricity consumed is more than $10 billion dollars, with cellular base-stations contributing to 60-80% of the energy consumption. In order to reduce the energy requirements of cellular networks, a promising new approach is to connect base-stations with energy harvesting and storage devices. The key benefits of such base-stations are threefold: (i) Green base-stations are suited for deploying off-power-grid base-stations, or where reliable power does not exist; (ii) they will reduce the operational cost for cellular providers, which could translate to lower costs for end-customers; (iii) they will reduce the carbon emission footprint of cellular infrastructure. A major challenge pursued in this project is to design a cost-effective green base-station based system that can adequately exploit these new harvesting and storage devices. Hence, the overarching goal of this project is to develop the mathematical foundations for the design and operation of cellular base-stations equipped with energy harvesting devices, and to develop practical solutions that can be implemented in real cellular systems. This project combines techniques from networking, algorithmic design, controls, optimization, and game theory to address critical issues in this important emerging area. Hence, graduate students trained on this project will be exposed to a variety of different disciplines, which in turn will be beneficial to them as they enter the global workforce. Further, the project will offer undergraduate students opportunities to be involved in the more practical aspects of the project. The PIs will also work closely with the industry to help impact real systems.
In order to achieve the aforementioned goal, research on the project will focus on the following three critical inter-related thrust areas: (i) Dynamic Energy Control: Developing control strategies for minimizing the energy costs of base-stations in cellular networks governed by a single operator. Some of the key difficulties that will be addressed here are: how to dynamically control the battery levels at each base-station taking into account energy harvesting and user-load profile dynamics; how to form associations between users and base-stations based on energy reserves; how to ensure both energy efficiency and user experiences are satisfied. (ii) Base-Station Availability Control: Developing energy management strategies that carefully balance the energy savings and the cost incurred in ON/OFF operations by jointly considering the traffic demand and renewable energy supply. (iii) Multi-operator Base-station Management: Cooperation and Competition: Designing mechanisms to allow energy cost reductions by resource-pooling between multiple operators taking into account the practical realities of the marketplace and incentives/penalties. The analytical models and algorithms developed during the course of this project will be validated via experiments on a testbed at OSU, and trace-driven emulations.
