The data exchanged over networks has grown over the years, and is expected to grow further moving forward. This has led to increases in the required bandwidth, storage, and computing power of the entities involved. Due to this, present-day networks have moved to cloud environments to achieve better scalability and reduce infrastructure costs. Cloud service providers host their cloud environments on large-scale data centers, which process huge volumes of data. Processing such huge volumes of data requires an enormous amount of energy, leading to increased carbon footprints and thus environmental impacts. Researchers are now looking for ways to reduce energy consumption in data center networks (DCNs).

The authors use software-defined networking (SDN) to optimize network topology and reduce energy consumption in DCNs. SDN enables dynamic configurations of a network and its resources. The authors propose a flow mapping algorithm that studies the network flows of the network infrastructure; based on that, it dynamically puts network elements such as network switches and links in active and inactive modes, and controls the speeds of network switches. This helps in controlling the energy consumption of the data center based on network demand. Three strategies--power on/off links, traffic mapping, and link speed adaptation--are studied through simulations. From their study, the authors were able to reduce energy consumption by up to 70.02 percent. Future work includes implementing the proposed strategies on an OpenFlow controller. Also, as data centers are distributed all over the world, complex factors such as topology optimization, server localization, and load balancing need to be addressed.

The flow of information is good, which makes the paper easy to understand and follow. This paper will be useful for researchers working to develop energy-efficient DCNs.