Recently, the networking industries have gone through tremendous changes. It demands high-speed operations and complex problem-solving abilities. To manage these evolutions Internet-of-Things (IoT) is a proposed solution from several technical corners. Numerous researchers and government organizations showing their interest to provide solutions with IoT implementation. Handling a huge amount of network data, its privacy and security, Quality of Service (QoS) requirements and heterogeneity of underlying networking components are the various challenges in IoT implementations. To provide the solution, Software Defined Networking (SDN) is becoming a bliss in managing such complex networking problems. The allocation of the Virtual Machines (VMs) into the end device is an NP-Hard combinatorial optimization problem. We formulate the problem by using simple Additive Weighing (SAW) or Weighted Sum Method (WSM) to allocate the VMs asymmetrically based on CPU Utilization and Memory usage to optimize the energy. The proposed algorithm ServerCons minimizes the number of live migrations and the number of nodes used as well as the energy usage is at par with the state of art algorithms such as First-Fit-Decreasing(FFD), Best-Fit-Decreasing (BFD), and Modified-Best-Fit-Decreasing (MBFD).
30 April 2012 in International Journal of Computer Applications
In the current era of technology, the number of users connecting to the internet has increased rapidly. If the situation is like Pandemic (COVID-19), the demand of the data services is much higher than the expectation. The major problem of this situation is load management. SDN plays an important role in this regard. Being a programmable network, it is widely admirable for its network intelligence. In this paper, we discuss SDN architecture, its components, various applications and major challenges. The unique architecture differentiates SDN from traditional network. It makes it possible to take routing decision without depending upon the underlying architecture. We also present a systematic review on the study of the different load balancing approaches, existing solutions and various load balancing techniques in SDN. With the network intelligence, SDN is able to take the load balancing decision. We categorize the load balancing techniques based on controller, server, communication path selection and cloud-based environment. Finally, we present different research challenges, existing solutions under each technique.
11 March 2018 in International Journal of Engineering & Technology
Energy efficiency in wired networks has received considerable attention over the past few years for its heavy economic impact. Energy consumption in such network is becoming a challenging issue. Among all these wired networks Software Defined Networking (SDN) is considered as an intelligent network in which use of programmability provides the best solution for network operation. The logical separation of data plane and control plane enhances its network functionality. Despite of all these mechanisms, energy consumption is also an alarming issue in SDN. Various researchers propose many approaches in order to utilize the energy in its best way. Herein we consider the normal scenario of a data center network, and propose a new energy efficient routing algorithms to improve the performance in SDN, called SERAT. In this work, we focus on discrete of data rate of the link in order to minimize energy consumption in the network. In wired network the underutilized links are even operated in high data rate, increasing energy consumption. In order to reduce energy consumption of the network the links must be operated as per their utilization factor. We have used an output buffer queue with two levels of threshold values in order to monitor the performance of the network. Simulation result shows that the proposed technique leads to major amount of energy savings without mush affecting the throughput.
The traditional Internet defined by the TCP/IP was no more able to accommodate huge amount of traffic in order to accommodate large number of devices. To overcome this and to solve the problem of wasting of network resources, SDN is considered as one of the best approaches. SDN traffic engineering gives the immense solution to enhance the network performance at both traffic and resource levels. Being the centralized authority SDN is able to give better results in network load balancing and improve delay and loss performance. In this paper, we use two layer controller architecture with the name of Load Balancing in SDN using Effective Traffic Engineering Method (LBTEM) in order to maintain the throughput of the network without any delay in packet reception. At the same time we are taking care to minimize the overall network energy consumption.