Server and network load balancing

Software Defined Networking (SDN), is an emerging networking technology. This thesis aims to develop a new Server and Network Load balancing scheme in content delivery datacenters using SDN-based architecture. The scheme, called Server and Network Load Balancing (SNLB), tends to distribute the traffic load more evenly across the network. The SNLB achieves even distribution of flows on the links and servers by utilizing real-time network statistics. Furthermore, SNLB classifies the network flows into mice (flows with small bandwidth) and elephant (flows with large bandwidth) flows and performs load balancing on these two classes of flows separately. A detailed comparison of SNLB with Global first fit, Round robin and Load based balancing is presented. Other objectives achieved in this thesis are the designs of overload traffic handling technique and Fault tolerance method. The overload traffic handling technique activates and de-activates servers according to the traffic load; the fault tolerance method can reduce the impact on network performance during the network fault.

Server and network load balancing

Software Defined Networking (SDN), is an emerging networking technology. This thesis aims to develop a new Server and Network Load balancing scheme in content delivery datacenters using SDN-based architecture. The scheme, called Server and Network Load Balancing (SNLB), tends to distribute the traffic load more evenly across the network. The SNLB achieves even distribution of flows on the links and servers by utilizing real-time network statistics. Furthermore, SNLB classifies the network flows into mice (flows with small bandwidth) and elephant (flows with large bandwidth) flows and performs load balancing on these two classes of flows separately. A detailed comparison of SNLB with Global first fit, Round robin and Load based balancing is presented. Other objectives achieved in this thesis are the designs of overload traffic handling technique and Fault tolerance method. The overload traffic handling technique activates and de-activates servers according to the traffic load; the fault tolerance method can reduce the impact on network performance during the network fault.

Load balancing algorithm based-SDN for the IoT applications

Nowadays, the emergence of IoT devices has wholly revolutionized the customer's communication habits. The information can be collected at anytime and anywhere. However, the mobility of communication devices in a dense network results in an unbalanced network load and an increase in bandwidth demands. To address these issues, this study proposed a load balancing algorithm based on SDN for enhancing the performance of mobile IoT devices communication over a Wi-Fi network. The use of the SDN makes possible the automatic configuration of the network through a centralized controller, it provides programmability, a global view of the network, it also optimizes resource allocation based on real-time network information that helps implement our algorithm. The proposed algorithm is evaluated through simulation using mininet. The results indicate that our proposed method provides an efficient network load balancing and improves the throughput of associated devices.

Wing of 5G IoT and Other Services

The day-to-day advancements have brought the biggest challenge to network providers as it has become difficult to keep up the traditional networks with the ever-advancing technologies for them. It also result as a motivation for vendors to grow by developing, innovating, deploying, and migrating in their services, upgrading to new hardware and infrastructure, as well as hiring newly trained people, which requires a large amount of money and time to implement. It results to a need of a new network architecture who has a capability of supporting future technologies along with solving all sorts of issues known as the network proposal by software. For meeting highly increasing demands, various proposals of load balancing techniques come forward in which highly dedicated balancers of loads are being required for ever service in some of them, or for every new service, manual recognition of device is required. In the conventional network, on the basis of the local information in the network, load balancing is being established.

Scaling complexity comparison of an ACO-based routing algorithm used as an IoT network core

This paper proposes a routing method that is based on an Ant Colony Algorithm (ACO) for minimizing energy consumption in Wireless Sensor Networks (WSNs). The routing method is used as the backbone of the Internet of Things (IoT) platform. It also considers the critical design issues of a WSN, such as the energy constraint of sensor nodes, network load balancing, and sensor density in the field. Special attention is paid to the impact of network scaling on the performance of the ACO-based routing algorithm.