Design and Implementation of Distributed Controller Clustering for Solving the Issue of Single Failure in SDN Networks

Software-Defined Networks (SDN) It is a centralized control structure in the network that opens up new possibilities that did not exist before. The significant characteristic of this innovative approach is the focus on the capability of proposing networks of high dynamicity and programmability to transform the intelligence of underlying systems to the networks via controllers. The main issue of the SDN approach is found in its security, mainly due to its central-controlling architecture since the entire network is controlled from a central point. This makes it very vulnerable to single-point failure. In this paper, a fully Distributed SDN controller is proposed for solving the one point failure which exists within the single SDN controller. In general, the concept involves forming cluster of distributed controllers whereby each controller controls its domain and can thereby share the load within the network. The experimental results of the proposed system show an increase and enhancement in the performance of the network. The single-point failure issues have been overcome. The throughput of the proposed system increased with 20% while the packet loss rate was minimize with 33%.

Convexified Graph Neural Networks for Distributed Control in Robotic Swarms

A network of robots can be viewed as a signal graph, describing the underlying network topology with naturally distributed architectures, whose nodes are assigned to data values associated with each robot. Graph neural networks (GNNs) learn representations from signal graphs, thus making them well-suited candidates for learning distributed controllers. Oftentimes, existing GNN architectures assume ideal scenarios, while ignoring the possibility that this distributed graph may change along time due to link failures or topology variations, which can be found in dynamic settings. A mismatch between the graphs on which GNNs were trained and the ones on which they are tested is thus formed. Utilizing online learning, GNNs can be retrained at testing time, overcoming this issue. However, most online algorithms are centralized and work on convex problems (which GNNs scarcely lead to). This paper introduces novel architectures which solve the convexity restriction and can be easily updated in a distributed, online manner. Finally, we provide experiments, showing how these models can be applied to optimizing formation control in a swarm of flocking robots.

SDoT-NFV: A Distributed SDN Based Security System with IoT for Smart City Environments

The Internet of Things (IoT) is a key developing innovation aimed at linking objects via the Internet. While, Software Defined Networking (SDN) is another modern network- ing domain intelligence innovation that increases network effi- ciency and enhances security, reliability, and protection through dynamic software programs. In this paper, we proposed a distributed secure SDoT-NFV architecture for smart cities with Network Function Virtualization (NFV) implementation. We integrated highly protected SDN that delivers better network ef- ficiency, protection, and privacy results. It also secures metadata within each layer as well as payload. In addition, this architecture attempted to implement a more efficient method for constructing a cluster via SDN. Moreover, SDN-IoT with the NFV ideas brings benefits in terms of energy conservation and load balancing to the relevant fields. In addition, several distributed controllers have suggested enhancing accessibility, integrity, anonymity, con- fidentiality, and so on. We also implemented an energy-efficient Cluster Head Selection (CHS) algorithm to make use of our proposed architecture. The network offers greater protection of each network layer as opposed to the traditional network in the proposed architecture. Lastly, we analyze the efficiency of the proposed architecture with different network parameters (throughput, RTT, and Time sequence) for smart cities. GUB JOURNAL OF SCIENCE AND ENGINEERING, Vol 7, Dec 2020 P 27-35

A Staged Finite-Time Control Strategy for Formation of Underactuated Unmanned Surface Vehicles

The formation control issue for a group of underactuated unmanned surface vehicles (USVs) is discussed in the paper, and a staged finite-time control strategy for the USVs is proposed. Firstly, we try to steer each USV to its own starting point in the formation for a limited time, under the initial condition that each of these vehicles is parked at random. To deal with the nonholonomic behavior of the system, the dynamics of the USV is transformed into cascade systems. Then, the finite-time controller is designed for each vehicle based on homogeneity theory. After each USV reaches its own starting point with desired orientation, the model of the vehicle is decomposed into two subsystems under the Serret-Frenet frame. In order to maintain the formation pattern, two finite-time distributed controllers are developed for the surge subsystem and the yaw subsystem, respectively. The settling time for the staged control strategy is limited. Numerical simulations are carried out to illustrate the effectiveness of the proposed formation control strategy.

Achieving Balanced Load Distribution with Reinforcement Learning-Based Switch Migration in Distributed SDN Controllers

Distributed controllers in software-defined networking (SDN) become a promising approach because of their scalable and reliable deployments in current SDN environments. Since the network traffic varies with time and space, a static mapping between switches and controllers causes uneven load distribution among controllers. Dynamic migration of switches methods can provide a balanced load distribution between SDN controllers. Recently, existing reinforcement learning (RL) methods for dynamic switch migration such as MARVEL are modeling the load balancing of each controller as linear optimization. Even if it is widely used for network flow modeling, this type of linear optimization is not well fitted to the real-world workload of SDN controllers because correlations between resource types are unexpectedly and continuously changed. Consequently, using the linear model for resource utilization makes it difficult to distinguish which resource types are currently overloaded. In addition, this yields a high time cost. In this paper, we propose a reinforcement learning-based switch and controller selection scheme for switch migration, switch-aware reinforcement learning load balancing (SAR-LB). SAR-LB uses the utilization ratio of various resource types in both controllers and switches as the inputs of the neural network. It also considers switches as RL agents to reduce the action space of learning, while it considers all cases of migrations. Our experimental results show that SAR-LB achieved better (close to the even) load distribution among SDN controllers because of the accurate decision-making of switch migration. The proposed scheme achieves better normalized standard deviation among distributed SDN controllers than existing schemes by up to 34%.

Guaranteeing String Stability of Multiple Interconnected Vehicles Using Heterogeneous Controllers

In this paper the issue of string stability for acceleration-controlled vehicles interconnected in a chain is studied. String stability is concerned with having bounded displacements between vehicles in such a way that displacements should not grow unboundedly with respect to the perturbation. Different definitions can be given to string stability: one that relates to the amplification of a local disturbance acting on one vehicle towards the whole vehicle chain, more strict definition that is related to the boundedness of vector norm of displacements with respect to the bounded vector norm of disturbance inputs acting on all vehicles; and, most practical definition that considers the boundedness of signal norm of each individual displacement with respect to the bounded signal norm of disturbance inputs acting on all vehicles, independently from the number of vehicles. It has been proven that these definitions are all impossible to be achieved using any linear homogeneous unidirectional distributed controllers with constant spacing policy. This paper proposes linear heterogeneous controllers where each vehicle behaves differently from others in a vehicle chain. We prove that three different definitions of string stability can be attained using the proposed heterogenous controller. We propose sufficient conditions to guarantee string stability and boundedness of acceleration of each vehicle. Finally, simulation results are given to illustrate the effectiveness of proposed heterogenous control synthesis.