Tarek K. Refaat

Dynamic Virtual Machine Migration in a vehicular cloud

Vehicular clouds are formed by incorporating cloud-based services into vehicular ad hoc networks. Amongst the several challenges in a vehicular cloud network, virtual machine migration (VMM) may be one of the most crucial issues that need addressing. In this paper, a novel solution for VMM in a vehicular cloud is presented. The vehicular cloud is modeled as a small corporate data center with mobile hosts, equipped with limited computational and storage capacities. The proposed scheme is called Vehicular Virtual Machine Migration (VVMM). The VVMM aims to achieve efficient handling of frequent changes in the data center topology, host heterogeneity, all while doing so with minimum Roadside Unit (RU) intervention. Three modes of VVMM are studied. The first mode, VVMM-U uniformly selects the destinations for VM migrations, which will take place shortly prior to a vehicles departure from the coverage of the RU. The second mode, VVMM-LW aims at migrating the VM to the vehicle with the least workload, and the third mode, VVMM-MA incorporates mobility awareness by migrating the VM to the vehicle with the least workload and forecasted to be within the geographic boundaries of the vehicular cloud. We evaluate the performance of our proposed framework through simulations. Simulation results show that VVMM-MA introduces significant reduction in unsuccessful migration attempts and results in an increased fairness in vehicle capacity utilization across the vehicular cloud system.

Virtual machine migration and management for vehicular clouds

Vehicular Cloud Computing is a growing research field which consolidates the benefit of cloud computing into vehicular ad hoc networks. However, few studies address vehicles as potential Virtual Machine hosts. Due to the rapidly changing environment of a vehicular cloud, a host can easily change or leave coverage. As such, Virtual Machine Management and Migration schemes are necessary to ensure cloud subscribers have a satisfactory level of access to the resources. This paper proposes several Vehicular Virtual Machine Migration (VVMM) schemes: VVMM-U (Uniform), VVMM-LW (Least Workload), VVMM-MA (Mobility Aware) and MDWLAM (Mobility and Destination Workload Aware Migration). Their performance is evaluated with respect to a set of metrics through simulations with varying levels of vehicular traffic congestion, Virtual Machine sizes and levels of load restriction. The most advanced scheme (MDWLAM), takes into account, the workload and mobility of the original host as well as those of the potential destinations. By doing so a valid destination will both have time to receive the workload and migrate the new load when necessary. The behavior of various algorithms is compared and the MDWLAM has been shown to demonstrate best performance, exhibiting migration drop rates that are negligibly small.

Network fabric redundancy in NCS

Fault-tolerance is becoming an increasingly crucial aspect of the design of Networked Control Systems (NCSs) in order to mitigate system downtime. However, the introduction of fault-tolerance is typically associated with significant traffic overhead. In this paper, an optimization to an Ethernet-based network fabric fault-tolerant NCS is proposed. The proposed optimization halves the amount of overhead traffic necessary for fault-tolerance while maintaining the same level of robustness. Moreover, based on the same optimization methodology, an expanded model with two in-line cells is presented and subsequently tested. The expanded two-cell model is designed to provide controller-level in addition to fabric-level fault-tolerance. Simulations using OPNET followed the traffic analysis, and proved the models to be fully reliable in the case of a single failure at a time, for both the single-cell model and the expanded two-cell network.

WSN goodput and energy consumption for various wireless communication standards

This paper proposes a methodology for calculating the energy consumption and the throughput of a Wireless Sensor Network (WSN) taking into account retransmissions and data dropped due to collisions. For various scenarios, simulations are conducted using OPNET and utilizing off-the-shelf wireless communications standards such as ZigBee, Low-Power Wi-Fi and Wi-Fi. A figure of merit is then introduced for a fair comparison between the mentioned standards.

On the performability of on-board train networks with fault-tolerant controllers

This paper studies a mixed load of control and entertainment on top of Gigabit Ethernet on-board a train wagon. The control load is comprised of sensors and actuators of different sampling periods. There is a dedicated server for control and a dedicated server for entertainment. It is proven that the system can tolerate the failure of one controller, shifting both entertainment and control loads to the remaining controller while guaranteeing correct performance. The maximum acceptable entertainment load is found. Enhancements to the network are introduced in order to minimize degradation of entertainment service due to a controller failure. OPNET simulations are used to study packet transmission/reception and end-to-end delay. Finally, a performability model is developed; this model will help system designers determine the cost- effectiveness of architectural features aiming at increasing performance.

Wi-Fi-based hierarchical Wireless Networked Control Systems

This paper proposes a novel architecture for a hierarchical Wireless Networked Control System (WNCS). It consists of three cascaded workcells each containing 30 sensors, 30 actuators and one controller. The wireless communication protocol used is IEEE 802.11g with multicasting. The hierarchy of the system is such that the lowest level is that of the sensors and actuators, the intermediate level is the controllers, and the highest level is a supervisory node. This supervisor can be either active or passive. System performance is measured using OPNET simulations and the results are confirmed analytically. The system is shown to tolerate all possible controller failure scenarios. The supervisor can handle the entire control load of all three controllers, should the need arise. The system exhibits zero packet drops and delay constraints are met in all scenarios. The effect of interference is then investigated and the maximum interference that can be tolerated by the system is quantified.

Simulation of parallel redundant WLAN with OPNET

Applying multiple redundant and diverse communication channels is an established method to achieve an improved overall communication channel. When applied for packet-based data transmission over channels with strongly nondeterministic behaviour due to environmental influence, such as Wireless Communications, timing performance can be greatly improved by this approach. The central element in such a system is the so called “Combiner” on the receiving side. In this work, a new specific type named “Timing Combiner” is described. The Parallel Redundancy Protocol (PRP) according to IEC 62439-3 realizes such a Timing Combiner on the Ethernet level. In this work, an OPNET simulation model is created and analysed for its performance characteristics. Also a quantitative analysis of the effect of different interference models in an industrial environment is presented.

Network fabric fault-tolerance for Ethernet-based Networked Control Systems

Networked Control Systems (NCS) are widely used in industrial applications. Ethernet has recently been used as the communication protocol in NCS. Since overcoming faults is crucial in control systems, fault-tolerance in NCS is being extensively investigated. This paper focuses on the fault-tolerance aspect of Ethernet-based NCS at the network fabric level. It proposes a novel architecture that achieves successful recovery from any single link or switch failure. The model was also tested and proved to be fault-tolerant for some multiple failures.

On the Performability of Fault-Tolerant NCSs with Video Sensors

Smart sensors in Networked Control Systems are increasing in abundance with Industry 4.0. The use of such smart sensors can offer various advantages to a traditional networked control system. Furthermore, the utilization of such smart sensors is more realizable specifically due to the use of networked control systems, because of the abundance of bandwidth. This paper presents a networked control system, utilizing switched Ethernet, with sensors, actuators and an intelligent controller. The system also adds a smart sensor, which processes video through an onboard camera and embedded controller. The research investigated the impact of the video sensor on the system and proposed a fault-tolerant model that studies the utilization of the sensor’s onboard controller as a backup to the system’s main controller. The model is successfully simulated using Riverbed Network Modeler. Finally, a performability analysis is conducted on the proposed model.

A Wireless Body Area Network architecture for a prosthetic arm

State-of-the-art prosthetics incorporate network communication for both sensing and control, offering various degrees of freedom to amputees. Existing implementations rely on wired networks, which can restrict mobility and are vulnerable to connection wear and tear. Wireless Body Area Networks (WBANs) are an alternative to wired approaches; they avoid the issues of mobility and connection wear and tear. This study proposes a novel WBAN architecture for a prosthetic arm. The system requirements are based on an existing prototype designed by Johns Hopkins. The proposed architecture is simulated using OMNeT++ in the presence of interference. The delays and packet loss values are shown to be within satisfactory limits, based on a 95% confidence analysis.