Towards SLA Policy Refinement for QoS Management in Software-Defined Networking

Towards SLA Policy Refinement for QoS Management in Software-Defined Networking

The ID-HILS setup considered in this study couples an engine-in-the-loop simulation EILS setup in Ann Arbor, MI, with a driver and vehicle model inWarren, MI. Towards SLA Policy Refinement for QoS Management in Software-Defined Networking the setup at the highest level. Systems with italic labels are modeled, and regular labels denote physical systems. Even though the ultimate goal is to use human drivers, a driver model is considered in this paper to increase repeatability of simulations and to eliminate the variation due to the human driver. The driver model follows the drive cycle up through the first two hills of this cycle. The coupling point for this integration is chosen at the end of the transmission, so the physical engine and the transmission model constitute the Ann Arbor site, and the rest of the vehicle and driver models constitute theWarren site.

The coupling variables exchanged by the two sites are the torque and speed of the transmission shaft. More specifically, the Ann Arbor site sends the transmission output torque, andWarren site sends the transmission output speed. In addition, the Warren site also sends the throttle command, which is fed to the drivetrain model and, through an idle controller, to the engine on the Ann Arbor site. The hardware components and models comprising this IDHILS system will be described in detail next. Towards SLA Policy Refinement for QoS Management in Software-Defined Networking The Ann Arbor site uses a diesel engine in combination with drivetrain, flywheel and idle controller models. A highfidelity AC electric dynamometer couples the physical engine with the models in real time.

Towards SLA Policy Refinement for QoS Management in Software-Defined Networkings

The sole purpose of the flywheel model, which is simply a small inertia element, is to reverse the causality between the drivetrain model and the engine, provide a speed input to the engine and receive torque from the engine. This is done to control the engine speed, which proves to be much more stable than trying to control the engine torque. In addition, an idle controller model, a simple PI controller with anti-windup, is used to override the engine’s own internal idle controller and to idle around to prevent the engine from stalling. The drivetrain model includes the torque converter, transmission, and shift logic. The torque converter model is a static model that takes pump and turbine speeds as inputs and generates pump and turbine torques according to the equations is the speed ratio between turbine and pump speeds, is a piecewise function approximating a desired capacity factor curve, and is a piecewise linear function approximating a desired torque ratio curve.

The transmission takes into account the transmission shaft inertia, stiffness, and damping, as well as the gear inefficiencies and torque losses due to fluid churning. Specifically, the speed reduction in each gear is assumed to be ideal, while the torque multiplication is assumed to be scaled by an efficiency factor. Furthermore, the torque lost due to fluid churning is modeled as variable nonlinear resistance of the form where are coefficients that change depending on the gear, and is the angular speed of the transmission shaft at the torque converter end. The inputs to the shift logic, the final element in the drivetrain model, are the transmission output shaft speed and the throttle demanded by the driver. A simple chart, as shown in, determines the current gear number and whether or not an upshift or downshift is to be initiated. The solid and dashed lines indicate upshift and downshift thresholds, respectively.The Warren site includes the driver and vehicle dynamics models. The driver model, which takes the desired and actual vehicle velocities as inputs, is a PI controller with saturation and anti-windup in conjunction with preview. Towards SLA Policy Refinement for QoS Management in Software-Defined Networking The vehicle dynamics model is a point mass representation of the vehicle and includes differentials, wheel inertia, a Coulomb and viscous-friction-based brakemodel, rolling resistance, aerodynamic drag, and tire slip.