This paper considers a distributed optimization problem over a multi-agent network, in which the objective function is a sum of individual cost functions at the agents. We focus on the case when communication between the agents is described by a directed graph. Existing distributed optimization algorithms for directed graphs require at least the knowledge of the neighbors' out-degree at each agent (due to the requirement of column-stochastic matrices). In contrast, our algorithm requires no such knowledge. Moreover, the proposed algorithm achieves the best known rate of convergence for this class of problems, $O(\mu^k)$ for $0 \lt \mu \lt {1}$, where $k$ is the number of iterations, given that the objective functions are strongly-convex and have Lipschitz-continuous gradients. Numerical experiments are also provided to illustrate the theoretical findings.

The paper considers the problem of a leader that seeks to optimally influence the opinions of agents in a directed network through connecting with a limited number of the agents ("direct followers"), possibly in the presence of a fixed competing leader. The settings involving a single leader and two competing leaders are unified into a general combinatoric optimization problem, for which two heuristic approaches are developed. The first approach is based on a convex relaxation scheme, possibly in combination with the $\ell_1$-norm regularization technique, and the second is based on a greedy selection strategy. The main technical novelties of this work are in the establishment of supermodularity of the objective function and convexity of its continuous relaxation. The greedy approach is guaranteed to have a lower bound on the approximation ratio sharper than $(1-1/e)$, while the convex approach can benefit from efficient (customized) numerical solvers to have practically comparable solutions possibly with faster computation times. The two approaches can be combined to provide improved results. In numerical examples, the approximation ratio can be made to reach $90$% or higher de

A pitch-plane model of a supercavitating vehicle is developed to account for the time delay in the propagation of the cavitator action from the vehicle nose to the vehicle aft. This time delay is an advection delay, which is on the order of the vehicle length divided by its speed. Unlike previous models with time-delay effects, in the present model, the effect of cavity rotation during forward motion is incorporated. Stability analyses and feedback control designs are carried out using this model. It is found that the open-loop system with and without the time delay is unstable. Feedback control laws that stabilize the delay-free system model are found to be ineffective in the presence of the time delay. The authors show that the delay leads to destabilization of the supercavitating vehicle dynamics in the sense that an operation at a stable trim condition is replaced by a stable limit-cycle motion that is commonly referred to as tail-slap. Feedback control designs are carried out by taking into account the time delay, and it is demonstrated that the supercavitating vehicle can be stabilized at trim conditions inside and outside the cavity. By using numerical studies of the nonlinear delay-dependent pitch-plane model of the supercavitating vehicle, the effectiveness of the new control designs are demonstrated.

Cloud computing is a novel paradigm, where the limitations of ubiquitous connected devices in terms of computing, data access, networking and storage are solved through the use of cloud infrastructure. The pervasive adoption of cloud computing results in a rising carbon footprint due to the high energy consumption of computing servers. This negatively affects the environment and entails an associated increase in electricity costs and consequently operational costs. Many works proposed scheduling algorithms using software-centric power models in order to predict electric power consumption in underlying data centers and to schedule cloud tasks so as to reduce energy consumption. Linear models which are based on the lowestand highest-power data points (referred to here as the “Power Endpoints Model” -PEM) and the simple linear regression (SLR) model are the most used in the literature. However, these models have traditionally been evaluated using different environments, experimental setups, workloads, and error calculation formulas. In this paper, a unified classification and evaluation for these linear power models is presented, under unified setup, benchmarking applications, and error formula with the main goal being to achieve an objective comparison. A new power model is proposed, named Locally Corrected Multiple Linear Regression (LCMLR), in order to increase prediction accuracy. A simulation framework for a cloud energy-aware scheduler is introduced. The framework combines the Energy-Aware Task Scheduling on Cloud Virtual Machines (EATSVM) with the LC-MLR power model, and facilitates performance measurement for cloud data centers. The scheduler with the new power model increases energy efficiency without degrading the qualities of service of the system. The workloads used for performance evaluation and comparisons in this work are generated using a diverse set of applications. The results show that LC-MLR outperforms the most-used models for simulation of power consumption of cloud data centers. The detailed performance analysis is elaborated in the paper.

The paper studies an extension to nonlinear systems of a recently proposed approach to the definition of modal participation factors. A definition is given for local mode-in-state participation factors for smooth nonlinear autonomous systems. While the definition is general, the resulting measures depend on the assumed uncertainty law governing the system initial condition, as in the linear case. The work follows Hashlamoun et al. (IEEE Trans Autom Control 54(7):1439–1449 2009) in taking a mathematical expectation (or set-theoretic average) of a modal contribution measure over an uncertain set of system initial state. Poincaré linearization is used to replace the nonlinear system with a locally equivalent linear system. It is found that under a symmetry assumption on the distribution of the initial state, the tractable calculation and analytical formula for mode-in-state participation factors found for the linear case persists to the nonlinear setting. This paper is dedicated to the memory of Professor Ali H. Nayfeh.

This paper studies distributed optimization problems where a network of agents seeks to minimize the sum of their private cost functions. We propose new algorithms based on the distributed subgradient method and the finite-time consensus protocol introduced by Sundaram and Hadjicostis (2007). In our first algorithm, the local optimization variables are updated cyclically through a subgradient step while the consensus variables follow a usual consensus protocol periodically interrupted by a predictive consensus estimate re-set operation. For convex cost functions with bounded subgradients, this algorithm is guaranteed to converge to a certain range of the optimal value if using a constant step size or to the optimal value if a diminishing step size is in place. For differentiable cost functions whose sum is convex and has a Lipschitz continuous gradient, convergence to the optimal value can be ensured when using a constant step size, even if some of the individual cost functions are nonconvex. In addition, exponential convergence to the optimal solution is achieved when the global cost function is further assumed to be strongly convex. In these cases, the local opt