Sharif Digital Repository

It is of fundamental importance to find algorithms obtaining optimal performance for learning of statistical models in distributed and communication limited systems. Aiming at characterizing the optimal strategies, we consider learning of Gaussian Processes (GP) in distributed systems as a pivotal example. We first address a very basic problem: how many bits are required to estimate the inner-products of some Gaussian vectors across distributed machines? Using information theoretic bounds, we obtain an optimal solution for the problem which is based on vector quantization. Two suboptimal and more practical schemes are also presented as substitutes for the vector quantization scheme. In... 

The sensor-source geometry has a significant effect on accuracy of source localization problems. In a sensor placement problem, one attempts to optimally place the sensors in the surveillance area so as to optimize a performance criterion. Sensor placement methods mainly solve the associated problems without taking any specific constraint on permissible location of sensors into account. In practical applications, however, possible location for deployment of the sensors are subject to such limitations as environmental, industrial, and communication constraints, which affects the optimal sensor-source geometry. In this article, we consider the problem of optimal sensor placement, based on time... 

In this paper, a model for the optimal placement of contingency-constrained phasor measurement units (PMUs) in electric power networks is presented. The conventional complete observability of power networks is first formulated and then, different contingency conditions in power networks including measurement losses and line outages are added to the main model. The communication constraints which would limit the maximum number of measurements associated with each installed PMU is considered as measurement limitations. The relevant formulations are also proposed to make the model more comprehensive. The IEEE standard test systems are examined for the applicability of proposed model. The...