Date

5-1-2018

Document Type

Thesis

Degree

Master of Science

Department

Electrical and Computer Engineering

First Adviser

Blum, Rick S.

Abstract

A modern wide area monitoring system (WAMS) supporting the future grid will include a vastly improved information and communications functionality that allows service providers to sense, monitor, and manage electricity flows throughout the grid [1]. While the cyber physical integration improves the performance and efficiency of the grid, it increases its vulnerability to potential cyber-attacks. Security of cyber-physical systems in the context of the power grid has received significant attention [2] - [4]. In this Master’s Thesis, we provide two sets of tests for the existing detection scheme which address the problem of cybersecurity in smart grid networks involving PMUs (Phasor Measurement Units) taking into account the dynamical nature of the power system [5]. A PMU can record synchrophasors at a high sampling rate, and the measurements are synchronized to an absolute time reference provided by the GPS. In general, a GPS spoofing attack refers to deception of the GPS receiver by transmitting spurious signals resembling the normal GPS signals, leading to timing synchronization errors [6]. In an electric grid with PMUs, GPS spoofing results in counterfeit time stamps at the synchrophasors and is referred to as a timing synchronization attack (TSA) [7]. While a TSA only alters the time stamps without inducing changes in the actual measurements, it results in confusing the grid command center with erroneous system operation status. Evaluating the threat to synchrophasor measurements and the countermeasures to combat TSAs have received considerable attention in the existing literature [8]- [11]. In this Master’s Thesis, we propose two sets of tests for the existing GPS spoofing attack detection scheme [5] to check the performance of the scheme under different circumstances. In the first sets of test (α test), we simulate the 9−bus, 3−generator IEEE power system in MATLAB and using this simulated system, we apply our test for checking the performance of detection scheme. In the other word, in the α test, we use the simulated data and test the detection scheme. We will investigate the performance of the detection scheme due to changes in attack parameter (which is the time delay made by attacker to spoof the authenticated GPS signal), window size (which is the number of sample in a window we want to check), and examine the performance in the case of unknown time of attack (which means that the time of attack is not known in the detection scheme). The second half of the thesis is dedicated to the second sets of test (β test), in which we use the data from Real-Time Renewable Microgrid Test-bed lab at Lehigh University. The key difference between α and β testing is the data used for the test. In the α test, data comes from the simulated power grid in MATLAB and in the β test, the data comes from one of the labs at Lehigh.

Share

COinS