Beam Forming from Randomly Spaced Elements in a Linear Array
The activation of antenna elements located at random positions on a linear aperture with constant currents so as to collaboratively transmit or receive in preferred directions is investigated. Assuming the elements are located in continuum spatially, the objective is to determine a probability distribution function for the element positions and characterize the random features of the resulting beam patterns. Two approaches are presented, one analytical and another that computationally determines the optimal positions using a firefly algorithm, where each firefly models an array with finite number of elements. The analytical model for the probability distribution is constrained to match the expected value of the beam response to the directivity function of a uniformly spaced element Dolph-Chebyshev array with a prescribed side lobe level. The variance in the side lobe amplitudes is shown to decrease as 1/(2*sqrt(N)) where N is the number of elements. The firefly algorithm simulates the exploratory and convergent behavior of a swarm of fireflies, attracted to one another through the light they emit, which is emulated through a fitness function. Multiple ensembles from the computational model yield a set of viable arrays that are shown to exhibit a lower variance than the analytical model. The computational model also yields a set of conditional probability distributions that capture the correlation between neighboring inter-element distances. Arrays constructed by sampling from these conditional distributions yield a reduced variance in the side lobe amplitudes, using fewer elements than that required under the independent sampling assumption of the analytical model.
Date and Time
Location
Hosts
Registration
- Date: 04 May 2017
- Time: 06:30 PM to 08:00 PM
- All times are (GMT-05:00) US/Eastern
-
Add Event to Calendar
- 2 Court Street
- Nashua, New Hampshire
- United States
- Building: Nashua Library
- Room Number: Downstairs Lecture Room
- Starts 06 April 2017 09:12 AM
- Ends 04 May 2017 06:30 PM
- All times are (GMT-05:00) US/Eastern
- No Admission Charge
Speakers
Nicholas Misiunas of University of Massachusetts-Lowell, Center for Advanced Computation and Telecommunications
Biography:
Nicholas Misiunas got his bachelor's in electrical engineering from Purdue in 2009 before coming to University of Massachussets at Lowell where he obtained his master's in Electrical and Computer Engineering and stayed to pursue his PhD, of which he is currently a PhD candidate. His focuses are metaheuristic algorithms and data analysis and modeling.
Agenda
Meeting start - 6:30 PM
The presentation will start little after 6:30 pm and will end around 7:30 pm.
Q&A session.
Meeting ends at 8:00 PM.