Optimization of antenna array layouts in planar and non-planar geometries for DoA estimation
(2024) EITM01 20241Department of Electrical and Information Technology
- Abstract
- The use of antenna arrays allows for direction-of-arrival (DoA) estimation in a radar system. The antenna elements in an array are conventionally positioned equidistantly on a grid, but they can also be placed relatively freely to form sparse antenna arrays. These non-conventional layouts could potentially offer improved DoA capabilities. In this work, a particle swarm optimization algorithm has been implemented to find optimal array layouts according to a set of requirements. These requirements have been based on the ambiguity function, which is related to the risk of error and to the resolution of the DoA estimations. The results indicate that the implemented method works as intended, and that it can be applied to two different... (More)
- The use of antenna arrays allows for direction-of-arrival (DoA) estimation in a radar system. The antenna elements in an array are conventionally positioned equidistantly on a grid, but they can also be placed relatively freely to form sparse antenna arrays. These non-conventional layouts could potentially offer improved DoA capabilities. In this work, a particle swarm optimization algorithm has been implemented to find optimal array layouts according to a set of requirements. These requirements have been based on the ambiguity function, which is related to the risk of error and to the resolution of the DoA estimations. The results indicate that the implemented method works as intended, and that it can be applied to two different geometries. The first geometry is planar and circular, while the second consists of multiple planar surfaces, positioned to cover 360° in azimuth. A comparison was made between the noise performance of the optimized arrays and that of a reference array, in which only small differences were seen. The chosen method was implemented to work for any given dimension of the specified geometries, and could potentially be used for restrictive dimensions and a varying number of antenna elements. (Less)
- Popular Abstract
- The existence of electromagnetic (EM) waves was first experimentally shown in the late 19th century, a discovery that would revolutionize the way we communicate. From the first radio to the wireless systems of today, our world has become increasingly dependent on EM waves in our everyday lives. Parallel to the communication applications, this discovery also led to the development of radar. A radar system works by sending out signals through an antenna, and listening for a reflected signal. Oftentimes, it is desired to determine in what direction you are receiving the reflected signal in order to know where your target is. To explain how this is done, it is worth taking a look at ourselves. Without much thought, humans are able to locate... (More)
- The existence of electromagnetic (EM) waves was first experimentally shown in the late 19th century, a discovery that would revolutionize the way we communicate. From the first radio to the wireless systems of today, our world has become increasingly dependent on EM waves in our everyday lives. Parallel to the communication applications, this discovery also led to the development of radar. A radar system works by sending out signals through an antenna, and listening for a reflected signal. Oftentimes, it is desired to determine in what direction you are receiving the reflected signal in order to know where your target is. To explain how this is done, it is worth taking a look at ourselves. Without much thought, humans are able to locate sounds that we hear, which is mainly thanks to us having two ears. The direction-of-arrival (DoA) can be estimated by comparing how loud the signal is at each ear, and by sensing the time difference between them. The latter is essentially what is used by a radar system to do a similar estimation. For the same reason that we need two ears to determine the DoA, a radar will also need more than one antenna to do the same thing. Something that we are unable to experience is that the ability to determine the DoA is dependent on how the antennas are placed. For instance, if the antennas are badly placed, it can “sound” like a single signal is coming from multiple directions. In order to help with this problem, a radar system often has more than two antennas and these antennas can be positioned in different ways. This raises a question, which is also the aim of this thesis: how do you place the antennas to get as good DoA estimation as possible? Unfortunately, there is often an infinite amount of possible solutions to this problem, so it is impossible to try all of them. Instead, we can once again be inspired by a natural phenomenon. This time, we look at the way that bird flocks collaborate when they hunt for insects. In such a flock, each individual bird wants to find a position in the air with the most amount of food. If they have once found such a place, they will tend to return there. At the same time, they will also look at where the other birds are flying, as there could be another spot with even more insects. While this might sound abstract, this same approach can be used to find good antenna positions. By allowing an algorithm to search for antenna positions, it was possible to find satisfactory layouts in different types of geometries. This method could prove useful when it is needed to place antennas in restrictive geometries, where it is difficult to find suitable placements. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9170053
- author
- Zanichelli, Johan LU
- supervisor
- organization
- course
- EITM01 20241
- year
- 2024
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Radar, DoA, MIMO, Particle Swarm Optimization, Sparse Arrays
- report number
- LU/LTH-EIT 2024-1010
- language
- English
- id
- 9170053
- date added to LUP
- 2024-08-19 14:15:29
- date last changed
- 2024-08-19 14:15:29
@misc{9170053,
abstract = {{The use of antenna arrays allows for direction-of-arrival (DoA) estimation in a radar system. The antenna elements in an array are conventionally positioned equidistantly on a grid, but they can also be placed relatively freely to form sparse antenna arrays. These non-conventional layouts could potentially offer improved DoA capabilities. In this work, a particle swarm optimization algorithm has been implemented to find optimal array layouts according to a set of requirements. These requirements have been based on the ambiguity function, which is related to the risk of error and to the resolution of the DoA estimations. The results indicate that the implemented method works as intended, and that it can be applied to two different geometries. The first geometry is planar and circular, while the second consists of multiple planar surfaces, positioned to cover 360° in azimuth. A comparison was made between the noise performance of the optimized arrays and that of a reference array, in which only small differences were seen. The chosen method was implemented to work for any given dimension of the specified geometries, and could potentially be used for restrictive dimensions and a varying number of antenna elements.}},
author = {{Zanichelli, Johan}},
language = {{eng}},
note = {{Student Paper}},
title = {{Optimization of antenna array layouts in planar and non-planar geometries for DoA estimation}},
year = {{2024}},
}