Sub-sampled exponential analysis applied to sparse planar antenna array configurations
| dc.contributor.advisor | De Villiers, Dirk | en_ZA |
| dc.contributor.author | Raal, Ridalise | en_ZA |
| dc.contributor.other | Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering. | en_ZA |
| dc.date.accessioned | 2023-03-03T07:27:56Z | en_ZA |
| dc.date.accessioned | 2023-05-18T07:13:24Z | en_ZA |
| dc.date.available | 2023-03-03T07:27:56Z | en_ZA |
| dc.date.available | 2023-05-18T07:13:24Z | en_ZA |
| dc.date.issued | 2023-03 | en_ZA |
| dc.description | Thesis (PhD)--Stellenbosch University, 2023. | en_ZA |
| dc.description.abstract | ENGLISH ABSTRACT: Considerable research efforts are currently ongoing for the planned Mid-Frequency perture Array of the Square Kilometre Array project, which is set to operate from 450- 1450 MHz. An important design criterion is the aperture array configuration, for which a choice has not yet been made. Until recently, the focus was mainly on dense regular or sparse random array configurations, since the strong grating lobes that are present in the beam pattern of a sparse regular array have made them historically unattractive. Random placement of antenna elements in a sparse array reduces the impact of grating lobes by spreading out their energy into sidelobes. A sparse configuration is attractive since fewer elements are required for the same effective area, which leads to an improved angular resolution and lower cost when compared to dense regular arrays. The increased spacing between elements additionally leads to reduced mutual coupling effects. A recently developed algorithm known as VEXPA (Validated EXPonential Analysis) allows recovery from aliasing after sub-Nyquist sampling of data. VEXPA was previously applied to the direction-of-arrival estimation problem when using a sparse regular linear array in a coprime configuration which consists of two sub-arrays that are designed to have the grating lobes of each add destructively. Whereas traditional coprime arrays require a computationally expensive pairing and matching step between solutions of the sub-arrays, VEXPA makes use of a shared linear structure of equations between them, which automatically pairs their output. It also provides additional features compared to traditional DOA algorithms, such as validation of the output and the ability to estimate the number of incoming signals of the problem without any prior knowledge. In this dissertation, VEXPA is adapted and used for two different applications concerning sparse array configurations. For the first application, the sparse regular linear antenna array case (1-D scanning) is extended to the sparse regular planar antenna array case (full 2-D scanning). Additionally, a non-collinear configuration is presented which allows the algorithm to be used with an existing sparse regular planar array that was not originally designed with the coprime configuration in mind. A small engineering test system was constructed and measured in an anechoic chamber. Due to limitations of the available measuring equipment, it was necessary to construct the planar array demonstrator to be sparse along one dimension and dense in the second dimension. Comparisons to full-wave simulations showed that the main source of error was due to mutual coupling between the densely spaced antenna elements, emphasising the importance of the use of a sparse antenna array configuration. The second application considers antenna position estimation in a sparse random array. The construction cost of an aperture array consisting of a large amount of antenna elements can be significantly reduced by relaxing the required placement accuracy, and allowing for errors when connecting the elements to the system back-end. The idea is to find the antenna element positions after the installation phase through low-cost in-situ measurement campaigns where odd harmonic signals are transmitted from an Unmanned Aerial Vehicle (UAV) flying above the array with a known position in the sky. The problem is formulated as an inverse direction-of-arrival method and a simplified version of VEXPA which does not require validation is used to solve the resulting sub-sampled exponential analysis problem. The method was first developed conceptually by assuming that the UAV is in the farfield of the array, which allows its zenith angle to be used to model the coprime scaling parameters that are required for recovery from aliasing. For large arrays, the UAV is in the near-field with a spherical phase front incident on elements in the array, which leads to the time delay between elements being described by a non-linear equation. Thus, an iterative approach was followed to estimate the antenna positions from a linearised approximation of the near-field equation. For each iteration, coprime scale parameters were extracted to allow for de-aliasing and subsequent updating of the approximation with a new estimation of the antenna position, until the solution converges. An optimal flight path for the UAV was considered and the method was shown to work well for toy examples simulated in Matlab. | en_ZA |
| dc.description.abstract | AFRIKAANS OPSOMMING: Aansienlike navorsingspogings is tans aan die gang vir die beplande Mid-Frekwensie Stralingsvlak Samestelling van die “Square Kilometer Array” projek, wat veronderstel is om van 450-1450 MHz te werk. ’n Belangrike ontwerpkriterium is die konfigurasie van die samestelling, waarvoor ’n keuse nog nie gemaak is nie. Tot onlangs was die fokus hoofsaaklik op digte-reëlmatige of yl-lukrake konfigurasies, aangesien die sterk roosterlobbe wat in die stralingspatroon van ’n yl-reëlmatige samestelling teenwoordig is, hulle histories onaantreklik gemaak het. Willekeurige plasing van antenna elemente in ’n yl samestelling verminder die impak van roosterlobbe deur hul energie in sylobbe in te versprei. ’n Yl konfigurasie is aantreklik aangesien minder elemente benodig word vir dieselfde effektiewe area, wat lei tot ’n verbeterde hoekresolusie en laer koste in vergelyking met digtereëlmatige samestellings. Die groter spasiëring tussen elemente lei ook tot verminderde wedersydse koppelingseffekte. ’n Algoritme was onlangs ontwikkel wat bekendstaan as VEXPA (“Validated EXPonential Analysis”) wat toelaat dat dubbelsinnige resultate as gevolg van sub-Nyquist-monstering uniek bepaal kan word. VEXPA is voorheen toegepas op die invalshoekafskattingsprobleem van ’n yl-reëlmatige lineêre samestelling in ’n kopriem-konfigurasie wat uit twee sub-samestellings bestaan wat ontwerp is sodat die roosterlobbe van elkeen destruktief inmeng met mekaar. Terwyl tradisionele kopriem-samestellings ’n duur paring- en passingsstap tussen oplossings van die sub-samestellings vereis, maak VEXPA gebruik van gedeelde lineêre vergelykings tussen hulle, wat hul oplossings outomaties pas. Dit bied ook bykomende werkverrigtinge in vergelyking met tradisionele invalshoekafskattingsalgoritmes, soos byvoorbeeld validasie van die oplossing en die vermoë om die aantal inkomende seine van die probleem af te skat sonder enige voorafkennis. In hierdie proefskrif word VEXPA aangepas en gebruik vir twee verskillende toepassings rakende yl samestellings. Vir die eerste toepassing word die yl-reëlmatige lineêre antenna samestelling (1-D-skandering) uitgebrei na die yl-reëlmatige planêre antenna samestelling (volle 2-D-skandering). Daarbenewens word ’n nie-kollineêre konfigurasie aangebied wat die algoritme toelaat om gebruik te word met ’n bestaande yl-reëlmatige planêre samestelling wat nie oorspronklik ontwerp is met die kopriem-konfigurasie in gedagte nie. ’n Klein ingenieurstoetsstelsel is gebou en gemeet in ’n weerkaatsvrye kamer. As gevolg van beperkings van die beskikbare meettoerusting, was dit nodig om die planêre samestellingstoestel te ontwerp om yl langs een dimensie en dig langs die tweede dimensie te wees. Vergelykings met volgolfsimulasies het getoon dat die hoofbron van foute te wyte was aan wedersydse koppeling tussen die dig-gespasiëerde antenna elemente, wat die belangrikheid van die gebruik van ’n yl antenna samestelling beklemtoon. Die tweede toepassing oorweeg antenna posisie afskatting in ’n yl-lukrake samestelling. Die konstruksiekoste van ’n stralingsvlaksamestelling wat uit ’n groot hoeveelheid antenna elemente bestaan, kan aansienlik verminder word deur die vereiste plasingsakkuraatheid te verslap, en voorsiening te maak vir foute wanneer die elemente aan die ontvangerstelsel gekoppel word. Die idee is om die antenna elementposisies na die installasiefase te vind deur middel van laekoste in-situ metingsveldtogte waar onewe harmonieke uitgesaai word vanaf ’n onbemande lugvoertuig wat bo die samestelling vlieg met ’n bekende posisie in die lug. Die probleem is geformuleer as ’n inverse rigtingafskattingsprobleem en ’n vereenvoudigde weergawe van VEXPA wat nie validasie vereis nie, word gebruik om die gevolglike ondergemonsterde eksponensiële analise probleem op te los. Die metode is eerstens konseptueel ontwikkel deur aan te neem dat die onbemande lugvoertuig in die ver-veld van die samestelling is, wat dit moontlik maak om sy senithoek te gebruik om die kopriem-skaleringsparameters te modelleer wat benodig word vir herstel van aliasering. Vir groot samestellings is die onbemande lugvoertuig in die naby-veld met ’n sferiese fase-front wat inval op elemente in die samestelling, en gevolglik word die tydsvertraging tussen elemente deur ’n nie-lineêre vergelyking beskryf. ’n Iteratiewe metode word dus gebruik om die antenna posisies af te skat vanaf ’n gelineariseerde benadering van die naby-veld vergelyking. Vir elke iterasie is kopriem-skaleringsparameters onttrek om voorsiening te maak vir die herstel van aliasering en daaropvolgende opdatering van die benadering met ’n nuwe afskatting van die antenna posisie, totdat die oplossing konvergeer. ’n Optimale vlugpad vir die onbemande lugvoertuig is oorweeg en die metode werk goed vir speelgoedvoorbeelde wat in Matlab gesimuleer is. | af_ZA |
| dc.description.version | Masters | en_ZA |
| dc.format.extent | ccvb | en_ZA |
| dc.identifier.uri | http://hdl.handle.net/10019.1/127273 | |
| dc.language.iso | en_ZA | en_ZA |
| dc.language.iso | en_ZA | en_ZA |
| dc.publisher | Stellenbosch : Stellenbosch University | en_ZA |
| dc.rights.holder | Stellenbosch University | en_ZA |
| dc.subject.lcsh | Aperture antennas | en_ZA |
| dc.subject.lcsh | Algorithm | en_ZA |
| dc.subject.lcsh | Signal processing | en_ZA |
| dc.title | Sub-sampled exponential analysis applied to sparse planar antenna array configurations | en_ZA |
| dc.type | Thesis | en_ZA |
Files
Original bundle
1 - 1 of 1
Loading...
- Name:
- louw_subsampled_2023.pdf
- Size:
- 16.08 MB
- Format:
- Adobe Portable Document Format
- Description: