Recurrent Deep Learning for Beam Pattern Synthesis in Optimized Antenna Arrays
This work proposes and describes a deep learning-based approach utilizing recurrent neural networks (RNNs) for beam pattern synthesis considering uniform linear arrays. In this particular case, the deep neural network (DNN) learns from previously optimized radiation patterns as inputs and generates...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/15/1/204 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841549433443975168 |
|---|---|
| author | Armando Arce Fernando Arce Enrique Stevens-Navarro Ulises Pineda-Rico Marco Cardenas-Juarez Abel Garcia-Barrientos |
| author_facet | Armando Arce Fernando Arce Enrique Stevens-Navarro Ulises Pineda-Rico Marco Cardenas-Juarez Abel Garcia-Barrientos |
| author_sort | Armando Arce |
| collection | DOAJ |
| description | This work proposes and describes a deep learning-based approach utilizing recurrent neural networks (RNNs) for beam pattern synthesis considering uniform linear arrays. In this particular case, the deep neural network (DNN) learns from previously optimized radiation patterns as inputs and generates complex excitations as output. Beam patterns are optimized using a genetic algorithm during the training phase in order to reduce sidelobes and achieve high directivity. Idealized and test beam patterns are employed as inputs for the DNN, demonstrating their effectiveness in scenarios with high prediction complexity and closely spaced elements. Additionally, a comparative analysis is conducted among the three DNN architectures. Numerical experiments reveal improvements in performance when using the long short-term memory network (LSTM) compared to fully connected and convolutional neural networks. |
| format | Article |
| id | doaj-art-3321bfafacca4b84936fd74151d4d305 |
| institution | Kabale University |
| issn | 2076-3417 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj-art-3321bfafacca4b84936fd74151d4d3052025-01-10T13:14:47ZengMDPI AGApplied Sciences2076-34172024-12-0115120410.3390/app15010204Recurrent Deep Learning for Beam Pattern Synthesis in Optimized Antenna ArraysArmando Arce0Fernando Arce1Enrique Stevens-Navarro2Ulises Pineda-Rico3Marco Cardenas-Juarez4Abel Garcia-Barrientos5Consejo Nacional de Humanidades, Ciencia y Tecnología (CONAHCYT), Facultad de Ciencias, Universidad Autónoma de San Luis Potosí (UASLP), San Luis Potosí 78295, MexicoCentro de Investigaciones en Óptica (CIO), A.C., León 37150, MexicoFacultad de Ciencias, Universidad Autónoma de San Luis Potosí (UASLP), San Luis Potosí 78295, MexicoFacultad de Ciencias, Universidad Autónoma de San Luis Potosí (UASLP), San Luis Potosí 78295, MexicoFacultad de Ciencias, Universidad Autónoma de San Luis Potosí (UASLP), San Luis Potosí 78295, MexicoFacultad de Ciencias, Universidad Autónoma de San Luis Potosí (UASLP), San Luis Potosí 78295, MexicoThis work proposes and describes a deep learning-based approach utilizing recurrent neural networks (RNNs) for beam pattern synthesis considering uniform linear arrays. In this particular case, the deep neural network (DNN) learns from previously optimized radiation patterns as inputs and generates complex excitations as output. Beam patterns are optimized using a genetic algorithm during the training phase in order to reduce sidelobes and achieve high directivity. Idealized and test beam patterns are employed as inputs for the DNN, demonstrating their effectiveness in scenarios with high prediction complexity and closely spaced elements. Additionally, a comparative analysis is conducted among the three DNN architectures. Numerical experiments reveal improvements in performance when using the long short-term memory network (LSTM) compared to fully connected and convolutional neural networks.https://www.mdpi.com/2076-3417/15/1/204deep learningneural networksantenna arraysantenna radiation patternssynthesisoptimization |
| spellingShingle | Armando Arce Fernando Arce Enrique Stevens-Navarro Ulises Pineda-Rico Marco Cardenas-Juarez Abel Garcia-Barrientos Recurrent Deep Learning for Beam Pattern Synthesis in Optimized Antenna Arrays Applied Sciences deep learning neural networks antenna arrays antenna radiation patterns synthesis optimization |
| title | Recurrent Deep Learning for Beam Pattern Synthesis in Optimized Antenna Arrays |
| title_full | Recurrent Deep Learning for Beam Pattern Synthesis in Optimized Antenna Arrays |
| title_fullStr | Recurrent Deep Learning for Beam Pattern Synthesis in Optimized Antenna Arrays |
| title_full_unstemmed | Recurrent Deep Learning for Beam Pattern Synthesis in Optimized Antenna Arrays |
| title_short | Recurrent Deep Learning for Beam Pattern Synthesis in Optimized Antenna Arrays |
| title_sort | recurrent deep learning for beam pattern synthesis in optimized antenna arrays |
| topic | deep learning neural networks antenna arrays antenna radiation patterns synthesis optimization |
| url | https://www.mdpi.com/2076-3417/15/1/204 |
| work_keys_str_mv | AT armandoarce recurrentdeeplearningforbeampatternsynthesisinoptimizedantennaarrays AT fernandoarce recurrentdeeplearningforbeampatternsynthesisinoptimizedantennaarrays AT enriquestevensnavarro recurrentdeeplearningforbeampatternsynthesisinoptimizedantennaarrays AT ulisespinedarico recurrentdeeplearningforbeampatternsynthesisinoptimizedantennaarrays AT marcocardenasjuarez recurrentdeeplearningforbeampatternsynthesisinoptimizedantennaarrays AT abelgarciabarrientos recurrentdeeplearningforbeampatternsynthesisinoptimizedantennaarrays |