Minimum Distance Receiver using Dynamic Programming-Based Sparse Samples
Document Type
Conference Proceeding
Date of Original Version
1-1-2024
Abstract
Sparse samples that maximize the detection of known deterministic signals in autoregressive Gaussian noise of order 1 can be precisely determined using dynamic programming. This study investigates the efficacy of these dynamic programming-based samples in signal classification using the minimum distance receiver algorithm. Initially, we employ dynamic programming to design sparse samples tailored for detecting a sinusoid with a predefined frequency, referred to as the design frequency. Subsequently, we consider sinusoids with five distinct frequencies: two above the design frequency, one matching it, and two below it. The minimum distance receiver algorithm is then utilized to classify these sinusoids in the presence of noise. Comparative analysis is conducted to assess the classification accuracy of the dynamic programming-based samples in comparison to other commonly used sparse sampling methods. Simulation results demonstrate that the detection-optimized samples exhibit superior performance in terms of classification accuracy when compared to other standard sparse samples.
Publication Title, e.g., Journal
2024 IEEE 5th World AI Iot Congress Aiiot 2024
Citation/Publisher Attribution
Kalaoun, Zaynah L., Kaushallya Adhikari, and Steven Kay. "Minimum Distance Receiver using Dynamic Programming-Based Sparse Samples." 2024 IEEE 5th World AI Iot Congress Aiiot 2024 (2024). doi: 10.1109/AIIoT61789.2024.10579019.