Non-coherent UWB Receiver Threshold Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-coherent ultra-wideband receivers face challenges in distinguishing binary '1' from '0' due to Additive White Gaussian Noise, especially with lower hardware and computational complexities compared to coherent receivers, and are susceptible to varying noise levels.
Innovation Solution
A non-coherent ultra-wideband receiver uses a lookup table with sorted time domain samples and estimated signal-to-noise ratio to determine a threshold for distinguishing between binary '1' and '0' by comparing selected samples, improving noise resistance and performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If energy detection is used in non-coherent UWB receivers, then hardware and computational complexities are reduced, but the receiver becomes susceptible to varying background noise and interference levels
Solution Approach 1:
The received signal is divided into multiple samples within each symbol period, and these samples are sorted by magnitude to separate the signal component from noise. This segmentation approach allows the receiver to identify the largest sample as the signal while treating smaller samples as noise, thereby maintaining low complexity while improving noise resistance.
Solution Approach 2:
The invention changes the detection parameter from simple energy summation to sorted sample magnitude comparison. By sorting samples and selecting specific samples based on their magnitude ranks, the receiver adapts to varying noise conditions without requiring complex adaptive algorithms, thus maintaining low complexity while improving reliability.
2Power
If the detection window is increased in energy detection receivers, then more signal energy is captured, but more noise samples are collected leading to a higher noise floor
Solution Approach 1:
Within the detection window, the received samples are segmented and sorted by magnitude. This allows the receiver to selectively process only the largest samples that correspond to the signal, while excluding smaller noise samples, thereby capturing signal energy without proportionally increasing noise impact.
Solution Approach 2:
The receiver processes only a subset of the total samples (specifically, the largest k samples out of N total samples) rather than all samples. This partial action approach captures sufficient signal energy while limiting noise accumulation, achieving an optimal balance between signal capture and noise rejection.
3Device complexity
If non-coherent detection based on amplitude only is used, then hardware complexity is reduced, but the ability to distinguish binary '1' from '0' in noisy channels deteriorates
Solution Approach 1:
The receiver performs preliminary sorting of all received samples by magnitude before making the binary decision. This preliminary action organizes the data in a way that makes the signal component easily identifiable as the largest sample, improving measurement precision while keeping the final decision logic simple and low-complexity.
Solution Approach 2:
The invention creates an ordered copy or ranking of the received samples based on magnitude, where the signal corresponds to the largest value. This copying approach transforms the detection problem into a simple comparison task, maintaining low hardware complexity while significantly improving the ability to distinguish binary states in noisy conditions.
Data Source
AI summary
The non-coherent ultra-wideband receiver receives an ultra-wideband (UWB) signal, consisting of pulses (or “symbols”) and uses on-off keying (OOK) modulation so that when a binary “0” is transmitted, the receiver collects noise-only samples. The receiver collects samples during the symbol (pulse) duration and sorts the samples by magnitude of voltage or energy. The receiver uses the known transmission rate and the estimated signal-to-noise ratio to retrieve a sample index from a look-up table. The receiver then compares the signal sample at the index value with a predetermined threshold voltage (or energy). If the selected sample exceeds the threshold, then it is assumed that all succeeding samples also exceed the threshold (assuming the sort is in ascending magnitude) and the pulse is present and binary “1”. Otherwise, the pulse is absent in the sampling period, and binary “0”. The process is repeated for the signal duration.