Performance Evaluation of Conventional and AI-Aided Detectors for AmBC in Multi-Tap Channels
9th International Balkan Conference on Communications and Networking, Balkancom 2026, Ulcinj, Karadağ, 16 - 19 Haziran 2026, ss.60-65, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/balkancom71095.2026.11605022
- Basıldığı Şehir: Ulcinj
- Basıldığı Ülke: Karadağ
- Sayfa Sayıları: ss.60-65
- Anahtar Kelimeler: Ambient backscatter, logistic regression, non-coherent detection, Rayleigh fading
- Kocaeli Üniversitesi Adresli: Evet
Özet
Ambient Backscatter Communication (AmBC) enables battery-free Internet of Things (IoT) devices by modulating and reflecting existing Radio Frequency (RF) signals. A central challenge is reliable tag-bit reception over multitap fading channels. This paper benchmarks four detection strategies, including a pilot-aided Energy Detector (ED), a Joint Correlation-Energy Detector (JCED), a genie-aided coherent Optimal Detector (OPT), and a 7-feature Logistic Regression Detector (LR). We evaluate these detectors under three ambient signal types: unmodulated carrier, On-Off Keying (OOK), and Binary Phase-Shift Keying (BPSK). Furthermore, we consider multi-tap Rayleigh fading channels with varying channel taps and reflection coefficients. Simulations confirm that OPT consistently provides the lowest Bit Error Rate (BER). Among practical non-coherent methods, LR outperforms ED/JCED (especially for OOK/BPSK and multipath), while JCED performs better than ED mainly in multipath settings.