Direct Target Localization With Low-Bit Quantization in Wireless Sensor Networks
Guoxin Zhang,Wei Yi,Michail Matthaiou,Pramod K. Varshney
DOI: https://doi.org/10.1109/tsp.2024.3411672
IF: 4.875
2024-07-26
IEEE Transactions on Signal Processing
Abstract:Wireless sensor networks (WSNs) have been recognized to have great potential in the field of target localization. To address the challenges of limited power consumption, channel throughput, and non-ideal transmission channel conditions in WSNs, this paper proposes a channel-aware direct localization method with low-bit quantization data. At its core, the received baseband signal is quantized into a small number of bits at each local sensor before being transmitted to the fusion center (FC). A channel-aware target localization method is designed at the FC to account for potential data distortion in the wireless channels due to the effects of fading and noise. The Cramér-Rao lower bound (CRLB) of the proposed method is derived to assess its localization performance. To achieve optimal localization performance, an optimization problem for quantizer design is formulated to guide each local sensor in quantizing the raw data. Since the formulated optimization problem is high-dimensional and non-convex, a low-dimensional optimization method based on position decoupling is proposed, which decouples the quantizer design problem into multiple low-dimensional sub-optimization problems. Specifically, for one-bit quantization, we prove that the optimal quantizer can be obtained analytically. For multi-bit quantization, we employ an efficient particle swarm optimization (PSO) based method for designing the quantizer. Our numerical results validate the theoretical analysis, demonstrating that the proposed method effectively addresses the low-bit quantization target localization problem especially in the presence of channel distortion.
engineering, electrical & electronic