Hard/Soft NLoS Detection via Combinatorial Data Augmentation for 6G Positioning

TL;DR

A novel NLoS detection algorithm termed combinatorial data augmentation-guided NLoS detection (CDA-ND), which builds upon prior work and achieves high reliability in indoor factory environments under frequency range 1, attaining NLoS detection accuracies of 96.6% and 91.1%.

Abstract

A key enabler for meeting the stringent requirements of 6G positioning is the ability to exploit site-dependent information governing line-of-sight (LoS) and non-line-of-sight (NLoS) propagation. However, acquiring such environmental information in real time is challenging in practice. To address this issue, we propose a novel NLoS detection algorithm termed combinatorial data augmentation-guided NLoS detection (CDA-ND), which builds upon our prior work. CDA-ND generates numerous preliminary estimated locations (PELs) by applying multilateration over many gNodeB (gNB) combinations using a single snapshot of range measurements. When a target gNB is in NLoS, the resulting PELs split into two clusters: one derived using the target gNB's range measurement and the other derived without it. Their displacement is summarized by a single vector, called the NLoS evidence vector (NEV), which is used to compute an NLoS likelihood score. Based on this score, two modes of NLoS detection are developed. First, each gNB is classified as LoS or NLoS, termed hard decision (HD), using a simple threshold test. Second, each gNB's NLoS confidence is probabilistically quantified, termed soft decision (SD), which extends HD with weak site-survey priors, namely empirical NLoS-score samples and the average NLoS probability. We then design positioning algorithms tailored to these two modes by excluding gNBs deemed NLoS and re-weighting the remaining gNBs for SD. The proposed CDA-ND achieves high reliability in indoor factory environments under frequency range 1, attaining NLoS detection accuracies of 96.6% and 91.1% when the proportion of NLoS gNBs is approximately 18% and 56%, respectively. As a result, integrating CDA-ND into positioning significantly reduces mean absolute error by 20.04% and 65.99% in LoS- and NLoS-dominant environments, respectively.

Figures (10)

• Figure 1: Graphical illustration of a single snapshot of PELs skewed by a target NLoS gNB. The black circles represent PELs constructed from the target gNB's measurements, while the gray circles represent PELs constructed without it. The yellow star indicates the true UE position.
• Figure 2: Receiver operating characteristic curves of the SD output as defined in Sec. \ref{['sec:soft_decision']}, under different numbers of gNBs $N\in\{6,8,\dots,18\}$, in the InF-DH scenario under frequency range 1 (settings in Sec. \ref{['sec:sim_setup']}). Here, "positive" corresponds to NLoS. Darker curves indicate larger $N$, with higher AUCs demonstrating better separability between LoS and NLoS links, i.e., higher detection reliability across thresholds.
• Figure 3: Illustration of CDA and its implications for NLoS detection. (a) CDA generates multiple PELs by applying multilateration to different subsets of gNB range measurements. (b) When a target gNB is NLoS, the PELs formed with that gNB and those formed without it exhibit a systematic spatial shift, yielding a discriminative geometric NLoS signature in the PEL distribution. (c) The PEL distribution empirically converges as the number of participating gNBs $N$ increases, quantified by the average squared MMD between the distribution induced by $N$ and $(N-1)$ gNBs.
• Figure 4: Examples of the NEV from \ref{['eq:NEV']} in a single snapshot across three different target gNB cases: (a) an NLoS case with both sufficient magnitude and directional alignment, (b) a LoS case with sufficient magnitude but directional misalignment, and (c) a LoS case with directional alignment but insufficient magnitude.
• Figure 5: The empirical NLoS probability with respect to score $\rho$ in InF-DH under FR1 (details in Sec. \ref{['sec:sim_setup']}), shown to fit a sigmoid curve. Besides, the probability density of the adaptive threshold $\eta$ is represented by blue histogram.

Theorems & Definitions (5)

• Definition 1: NLoS Evidence Vector (NEV)
• Remark 1: Incomplete Reference Vector
• Remark 2: Sigmoid-like NLoS Probability
• Remark 3: HD vs. RE&RS Filtering
• Remark 4: Role of SD for Positioning