Effective signal detection in a LoRa-networked ship for data-driven ship management

1. Introduction

With advancements in Internet of Things (IoT) technology, the market demand for various IoT technologies has increased. For achieving better network connectivity, many studies have been conducted to determine optimal power consumption strategies to increase the lifetime of each node [1]. In an IoT application, a large number of nodes equipped with communication devices and sensors requires access to the Internet and long-distance communication with limited energy [2].

Recently, data-based ship management with in-ship data is increasingly being used to improve efficiency and safety. Networking solutions that enable data-driven management in a ship should satisfy the requirements of a ship [3]. Passenger and non-passenger ships with a gross tonnage of 3000 or more that are constructed on or after July 1, 2002, must carry voyage data recorders (VDRs) to assist in accident investigations under regulations adopted in 2000 [3]. The VDR is a data recording system that collects data from various sensors onboard a vessel; it is designed for all vessels required to comply with the International Maritime Organization’s (IMO’s) International Convention for the Safety of Life at Sea (SOLAS) Requirements (IMO Res. A. 861 (20)) [4].

An IoT wireless sensor network on a ship must adopt techniques to reduce the energy consumption of distributed devices and to extend communication coverage to accurately collect various data. One study [2] noted that a long-range wide area network (LoRaWAN) [5]-[6] is one of the most promising technologies among various commercially available IoT networking services because it is a data link layer with a long range, low power consumption, and suitable bit rate for IoT applications.

In ships, it is difficult to guarantee a line-of-sight (LOS) communication environment because of the presence of shielding and obstacles in narrow spaces, as depicted in Figure 1. To ensure stable connectivity in such an environment, more relay nodes or gateways need to be deployed. However, this directly increases the installation and maintenance costs. Therefore, effective signal detection is essential to ensure a stable connection at a minimum cost, especially in life- and safety-related applications. In this light, this study develops an effective signal detection method for data-driven management in the communication environment of a ship.

Figure 1:

LoRaWAN communication environment in a ship

This paper is organized as follows: Section 2 proposes an applicable solution for effective signal detection in a LoRa-networked ship. Finally, concluding remarks are presented in Section 3.

2. Applicable solutions for effective signal detection in a LoRa-networked ship

The first part of a LoRa packet is the preamble. It consists of a variable number N_pr of upchirps, thus enabling effective preamble detection for a very large range of effective signal-to-noise ratios (SNRs) [7]. If the received signal strength indicator (RSSI) is below the noise floor, it is impossible to demodulate the signal [8]. However, LoRa can demodulate signals that are below the noise floor with dynamic spreading factors (SFs) informed by the RegModulationCfg parameter. The minimum SNR required for demodulation is shown in Table 1 [9].

Table 1:

Minimum SNR for demodulation with different SFs [9]

As shown in this table, LoRa signals with an SNR of up to -20 dB can be demodulated by adjusting the SF. Beyond the range of the minimum required SNR, it is necessary to increase the transmission (Tx) power, increase N_pr, or use a combination of these methods for signal detection. However, because adjusting the Tx power, SF, or preamble length causes the loss of transmission or processing power, the usage of these methods might be limited in the energy-restricted IoT environment.

Among various IoT application scenarios in a ship, life- or safety-related applications require effective preamble detection even in a harsh communication environment. In most communication systems, misdetection errors (P_m) and false alarms are assumed to be of equal importance. However, in such applications, the detection probability (P_D) must be maximized for a given false alarm probability. Therefore, an effective signal detection solution in a LoRa-networked ship subject to the false alarm constraint is proposed as follows.

The Neyman–Pearson (NP) criterion is used to maximize P(D₁|H₁) subject to the constraint P(D₁|H₀), where P(D₁|H₁) and P(D₁|H₀) are the probability that D₁ (signal is detected (‘1’)) is true when H₁ (hypothesis ‘1’: with signal) is true and H₀ (hypothesis ‘0’: without signal) is true, respectively [10]. In this criterion, the threshold has to be chosen to satisfy the false alarm constraint. As shown in Figure 2, the decision threshold (th_LoRa) to detect a signal is determined when the measured SNR is less than -20 dB and the applied SF is 12 according to the minimum SNR in Table 1.

Figure 2:

Proposed method for effective signal detection in a LoRa-networked ship

The false alarm probability (P_f), P(D₁|H₀) can be given by Equation (1), where p₀(y) is a probability density function with no signal present and R₁ is the region that results in H1 being chosen.

Assuming Gaussian noise, the probability density function p₀(y) is given by Equation (2).

Accordingly, if the tolerable false alarm probability (i.e., constraint) is set, the decision threshold th_LoRa can be chosen for effective signal detection in a LoRa-networked ship.

Figure 3 shows the false alarm probability with varying threshold values when the random variable y is defined as a signal with noise, where the noise is a Gaussian random variable having a mean of 0 and variance of 2. In this figure, the false alarm probability is seen to increase as the decision threshold decreases. However, by adjusting th_LoRa, the detection probability can be maximized by minimizing the misdetection probability according to the N-P criterion. The decision threshold can be adjusted through the P_f constraint shown in Figure 3.

Figure 3:

False alarm probability (Pf) with varying threshold (thLoRa) values

3. Conclusion

This study proposes an effective signal detection method in a LoRa-networked ship for data-driven ship management. For life- and safety-related applications, the detection performance is a more important criterion. Because the LoRaWAN protocol has been applied in various areas and meets the requirements for safety and efficiency, the proposed scheme improves its signal detection performance for ship applications and facilitates its easy adoption in an onboard IoT application.

Author Contributions

Conceptualization, Y. -I. Joo; Methodology, Y. -I. Joo; Writing-Original Draft Preparation, Y. -I. Joo; Validation, Y. -I. Joo; Writing-Review & Editing, Y. -I. Joo.

References

• J. de Carvalho Silva, J. J. P. C. Rodrigues, A. M. Alberti, P. Solic, and A. L. L. Aquino, “LoRaWAN — A low power WAN protocol for Internet of Things: A review and opportunities,” in Proceeding of 2017 2nd International Multidisciplinary Conference on Computer and Energy Science (SpliTech), pp. 1-6, 2017.
• Y. -I. Joo, “Networking strategy for data-driven management in a ship: Requirements and applications,” Journal of Advanced Marine Engineering and Technology, vol. 46, no. 4, pp. 193-195, 2022. [https://doi.org/10.5916/jamet.2022.46.4.193]
• Voyage Data Recorders, https://www.imo.org/en/OurWork/Safety/Pages/VDR.aspx, , Accessed May 11, 2022.
• Voyage Data Recorders, https://en.wikipedia.org/wiki/Voyage_data_recorder, , Accessed May 11, 2022.
• LoRa Alliance, LoRaWAN L2 1.0.4 Specification, TS001-1.0.4, October, 2020.
• SEMTECH, LoRa and LoRaWAN: A technical overview, Semtech Corporation, December, 2019.
• J. Tapparel, O. Afisiadis, P. Mayoraz, A. Balatsoukas-Stimming, and A. Burg, “An open-source LoRa physical layer prototype on GNU radio,” Proceeding of IEEE International Workshop on Signal Processing Advances in Wireless Communications (SPAWC), pp. 1-5, 2020. [https://doi.org/10.1109/SPAWC48557.2020.9154273]
• The things network, https://www.thethingsnetwork.org/docs/lorawan/rssi-and-snr/, , Accessed November 8, 2022.
• SEMTECH, Semtech datasheet SX1276-77-78-79, rev. 7, May 2020.
• R. N. McDonough and A. D. Whalen, Detection of signals in noise, 2nd edition, Academic Press, 1995.