Satellite Link Status Prediction: LSTM Approach Using RSSI and Doppler Data

Sankani Sarathchandra; Asad Ullah; Antti Anttonen; Konstantin Mikhaylov; Hirley Alves

doi:10.1109/APWiMob67231.2025.11269154

Satellite Link Status Prediction: LSTM Approach Using RSSI and Doppler Data

Sankani Sarathchandra
, Asad Ullah
, Antti Anttonen
, Konstantin Mikhaylov
, Hirley Alves

University of Oulu

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

Abstract

—The performance of satellite communication systems strongly depends on line-of-sight (LoS) conditions between the satellite and the ground terminal. This paper presents a data-driven approach for predicting satellite visibility periods using a long short-term memory (LSTM) network. The model is trained on simulated beacon signals characterized by received signal strength indicator (RSSI) and Doppler shift values, generated for one-year period based on realistic orbital parameters and the 3GPP TR 38.811 S-band propagation model. The dataset includes both LoS and non-line-of-sight (NLoS) conditions, incorporating obstruction effects caused by a cylindrical obstacle. Once trained, the proposed LSTM model predicts satellite visibility for a stationary ground terminal solely from downlink radio signal features, without requiring two-line element (TLE) data, ephemeris information, or terminal location. This enables efficient, real-time classification of LoS and NLoS conditions for satellite communication links.

Original language	English
Title of host publication	2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob)
Publisher	IEEE Institute of Electrical and Electronic Engineers
Pages	167 - 172
Number of pages	6
ISBN (Electronic)	979-8-3315-8130-5
DOIs	https://doi.org/10.1109/APWiMob67231.2025.11269154
Publication status	Published - 5 Dec 2025
MoE publication type	A4 Article in a conference publication
Event	2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob) - Bali, Indonesia Duration: 6 Nov 2025 → 8 Nov 2025

Conference

Conference	2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob)
Abbreviated title	APWiMob
Country/Territory	Indonesia
City	Bali
Period	6/11/25 → 8/11/25

Funding

This research was supported by Business Finland through the 6G-SatMTC project and by the Research Council of Finland via 6G Flagship (grant 369116).

Keywords

Doppler shift
RSSI
satellite communication
line-of-sight
long short-term memory

Access to Document

10.1109/APWiMob67231.2025.11269154

Cite this

@inproceedings{390b472a94914cd182555b9bb6899acf,

title = "Satellite Link Status Prediction: LSTM Approach Using RSSI and Doppler Data",

abstract = "{\textemdash}The performance of satellite communication systems strongly depends on line-of-sight (LoS) conditions between the satellite and the ground terminal. This paper presents a data-driven approach for predicting satellite visibility periods using a long short-term memory (LSTM) network. The model is trained on simulated beacon signals characterized by received signal strength indicator (RSSI) and Doppler shift values, generated for one-year period based on realistic orbital parameters and the 3GPP TR 38.811 S-band propagation model. The dataset includes both LoS and non-line-of-sight (NLoS) conditions, incorporating obstruction effects caused by a cylindrical obstacle. Once trained, the proposed LSTM model predicts satellite visibility for a stationary ground terminal solely from downlink radio signal features, without requiring two-line element (TLE) data, ephemeris information, or terminal location. This enables efficient, real-time classification of LoS and NLoS conditions for satellite communication links.",

keywords = "Doppler shift, RSSI, satellite communication, line-of-sight, long short-term memory",

author = "Sankani Sarathchandra and Asad Ullah and Antti Anttonen and Konstantin Mikhaylov and Hirley Alves",

year = "2025",

month = dec,

day = "5",

doi = "10.1109/APWiMob67231.2025.11269154",

language = "English",

pages = "167 -- 172",

booktitle = "2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob)",

publisher = "IEEE Institute of Electrical and Electronic Engineers",

address = "United States",

note = "2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob), APWiMob ; Conference date: 06-11-2025 Through 08-11-2025",

}

Sarathchandra, S, Ullah, A , Anttonen, A, Mikhaylov, K & Alves, H 2025, Satellite Link Status Prediction: LSTM Approach Using RSSI and Doppler Data. in 2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob). IEEE Institute of Electrical and Electronic Engineers, pp. 167 - 172, 2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob), Bali, Indonesia, 6/11/25. https://doi.org/10.1109/APWiMob67231.2025.11269154

Satellite Link Status Prediction: LSTM Approach Using RSSI and Doppler Data. / Sarathchandra, Sankani; Ullah, Asad ; Anttonen, Antti et al.
2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob). IEEE Institute of Electrical and Electronic Engineers, 2025. p. 167 - 172.

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - Satellite Link Status Prediction: LSTM Approach Using RSSI and Doppler Data

AU - Sarathchandra, Sankani

AU - Ullah, Asad

AU - Anttonen, Antti

AU - Mikhaylov, Konstantin

AU - Alves, Hirley

PY - 2025/12/5

Y1 - 2025/12/5

N2 - —The performance of satellite communication systems strongly depends on line-of-sight (LoS) conditions between the satellite and the ground terminal. This paper presents a data-driven approach for predicting satellite visibility periods using a long short-term memory (LSTM) network. The model is trained on simulated beacon signals characterized by received signal strength indicator (RSSI) and Doppler shift values, generated for one-year period based on realistic orbital parameters and the 3GPP TR 38.811 S-band propagation model. The dataset includes both LoS and non-line-of-sight (NLoS) conditions, incorporating obstruction effects caused by a cylindrical obstacle. Once trained, the proposed LSTM model predicts satellite visibility for a stationary ground terminal solely from downlink radio signal features, without requiring two-line element (TLE) data, ephemeris information, or terminal location. This enables efficient, real-time classification of LoS and NLoS conditions for satellite communication links.

AB - —The performance of satellite communication systems strongly depends on line-of-sight (LoS) conditions between the satellite and the ground terminal. This paper presents a data-driven approach for predicting satellite visibility periods using a long short-term memory (LSTM) network. The model is trained on simulated beacon signals characterized by received signal strength indicator (RSSI) and Doppler shift values, generated for one-year period based on realistic orbital parameters and the 3GPP TR 38.811 S-band propagation model. The dataset includes both LoS and non-line-of-sight (NLoS) conditions, incorporating obstruction effects caused by a cylindrical obstacle. Once trained, the proposed LSTM model predicts satellite visibility for a stationary ground terminal solely from downlink radio signal features, without requiring two-line element (TLE) data, ephemeris information, or terminal location. This enables efficient, real-time classification of LoS and NLoS conditions for satellite communication links.

KW - Doppler shift

KW - RSSI

KW - satellite communication

KW - line-of-sight

KW - long short-term memory

UR - https://www.scopus.com/pages/publications/105032646362

U2 - 10.1109/APWiMob67231.2025.11269154

DO - 10.1109/APWiMob67231.2025.11269154

M3 - Conference article in proceedings

SP - 167

EP - 172

BT - 2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob)

PB - IEEE Institute of Electrical and Electronic Engineers

T2 - 2025 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob)

Y2 - 6 November 2025 through 8 November 2025

ER -

Satellite Link Status Prediction: LSTM Approach Using RSSI and Doppler Data

Abstract

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this