Machine Learning-Based Satellite Routing for SAGIN IoT Networks

Abstract

Due to limited coverage, radio access provided by ground communication systems is not available everywhere on the Earth. It is necessary to develop a new three-dimensional network architecture in a bid to meet various connection requirements. Space–air–ground integrated networks (SAGINs) offer large coverage, but the communication quality of satellites is often compromised by weather conditions. To solve this problem, we propose an extended extreme learning machine (ELM) algorithm in this paper, which can predict the communication attenuation caused by rainy weather to satellite communication links, so as to avoid large path loss caused by bad weather conditions. Firstly, we use Internet of Things (IoT)-enabled sensors to collect weather-related data. Then, the system feeds the data to the extended ELM model to obtain a category prediction for blockage caused by weather. Finally, this information helps the selection of the data transmission link and thus improves the satellite routing performance.

A systematic literature review of machine learning applications in IoT

“…Data aggregation, load balancing, fault tolerance, network longevity, energy efficiency, scalability, and dependability are just a few of the diverse goals that network clustering is utilized toward. ML‐based routing : Compablack to traditional routing, ML‐based solutions have the advantages of learning useful information from input historical data about the network in order to pblackict new conditions which improve the routing decision and hence improve the network performances. For example, a ML‐based model can pblackict the congestion of the communication paths between the satellites and the ground stations, by the bias of historical weather data, which can improve the quality of the communication in space‐air‐ground integrated networks and increase the packet delivery probability 82 . According to earlier studies, 14,79,81 supervised learning approaches train real data traces.…”

“…Processing enormous volumes of large datasets from IoT sensing devices is a challenging issue, though. ML‐based routing solutions : Despite the benefits gained from applying ML‐based models in optimizing the routing process, there are some challenges related to training data, computation complexity, and latency that should be well addressed. Considering training data, it is shown in the reviewed solutions that collecting historical data generated from simulating traditional routing schemes, such as in Sharma et al, 39 or environment observations, such as in Yuan et al, 82 is requiblack to pblackict accurate results. However, several studies argue that it is not an easy task to collect data for real‐world network environments in practice 5,92 .…”

“…Yuan et al 82 proposed an extended extreme learning machine (ELM) algorithm to improve the communication quality in space‐air‐ground integrated networks (SAGIN). An architecture that relies on satellite‐based IoT (SIoT) networks is put in use.…”

“…Figure 8 shows an example of the network architecture and relay nodes selection after running the proposed solution. Yuan et al 82 proposed an extended extreme learning machine (ELM) algorithm to improve the communication quality in space-air-ground integrated networks (SAGIN). An architecture that relies on satellite-based IoT (SIoT) networks is put in use.…”

Space-Air-Ground Integrated Network for Disaster Management: Systematic Literature Review

“…However, they also have certain limitations. To overcome all the limitations of .5 [72] 2021 Proposed solution 1 1 0 0 2 [73] 2019 Proposed solution 1 1 1 1 4 [74] 2022 Framework 1 1 0.5 2 4.5 [75] 2022 Proposed solution 0 0.5 0 1.5 2 [76] 2019 Proposed solution 1 0.5 1 0 2.5 [77] 2020 Survey 1 0.5 1 2 4.5 [78] 2021 Model 1 0.5 1 2 4.5 [79] 2019 Framework 0 0.5 1 2 3.5 [80] 2021 Model 0 0.5 0.5 1 2 [81] 2020 Proposed solution 1 2021 Proposed solution 1 1 0.5 2 4.5 [90] 2018 Framework 0 0.5 1 2 3.5 [91] 2022 Survey 1 1 0.5 2 4.5 [92] 2021 Survey 1 0.5 1 2 4.5 [93] 2021 Proposed solution 1 0.5 0 2 3.5 [30] 2021 Proposed solution 1 1 0.5 1.5 4 [94] 2017 Survey 1 0.5 0.5 0 2 [95] 2017 Framework 1 0.5 0 0 1.5 [96] 2020 Proposed solution 1 0. In recent years, SAGIN has been used by multiple organizations including SpaceX [134] and Global Information Grid [135].…”