Cooperative Learning for Spectrum Management in Railway Cognitive Radio Network

Wu, Cheng; Wang, Cheng; Sheng, Jie; Wang, Yiming

doi:10.1109/tvt.2019.2912673

Cited by 13 publications

(2 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Techniques such as paral-lel multiple-agent policy updates [171], swarm-based techniques [172], and learning state-space division among agents [173] have been proposed. In the HST scenario, a recent work in [174] established a Bayesian-adaptive partially observable Markov decision process and applied multi-agent Bayesian reinforcement learning for joint spectrum management of BS groups to tackle the problem of spectrum volatility due to high mobility and frequent handover. The accuracy of AI and ML algorithms depends on whether the data used to train the neural network are drawn from the same or similar distribution as the true application.…”

Section: ) Ai/ml Assistance In Hst Signal Processingmentioning

confidence: 99%

Physical Layer Enhancement for Next-Generation Railway Communication Systems

Sibel

Berbineau

et al. 2022

IEEE Access

View full text Add to dashboard Cite

This paper presents an overview of the challenges and state-of-the-art physical layer enhancement designs for next-generation railway communication, also known as high-speed train (HST) communication. The physical layer design for the HST should adapt from its counterpart in the generalpurpose network because of the harsh propagation environment and extreme conditions, stringent latency and reliability requirements of dedicated railway applications, and frequency band scarcity caused by regulation. In this survey, we examine how conventional multiple-input-multiple-output (MIMO) family techniques such as beamforming, multi-cell MIMO, and relays can enhance the physical layer performance for HST. Physical layer enhancement assisted by novel reconfigurable intelligent surface (RIS) technology was also analyzed from different perspectives. Dedicated control channels, reference signals, waveforms, and numerology designs for train-to-infrastructure (T2I) and train-to-train (T2T) communication in sidelinks are also reviewed. Finally, a brief introduction to artificial intelligence (AI)/machine learning (ML)-aided HST physical layer design is provided. Several promising research avenues have also been suggested.

show abstract

Section: ) Ai/ml Assistance In Hst Signal Processingmentioning

confidence: 99%

Physical Layer Enhancement for Next-Generation Railway Communication Systems

Sibel

Berbineau

et al. 2022

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Previous studies [7][8][9][10] proposed spectrum management schemes in railway networks. These schemes provide significant improvement to network performance due to their knowledge of the the trajectory and speed of the train.…”

Section: Existing Spectrum Handover Schemesmentioning

confidence: 99%

Cooperation-Assisted Spectrum Handover Mechanism in Vehicular Ad Hoc Networks

Chuang

2021

Electronics

View full text Add to dashboard Cite

This paper proposes a cooperation-assisted spectrum handover (CASH) scheme in vehicular ad hoc networks (VANETs). In CASH, each vehicle uses cognitive radio technology to collect the surrounding spectrum information, computes the stability of the spectrum, exchanges neighbor information by vehicle-to-vehicle collaboration technology, and then executes a partial prescan mechanism to reduce the handover delay. The proposed method can improve the spectrum utilization, shorten the delay time of the spectrum handover, and reduce the total number of handovers. Finally, the simulation results show that the proposed method outperforms other existing schemes. In spectrum utilization, CASH is about 20% better than that of the traditional full scan method. In average scan time, the time of CASH is about 5 times less than the time of the traditional full scan method. In total number of handovers, CASH improves the performance of the full scan scheme by 33%.

show abstract