DocumentCode :
3732797
Title :
Learning the vehicular channel through the self-organization of frequencies
Author :
Razvan-Andrei Stoica;Stefano Severi;Giuseppe Thadeu Freitas de Abreu
Author_Institution :
Focus Area Mobility, Jacobs University Bremen, Campus Ring 1, 28759, Germany
fYear :
2015
Firstpage :
68
Lastpage :
75
Abstract :
Connected Vehicles is a currently hot topic, which can be said to have emerged under the inspiration of the Internet of Things. Motivated by a myriad of traffic safety, decongestion, and respectively, driver comfort applications, the subject is gaining more and more popularity among both researchers and industry. Our interests in the topic are focused on both the communication links and the precise localization of the vehicles. These topics are catalyzed by applications such as high precision collaborative vehicular positioning and autonomous driving. The robustness of the radio links in this context is hence at the backbone of vehicular ad-hoc networks, and respectively, of initial cooperative positioning estimates. The subsequent communication protocol is based on the IEEE 802.11p standard of the WAVE (Wireless Access in Vehicular Environments) technology. As part of the post-transmission channel modeling, we introduce in this paper a new channel estimator structure based on a low complexity adaption of Kohonen´s Self-Organizing Map complemented by a filtered decision feedback layer. Furthermore, we study the key factors that would lead to a further performance optimization of our estimator while also comparing it to the state of the art solutions existing already in the literature.
Keywords :
"Channel estimation","OFDM","Training","Standards","Estimation","Frequency estimation","Mathematical model"
Publisher :
ieee
Conference_Titel :
Vehicular Networking Conference (VNC), 2015 IEEE
Electronic_ISBN :
2157-9865
Type :
conf
DOI :
10.1109/VNC.2015.7385549
Filename :
7385549
Link To Document :
بازگشت