A Tensor based Precoder and Receiver for MIMO GFDM systems

Generalized Frequency Division Multiplexing (GFDM) is a multi-domain communication scheme where data symbols are transmitted over a time-frequency block. Tensors, which are multi-way arrays, can be efficiently used to model such systems. This paper presents a system model for a multiple input multip...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE International Conference on Communications (2003) S. 1 - 6
Hauptverfasser: Pandey, Divyanshu, Leib, Harry
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.06.2021
Schlagworte:
Capacity
Covariance matrices
Eigenvalues and eigenfunctions
Frequency division multiplexing
MIMO communication
MIMO GFDM
Multi-linear MMSE
Precoding
Tensor Channel
Tensors
Transmitting antennas
ISSN:1938-1883
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Generalized Frequency Division Multiplexing (GFDM) is a multi-domain communication scheme where data symbols are transmitted over a time-frequency block. Tensors, which are multi-way arrays, can be efficiently used to model such systems. This paper presents a system model for a multiple input multiple output (MIMO) GFDM system using the Einstein product of tensors. The input and output are modelled as order 3 tensors where the three modes correspond to space, time and frequency domains. The equivalent channel between the input and output obtained from a cascade of transmit filter, physical channel and receive filter, is defined as an order 6 tensor which takes into account interference across all the domains. An information theoretic analysis of such a tensor channel is presented which is then used to develop a tensor based precoding scheme for MIMO GFDM systems. The effect of various GFDM pulse shape parameters on the capacity of the equivalent channel is explored. A multi-linear minimum mean square error (MMSE) receiver using the tensor framework is also presented.
ISSN:1938-1883
DOI:10.1109/ICC42927.2021.9500957