DoF Analysis in a Two-Layered Heterogeneous Wireless Interference Network

Degrees of freedom (DoF) is studied in the downlink of a heterogenous wireless network modeled as a two-layered interference network. The first layer of the interference network is the backhaul layer between macro base stations (MB) and small cell base stations (SB), which is modeled as a Wyner type...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Bande, Meghana, Veeravalli, Venugopal, Tolli, Antti, Juntti, Markku
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Degrees of freedom (DoF) is studied in the downlink of a heterogenous wireless network modeled as a two-layered interference network. The first layer of the interference network is the backhaul layer between macro base stations (MB) and small cell base stations (SB), which is modeled as a Wyner type linear network. The second layer is the transmission layer between SBs and mobile terminals (MTs), which is modeled as a linear Wyner $L_T$ network, i.e., each MT is connected to $L_T+1$ SBs. The SBs are assumed to be half-duplex, thus restricting the per user degrees of freedom (puDoF) in the system to $1/2$. The puDoF can be further restricted by the number of antennas at the MB. For $L_T \in \{1,2\}$, the optimal puDoF can be achieved by using simple interference avoidance schemes. The increase in the connectivity of transmission layer beyond $L_T=2$ limits the achievable puDoF using only zero-forcing schemes to less than 1/2, even in the presence of large number of antennas at each MB but the optimal puDoF can be achieved by making each message available at multiple SBs. This is done by sending an appropriate linear combination to the SB to zero-force interference at the intended user. The maximum per user DoF of 1/2 can be achieved in the linear network with sufficient number of antennas using only interference avoidance schemes. These results are also extended to a more realistic hexagonal cellular model as well.
DOI:10.48550/arxiv.1609.02117