Paper Info
Title
Interference Cancellation Employing Replica Selection Algorithm And Neural Network Power Control For Mimo Small Cell Networks
Abstract
In a network with dense deployment of multiple-input multiple-output (MIMO) small cells, coverage overlap between the small cells produces intercell-interference, which degrades system capacity. This paper proposes an intercell-interference management (IIM) scheme that aims to maximize system capacity by using both power control for intercell-interference coordination (ICIC) on the transmitter side and interference cancellation (IC) on the receiver side. The power control determines transmit power levels at the base stations (BSs) by employing a neural network (NN) algorithm over the backhaul. To further improve the signal to interference plus noise ratio (SINR), every user terminal (UT) employs a multiuser detector (MUD) as IC. The MUD detects not only the desired signals, but also some interfering signals to be cancelled from received signals. The receiver structure consists of branch metric generators (BMGs) and MUD. BMGs suppress residual interference and noise in the received signals by whitening matched filters (WMFs), and then generate metrices by using the WMFs' outputs and symbol candidates that the MUD provides. On the basis of the metrices, the MUD detects both the selected interfering signals and the desired signals. In addition, the MUD determines which interfering signals are detected by an SINR based replica selection algorithm. Computer simulations demonstrate that the SINR based replica selection algorithm, which is combined with channel encoders and packet interleavers, can significantly improve the system bit error rate (BER) and that combining IC at the receiver with NN power control at the transmitter can considerably increase the system capacity. Furthermore, it is shown that choosing the detected interfering signals by the replica selection algorithm can obtain system capacity with comparable loss and less computational complexity compared to the conventional greedy algorithm.
Year
DOI
Venue
2016
10.1587/transcom.2015EBP3536
IEICE TRANSACTIONS ON COMMUNICATIONS
Keywords
Field
DocType
MIMO, small cells, intercell interference management, power control, neural network, multiuser detector, interference cancellation, system capacity
Replica,Computer science,Single antenna interference cancellation,Selection algorithm,Power control,Computer network,MIMO,System capacity,Small cell,Artificial neural network
Journal
Volume
Issue
ISSN
E99B
11
0916-8516
Citations 
PageRank 
References 
0
0.34
13
Authors
3
Name
Order
Citations
PageRank
Michael Andri Wijaya130.74
Kazuhiko Fukawa221239.25
Hiroshi Suzuki33332.69