Effect of Power Allocation Schemes on MIMO Two-Way Multi-Hop Network

Jonghyun LEE  Gia Khanh TRAN  Kei SAKAGUCHI  Kiyomichi ARAKI  

IEICE TRANSACTIONS on Communications   Vol.E93-B   No.12   pp.3362-3370
Publication Date: 2010/12/01
Online ISSN: 1745-1345
DOI: 10.1587/transcom.E93.B.3362
Print ISSN: 0916-8516
Type of Manuscript: Special Section PAPER (Special Section on Wireless Distributed Networks)
MIMO two-way multi-hop network,  power allocation,  eigenvector based power allocation (EPA),  optimal power allocation (OPA),  distributed power allocation (DPA),  

Full Text: PDF(891.4KB)>>
Buy this Article

Recently, wireless multi-hop network using MIMO two-way relaying technique has been attracted much attention owing to its high network efficiency. It is well known that the MIMO two-way multi-hop network (MTMN) can provide its maximum throughput in uniform topology of node location. However, in realistic environments with non-uniform topology, network capacity degrades severely due to unequal link quality. Furthermore, the end-to-end capacity also degrades at high SNR due to far (overreach) interference existing in multi-hop relay scenarios. In this paper, we focus on several power allocation schemes to improve the end-to-end capacity performance of MTMN with non-uniform topology and far interference. Three conventional power allocation schemes are reformulated and applied under the system model of MTMN. The first two are centralized methods, i.e., Eigenvector based Power Allocation (EPA) which employs linear algebra and Optimal Power Allocation (OPA) using convex optimization. The last one is Distributed Power Allocation (DPA) using game theory. It is found from numerical analyses that the power allocation schemes are effective for MTMN in terms of end-to-end capacity improvement, especially in non-uniform node arrangement and at high SNR.