Premium
Energy‐efficient partner selection in cooperative wireless networks: a matching‐theoretic approach
Author(s) -
Baidas Mohammed W.,
Afghah Masoud M.
Publication year - 2016
Publication title -
international journal of communication systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.344
H-Index - 49
eISSN - 1099-1131
pISSN - 1074-5351
DOI - 10.1002/dac.3114
Subject(s) - computer science , node (physics) , disjoint sets , energy consumption , matching (statistics) , selection (genetic algorithm) , energy (signal processing) , algorithm , stability (learning theory) , wireless network , efficient energy use , time complexity , wireless , mathematical optimization , computer network , telecommunications , mathematics , discrete mathematics , artificial intelligence , statistics , ecology , electrical engineering , structural engineering , machine learning , engineering , biology
Summary In this paper, the problem of stable energy‐efficient partner selection in cooperative wireless networks is studied. Each node aims to be paired with another node so as to minimize the total energy consumption required to meet a target end‐to‐end signal‐to‐noise ratio requirement and thus maintain quality of service. Specifically, each node ranks every other node in the network according to their energy saving achievable through cooperation. Two polynomial time complexity algorithms based on the stable roommates matching problem are proposed through which nodes are paired according to their preference lists. The first algorithm, denoted Irving's stable matching, may not always have a stable solution. Therefore, the second algorithm—which is a modified version of Irving's algorithm and denoted maximum stable matching—is proposed to find the maximum number of stable disjoint pairs. Simulation results are provided to validate the efficiency of the proposed algorithms in comparison with centralized energy‐efficient partner selection as well as other matching algorithms, yielding a trade‐off between stability and total energy consumption, but comparable symbol error rate performance and network sum rate. Copyright © 2016 John Wiley & Sons, Ltd.