Joint power allocation and relay selection for relay assisted D2D communication with channel uncertainties

Abstract

Device to device(D2D) communication exploits the proximity between mobile devices to allow direct communication. Doing so offers advantages which include less energy consumption, hop gain, reduced load on the base station among others. For instances where the channel between the D2D devices does not allow for efficient direct communication, a suitably selected idle mobile device can act as a relay to forward intended signals. When the relaying mobile is in motion, channel uncertainties may arise and so constitute channel estimation errors at the transmitter and receiver pairs affecting the relay selection process. Recently matching theory has been studied as an approach for allocating resources in wireless communication system. In this paper joint power allocation and relay selection is considered with focus on comparing two matching theory schemes namely the Hungarian assignment(HA) approach and the stable matching(SM)approach. Full duplex communication with multiple D2D user pairs are considered in this work. The relay selection problem is formulated as an optimization problem which is decomposed into power allocation and actual relay selection. We formulate the relay selection problem as a one to one weighted bipartite matching. Both HA and SM approaches outperform random selection by offering reduced total transmit power. Whereas the performance of HA and SM are mostly similar, with respect to channel uncertainty errors the HA algorithm is more robust against errors than SM.