Exploiting Local Channel State Information for Random Access in Cooperative Networks
Shu-Hsien Wang,An-Dee Lin,Y. -W. Peter Hong
DOI: https://doi.org/10.1109/tvt.2013.2252638
IF: 6.8
2013-01-01
IEEE Transactions on Vehicular Technology
Abstract:In multiuser wireless systems, the variation in channel quality among different users can be explored to enhance system throughput and reliability. In particular, multiuser diversity can be exploited by scheduling the user with the best channel to transmit in each time slot, whereas cooperative diversity can be exploited by having users relay messages for each other. In the past, these advantages have been mostly separately addressed in the literature. In this paper, both multiuser diversity and cooperative diversity are jointly considered in the design of distributed channel-aware transmission control (CATC) and partner selection policies for random access cooperative networks. In the cooperative system under consideration, each user is allowed to select adaptively one of the other users in the system (i.e., a partner) to help forward a packet that has failed to reach the access point (AP) in the current time slot. The proposed policies allow each user to determine its transmission probability and partner selection based only on its local uplink and interuser channel state information (CSI), i.e., the channel between itself and the other users in the system. By maximizing the weighted sum throughput, the optimal policies are first derived for a two-user system, where each user only needs to determine whether or not to transmit and whether or not to utilize relaying by the other user. The policies are later generalized to systems with an arbitrary number of users. When there are more than two users, a selective relaying scheme is employed where each user requests assistance from only one other user in each time slot. The optimal partner selection policy is obtained by maximizing the local effective channel equality at each user, whereas the optimal transmission control function takes on the form of a threshold function, which allows a user to transmit only when its local effective channel is sufficiently reliable. The proposed distributed channel-aware transmission policies are also extended to the cases with only imperfect CSI at the users.