Abstract
A feedback scheme based on an iterative group splitting along with an opportunistic scheduling in a time division multiplexed wireless system with adaptive modulation is proposed in this paper. Considering a dynamic behavior of users in joining and leaving the multiuser networks, the proposed scheme does not assume any prior knowledge on users' channel statistics and therefore allows more robust and practical design. During a guard period, the proposed scheme exploits user feedback collisions in order to firstly find the signal-to-noise ratio (SNR) region in the adaptive modulation to which the best user belongs and secondly narrow down the range of user candidates by splitting user groups iteratively. As soon as one of the qualified users, whose channel quality falls into the SNR region for the best user, is found during the procedure, this user is selected by the scheduler and no more searching is pursued. Using an iterative group splitting, it is shown that the proposed scheme achieves a significant reduction in the number of feedbacks and thus scales with a large number of users.