Grant-Free Rateless SCMA for Cellular Internet of Things Networks

Abstract

For supporting the massive connectivities and better adapting the high dynamical channel states of the cellular Internet of things networks, in this paper we proposed a new grant-free rateless sparse code multiple access (RSCMA) protocol. In the proposed grant-free RSCMA, users switch from the inactive mode to the active mode as soon as they receive data packages, and start to transmit the data without the request-grant process between the base station (BS) and users. In each transmission, the active users firstly choose a certain number of the modulated SCMA symbols according to the pre-assigned access control function (ACF), and then transmit the linearly superimposed SCMA symbols to the BS over the available resource elements. At the BS, owing to the sparsity of RSCMA, message passing algorithm which allows multi-user detection with near-optimal performance can be utilized to decode the messages. The transmission will continue until the user successfully receives an ACK signal sent back from the BS. Theoretical analysis in terms of achievable constellation constrained throughput is provided to show the maximal sum rate of RSCMA and to optimize the ACF design under the trade-off between the complexity and throughput performance. Both theoretical and simulation results show that the RSCMA can achieve higher throughput than the existing rateless multiple access schemes, such as hybrid automatic repeat request SCMA, LT code coded SCMA, and rateless multiple access.