Noncoherent multiuser massive SIMO with mixed differential and index modulation

This paper proposes a new mixed differential and index modulation framework for noncoherent multiuser massive single-input multiple-output (SIMO) systems. While differential modulation and detection is a popular noncoherent scheme, its constellation collisions limit the resulting error performance. To address this issue, we introduce a user with binary index modulation (IM) among the differential users, achieving much reduced collisions. We then analyze a three-user SIMO system with binary modulations, attained a closed-form bit error rate (BER) expression with a fast noncoherent maximum-likelihood (ML) detection algorithm for each user. Furthermore, a closed-form optimal power loading vector is derived by minimizing the worst-case BER under individual power constraints. Finally, an efficient one-dimensional bisection search algorithm is employed to optimize constellations for arbitrary numbers of differential users and constellation sizes by minimizing the system BER. Simulation results validate the theoretical analysis and demonstrate the superiority of the proposed scheme compared to existing differential schemes.