Generalized Differential Index Modulation for Pilot-Free Communications

Short packet transmission is a key ingredient for the Internet-of-Things (IoT) communications that require extremely low latency. The pilot-based short packet transmission makes lower spectral efficiency and larger transmission delays. To resolve this problem, in this article, we present a generalized differential subcarrier index modulation (GDIM) for pilot-free short packet transmission. The central idea of GDIM is to map information bits to possible indices that can be constructed by joint row and column permutations of a unitary matrix. We show that the proposed GDIM achieves higher spectral efficiency than that attained by the existing differential modulation techniques, including differential subcarrier index modulation and differential Alamouti code. We present a noncoherent maximum-likelihood (ML) detection method and a low-complexity variant of it for GDIM. Furthermore, by analyzing the average bit-error-rate when applying GDIM, we propose an algorithm to optimize GDIM constellation sets. Simulation results demonstrate that the proposed GDIM outperforms the existing differential modulation techniques in terms of bit-error-rates for numerous target spectral efficiencies.