Multiplicative superposition signaling based detection schemes for heterogeneous-speed users

In a general cellular network, different users have different abilities to estimate the channel when it changes due to vehicle movements. On the two-user multiple-input-multiple-output (MIMO) broadcast channel, when one user with low mobility (called the static user) knows the exact channel state information (CSI) whereas the other user with high mobility (called the dynamic user) does not know CSI, a novel multiplicative superposition signaling scheme could obtain additive degree of freedom. Based on this novel signaling scheme, some detection schemes of the static user and the dynamic user are investigated in this paper. The signal desired by the dynamic user applies unitary space-time code and the signal desired by the static user is designed as the symbol matrix with the full spatial multiplexing, each element of which is a multiple phase-shift keying (PSK) or quadrature amplitude modulation (QAM) symbol. The final transmitted signal is the product of these two signals. At the dynamic user, a non-coherent maximum likelihood detection (NCMLD) is used. At the static user, we propose two hard detection schemes based on NCMLD-ML and NCMLD-Minimum-Mean-Square-Error (MMSE), respectively. To avoid the error-propagation problem in the hard detection schemes, we also propose a low-complexity soft detection scheme based on the conditional expectation and MMSE (CE-MMSE). Simulation results show that the NCMLD-ML scheme outperforms the NCMLD-MMSE scheme and CE-MMSE scheme. Furthermore, CE-MMSE scheme has a better error performance than NCMLD-MMSE in the low signal-to-noise ratio (SNR) region at the expense of higher complexity which is inconsiderable when it is compared with NCMLD-ML scheme.