TY - GEN
T1 - Low-Complexity Detection for Generalized Pre-Coding Aided Spatial Modulation
AU - Perovic, Nemanja Stefan
AU - Haselmayr, Werner
AU - Springer, Andreas
PY - 2015
Y1 - 2015
N2 - In this paper we consider Generalized Pre-coding aided Spatial Modulation (GPSM), which was recently proposed as a promising alternative to conventional Multiple Input Multiple Output (MIMO) transmission schemes. In GPSM only a part of the receive antennas is activated with the aid of pre-coding at the transmitter. Hence, information bits are mapped to a spatial symbol, corresponding to a particular activation pattern, and to
modulation symbols. Optimal performance is achieved with a Maximum Likelihood (ML) detector, but its exhaustive search leads to an intractable complexity. In this paper we present a novel detector, referred to as Soft MMSE with Exhaustive Search (SOMES) detector, that computes soft information for each symbol by employing a soft-output Minimum Mean Square Error (MMSE) detector. The soft information is used to determine the activation pattern using a small exhaustive search and to
obtain the symbols in the particular activation pattern. Link level simulations show that the proposed algorithm possesses the near-optimal Bit Error Rate (BER) performance while achieving a remarkable reduction in complexity.
AB - In this paper we consider Generalized Pre-coding aided Spatial Modulation (GPSM), which was recently proposed as a promising alternative to conventional Multiple Input Multiple Output (MIMO) transmission schemes. In GPSM only a part of the receive antennas is activated with the aid of pre-coding at the transmitter. Hence, information bits are mapped to a spatial symbol, corresponding to a particular activation pattern, and to
modulation symbols. Optimal performance is achieved with a Maximum Likelihood (ML) detector, but its exhaustive search leads to an intractable complexity. In this paper we present a novel detector, referred to as Soft MMSE with Exhaustive Search (SOMES) detector, that computes soft information for each symbol by employing a soft-output Minimum Mean Square Error (MMSE) detector. The soft information is used to determine the activation pattern using a small exhaustive search and to
obtain the symbols in the particular activation pattern. Link level simulations show that the proposed algorithm possesses the near-optimal Bit Error Rate (BER) performance while achieving a remarkable reduction in complexity.
UR - https://ieeexplore.ieee.org/document/7391015
U2 - 10.1109/VTCFall.2015.7391015
DO - 10.1109/VTCFall.2015.7391015
M3 - Conference proceedings
T3 - IEEE Vehicular Technology Conference
BT - Vehicular Technology Conference, 82th IEEE Conference on
ER -