Spectral Efficiency Improvement in Spatial Modulation Systems

Abstract

A novel energy efficient Multiple-Input-Multiple-Output (MIMO) technique is called Spatial Mod- ulation (SM). It uses only one radio frequency (RF) chain that reduces the hardware complexity and cost of the system. The concept of SM is to transmit the data information using modulation constellation and spatial constellation. It increases the spectral efficiency by keeping only one an- tenna active per every symbol period. Spatial Multiplexing (SMX) uses equal number of RF chains with transmit antennas that increases the hardware complexity and cost of the system. SM is free from Inter Channel Interference (ICI) and Inter Symbol Interference (ISI). Generalised Spatial Modulation (GSM) and Multiple Active Spatial Modulation (MASM) techniques were developed to increase the spectral efficiency of SM by increasing the number of RF chains. In GSM, the same modulation symbol transmits on different antennas hence, it avoids ISI. In MASM, multiple symbols are transmitted on different antenna con guration. A modi ed version of SM technique called Quadrature Spatial Modulation (QSM) uses both in-phase and quadrature dimensions to transmit the data symbol in one time instant. Hence, QSM enhance the spectral efficiency over SM. Recently, Generalised QSM (GQSM) scheme was developed to increase spectral efficiency by grouping the transmit antennas according to QSM principle. We propose a modi ed Generalised QSM (mGQSM) scheme without antenna grouping and we use multiple RF chains to enhance the spectral efficiency. The proposed scheme provides ex- tra one bit per channel use (bpcu) spectral efficiency over GQSM scheme with the constraint of flog2 Nt Nrf

g 0:5, where f g denotes the fractional part of the decimal value, Nt denotes number of transmit antennas, and Nrf denotes number of RF chains. In mGQSM system, multiple data symbols are divided into real and imaginary parts and these parts are transmitted as in-phase and quadrature components by selecting any possible antenna activation patterns available in mGQSM, resulting in the choice of antenna activation patterns being doubled in mGQSM compared to GQSM which yields the extra one bpcu spectral efficiency over GQSM. Using the ML detection algorithm, we study the performance via numerical simulations using half code rate convolutional encoder at the transmitter and Viterbi decoding algorithm at the receiver to estimate the transmitted bits. We compute the computational complexity of ML- decoding in terms of real valued multiplications and introduce a variant of mGQSM called Reduced Codebook mGQSM (RC-mGQSM) to reduce the complexity by decreasing the spectral efficiency. In summary, a novel scheme mGQSM is proposed which improves the spectral efficiency of known scheme, GQSM by one bpcu.