Generalized Index Modulation System and Application for Satellite Communications

The generalized index modulation (G-IM) technique provides an improvement in bit error rate (BER), spectral efficiency (SE), and/or energy efficiency (EE) over traditional digital modulation schemes. However, G-IM techniques suffer from the propagation of errors due to the incorrect detection of indices for higher modulation order (M > 4). To alleviate the risk of error propagation, the enhancement of the correct indices detection probability is a crucial requirement for G-IM systems. This paper presents an error propagation-free generalized index modulation (G-IM) system with adaptive coding modulation in which the indices bits are coded independently on the information bits. The main focus of this study is to derive a quantitative relation between the target system BER and the required indices error rate and, hence, to derive the overall system error probability. In particular, an analytical total BER formula for the error propagation-free G-IM system is provided and compared to the total BER of the classical counterpart modulation techniques. Hence, a direct relationship between the BER of index bits and the ordinary modulation BER is represented, providing an approximate threshold after which the target BER of the G-IM system outperforms the traditional modulation techniques. To validate our established results, the G-IM system with low density parity check code with binary phase shift keying modulation technique (LDPC-BPSK), as a simple satellite coding-modulation scheme which is used in satellite digital video broadcasting TV (DVB-T2) and deep-space communications, is described as a possible direct application for the G-IM system. once the index BER exceeds the threshold, the LDPC-BPSK modulated G-IM system significantly outperforms the classical BPSK modulation while keeping the same EE or SE yet on the expense of added encoder complexity. Moreover, once the index error tends to the zero point, the total G-IM BER reaches its lower limit, a 3 dB gain, thanks to power reallocation in the information bits which represents, on average, 50% of the entire frame.