Research Article
An Energy Sequencing Based Partial Maximum Likelihood Detection Method for Space-Frequency Joint Index Modulation System
@INPROCEEDINGS{10.1007/978-3-030-14657-3_32, author={Xiaoke Niu and Xingle Feng and Kun Hua and Guobin Duan and Shizhe Gao}, title={An Energy Sequencing Based Partial Maximum Likelihood Detection Method for Space-Frequency Joint Index Modulation System}, proceedings={IoT as a Service. 4th EAI International Conference, IoTaaS 2018, Xi’an, China, November 17--18, 2018, Proceedings}, proceedings_a={IOTAAS}, year={2019}, month={3}, keywords={Maximum Likelihood (ML) Space-Frequency joint index modulation Bit Error Rate (BER) Minimum Mean Square Error (MMSE)}, doi={10.1007/978-3-030-14657-3_32} }- Xiaoke Niu
Xingle Feng
Kun Hua
Guobin Duan
Shizhe Gao
Year: 2019
An Energy Sequencing Based Partial Maximum Likelihood Detection Method for Space-Frequency Joint Index Modulation System
IOTAAS
Springer
DOI: 10.1007/978-3-030-14657-3_32
Abstract
In this paper, an energy sequencing based partial Maximum Likelihood (ML) detection algorithm is proposed for the complex characteristics of receiver detection in space-frequency joint index modulation system. This algorithm can solve the problems of high complexity from ML detection and poor Bit Error Rate (BER) performances by Minimum Mean Square Error (MMSE) detection. The major idea of the proposed algorithm is to demodulate the activated sub-carrier sequence number, antenna sequence number and constellation symbol step by step, where the sub-carrier sequence number is equalized with MMSE and the energy value of each sub-carrier is calculated and sorted. And the P value is set as the number of candidate sub-carriers. Finally, the sequence numbers of alternative sub-carriers, antenna serial numbers and constellation symbols are detected by ML. Simulation results show that the proposed algorithm can reduce both search range of traditional ML methods and the complexity according to the selection of P value. For example, when , the BER can be reduced to 10 at the SNR of 20 dB in the proposed algorithm.
