A Novel Performance Evaluation Framework for OTFS System

Orthogonal Time Frequency Space (OTFS) systems have been widely investigated to achieve better performances in high-speed mobile scenarios. Bit Error Rate (BER) is always considered as the core indicator to measure communication performance in different conditions. However, the main issues of BER assessments are limited evaluation indicators and long computation time. Therefore, in this paper, we propose a novel communication performance evaluation framework for OTFS systems, which complements and extends traditional communication performance evaluation based on BER, and thus faster evaluation as well as parameter selections can be obtained. Specifically, the proposed evaluation framework is constituted by three main components, which includes indicator selection, weight calculation, and performance ranking. Firstly, waveform parameters, channel information and receiving algorithm of OTFS systems are considered as evaluation characteristics, and thus four key indicators are integrated and extracted for subsequent assessments. Secondly, subjective and objective weighting algorithms are respectively employed to determine the weights of different indicators, and then game theory is considered to obtain a joint weight. The Analytic Hierarchy Process (AHP) is enhanced through cloud model and optimal transfer matrix, to address the issue of multi-expert collaboration and the repeated consistency checks. In addition, the correlation coefficient matrix is applied in the coefficient of variation (CV) to solve the problem of indicator independence. Finally, the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) is selected as the ranking algorithm to judge performances of different schemes, and thus evaluation results can be obtained. In simulation results, eleven different schemes with different configurations are chosen for performance assessments, which demonstrates that the proposed evaluation framework effectively compares communication performances of different schemes. Additionally, the evaluation results can also point out the optimal solution, and thus the optimal parameter selections can be guided.