A technique based on five brain rhythms (δ, θ, α, β, and γ) presented in a sequential format has been proposed for Electroencephalography (EEG)-based emotion recognition. Its production employs the prominent rhythm having maximum instantaneous power at each 0.2 s timestamp. For this purpose, smoothed pseudo Wigner-Ville distribution (RSPWVD) method is used. In total, 32 subjects from the emotional EEG database (DEAP) are applied for experimental validation, and for each subject, 640 rhythmic features derived from the time-related properties are extracted from 32 channels. After performance evaluation through support vector machine (SVM) classifier, the one that offers the highest accuracy can be found and then denoted as the optimal feature. By this means, the accuracies of EEG-based emotion recognition accomplish 78.36 ± 5.56% for arousal and 75.78 ± 3.73% for valence. Therefore, the results disclosed that a single optimal feature from a representative channel is competent to recognize the emotional EEG data.