<p>In recent times, the use of artificial agents in various aspects of human social life has led to increasing attention on their emotional expressions from both the human-computer interaction and robotics communities. The ability of these agents to express emotions can facilitate engaging and believable communication with humans. As speech is a primary medium for conveying emotions among humans, it is essential to synthesize emotion-aware speech in artificial agents. This paper aims to provide a overview of methods for generating emotion-aware speech, serving as a guide for researchers in the field. Given the significant role of prosody in producing human-like and engaging speech, we investigate state-of-the-art methods that convey emotions through prosody. We present a general categorization of emotion-aware speech generation methods and review state-of-the-art methods in this field. Then, a comparative analysis of different emotion-aware speech generation methods is conducted. Finally, we categorized and discussed challenges that need to be addressed for further research. The main contribution of the paper is to provide insight and extensive resources for researchers focused on emotional speech generation.</p>