The most prevalent sleeping respiratory disease, known as sleep apnea (SA), sometimes known as obstructive sleep apnea (OSA), is characterized by pauses in airflow to the lungs. If left unchecked, it can result in accidents, daytime hypersomnia, and persistent, catastrophic illnesses. Snoring and persistent sleepiness are hallmarks of the sleep disease known as sleep apnea (SA), which can cause catastrophic disorders like excessive blood pressure, heart failure, and cardiomyopathy. A polysomnography (PSG) overnight sleep investigation conducted in a laboratory is the most reliable method for identifying obstructive sleep apnea. Nevertheless, the patient may experience more agony because to the lengthy and laborious nature of the operation. The emergence of computer-aided diagnosis (CAD) has sparked growing interest among sleep disorder researchers in the automatic detection of OSA. This technology has significant implications for both diagnostic and therapeutic decision-making. Although it may show aberrant cardiac activity, the various biomedical play a crucial role in detecting SA. This paper presented current studies on SA identification using various biomedical signals, which have concentrated on feature engineering approaches that extract certain traits from the signals and feed them into machine learning or deep learning-based classification models.