Examinations are integral to any education system; however, maintaining their integrity in an online setup poses a significant challenge. Online exams are susceptible to numerous forms of cheating behavior and thus hinder the proper evaluation of the student’s performance. A plethora of research has been conducted in this field. It will expedite further research if we know what challenges are already solved or need improvements and which challenges need to be solved. In this paper, we systematically reviewed the research on online proctoring from 2020 to June 2024, considering eight research questions that address various criteria. These criteria include different formats of online exam proctoring, types of cheating behavior addressed, hardware requirements, methods of dataset collection, availability of datasets, techniques, and strategies for identifying cheating behaviors. Our study reveals several issues in the current research works: (I) Lack of proper definition of cheating behavior; (II) Students are often penalized for normal movements, which should not occur; (III) Proctoring systems should not be implemented at the cost of student discomfort and additional anxiety; (IV) A pattern of movements that signify cheating behavior should be identified, and penalties should be based on these patterns; (V) Proctoring systems should assist human proctors rather than being fully automated; (VI) There is a need for benchmark datasets. To gain a comprehensive understanding of proctoring types and strategies, this study is a valuable resource for academics and researchers in the field.