In the past ten years, the Internet of Things (IoT) has become a massive force in our daily lives, providing countless smart services that have improved human existence. Nevertheless, the increased availability and rapidly escalating demand for intelligent devices and networks mean that IoT is currently encountering unprecedented security challenges. While there are established security protocols in place for IoT, these conventional methods fall short in effectively addressing the escalating frequency and severity of attacks. The term 'security' is used to encompass various concepts, including integrity, confidentiality, and privacy. Therefore, a strong and up-to-date security system is necessary for the next generation of industrial IoT. One area of technological advancement that can address the ongoing and future challenges of IoT security is Soft Computing. It has opened up many possible research avenues for detecting attacks and identifying abnormal behaviors in smart devices and networks. This paper explores the architecture of IoT and provides a comprehensive literature review of Soft Computing based approaches to IoT security, including the different types of possible attacks. It also presents potential Soft Computing-based solutions for IoT security and discusses open research issues along with future research scope.