This paper offers an in-depth analysis of malware detection methods on smartphones, highlighting current challenges, existing solutions and prospects. The article reviews recent scientific work and performance evaluations of different detection approaches by exploring the fundamentals of smartphone security and examining the main categories of malware. The use of static malware detection is the most widespread, but there is currently a growing trend towards Deep Learning-based approaches, which on average have a malware detection rate approaching 97%. The paper also discusses persistent challenges such as the rapidly evolving malware operating mechanism and inherent privacy issues, while identifying future opportunities for improving malware detection on smartphones. This paper thus provides a comprehensive and up-to-date overview of the field of intelligent malware detection, offering valuable pointers for both researchers and practitioners concerned with protecting smartphone users from digital threats.