Unmanned Aerial Vehicles (UAVs) have rapidly been adopted in industries from logistics and agriculture to surveillance and defence. Despite this, integrating UAVs into critical operations has revealed the serious cybersecurity challenges that need to be overcome before safe and robust deployment of these autonomous systems is practical. This paper explores the emerging threats and advanced solutions in UAV cybersecurity, focusing on three key areas: Blockchain technology, machine learning, safe communication protocols. Blockchain is a decentralized and immutable architecture preventing tampering, eavesdropping, and unauthorized access during UAV data exchanges. In addition, machine learning and deep learning techniques, like intrusion detection and anomaly detection systems, are presented as a potential solution to proactively identify and mitigate harmful activities like GPS spoofing, signal jamming and unauthorized UAV operations. Furthermore, the vulnerabilities of UAV radio links are analyzed and advanced communication protocols are evaluated for the ability to mitigate such vulnerabilities to enable secure data transmission between UAVs and ground control systems. This paper addresses these interconnected. Recommendations are made for future research directions to further improve UAV security, including the need for interdisciplinary approaches and novel technologies to protect UAV operations over a wide range of applications.