<p>Blockchain technology has been widely adopted in application domains such as finance, the Internet of Things (IoT), and vehicular networks, where scalability requirements are continuously increasing. As a result, improving consensus efficiency while preserving security and consistency has become a central challenge for modern blockchain systems. Although existing surveys have reviewed mainstream consensus algorithms, a systematic analysis of the HotStuff protocol is still lacking. Derived from Practical Byzantine Fault Tolerance (PBFT), HotStuff achieves linear communication complexity and pipelined stage progression, leading to significant improvements in throughput and latency. This paper presents a systematic review of the performance evolution and optimization pathways of the HotStuff consensus protocol. It begins with a comparative analysis of the design principles and architectural structures of PBFT and HotStuff, and subsequently surveys existing studies from five key dimensions: protocol mechanism optimization, parallelization strategies, security enhancement, network adaptability, and application-driven extensions. Representative variants, including Fast-HotStuff, Dolphin, and Swift HotStuff, are analyzed in terms of performance characteristics and scalability improvements. Finally, future research directions are discussed, covering multi-layer pipelining, cross-layer collaboration, heterogeneous scalability, and post-quantum security. This survey provides a structured and comprehensive reference for understanding the optimization landscape and future evolution of HotStuff-based consensus protocols, thereby supporting their efficient and secure deployment in large-scale distributed environments.</p>