A Comparative Study of Various Intrusion Detections In Smart Cities Using Machine Learning

Smart Cities demand highly scalable and connected technologies to operate at multiple distributed locations. Despite the enormous potential it brings to citizens' lives, security as well as privacy concerns must still be addressed. A huge volume of data is formed as a result of technological advancements, and machine learning techniques are used to understand useful patterns. Recent advances in ML methods make it possible to handle the huge amount of complex data, enable fast processing, robustness by decentralization, and scalability of real-world systems. In recent decades, cyber security has appeared as an important consideration throughout scientific investigation. According to the World Internet Statistics Report, the Internet grew by 1,114 % from 2000 to 2019, with more than 2 quintillion bytes of data formed daily basis. It also demonstrates that rate of data advancement out of multiple sources is incredibly quick, while hacking tools as well as techniques are also developing at a rapid pace. Machine Learning techniques identify network intrusions and threats by forecasting hazard utilizing data training. The demand for the internet is growing by the day, raising concerns about network security. By blocking malicious things in the network system, an Intrusion Detection System (IDS) can furnish solutions to several rapidly growing network attacks (e.g., DDoS attack, Ransom ware attack, Botnet attack, and so on). This systematic review summarizes latest framework studies in sensing DDoS threats and investigates a comprehensive evaluation of numerous Ml techniques employed throughout Smart City applications. Furthermore, the goal of this work is to categorize the intrusions utilizing ML algorithms.