Different node attacks like black-hole attacks, hello flooding attacks, and denial of service (DoS) in a MANET are modeled and simulated in OMNET++ for different mobility models and node speed. Obtained results are analyzed in terms of data packet delivery ratio (PDR), end-to-end delay, network throughput, and energy consumption. This helps us understand the impact of different node attacks in MANET and can be used to enhance the performance of the existing protocols. Moreover, this project also presents an offline intrusion detection system (IDS) using an artificial neural network (ANN) that identifies the attacker node in the MANET. This is achieved by classifying normal and threat patterns, and the approach is validated using data obtained from OMNET++ simulation. The results show that the IDS is 97.80% accurate and can detect attacker nodes successfully.