Intrusion Detection Systems (IDS) detect computer security intruders and notify system administrators of malicious activities. IDS identifies intruders and notifies system administrators about the threat. Wireless IDS must be reviewed separately from cable LAN because most wireless LAN security vulnerabilities are intrinsic. The researchers have not published the Aegean Wi-Fi Intrusion Dataset (AWID) or analyzed it using machine learning. AWIDs contain numeric, text, and hexadecimal data. Following the preprocessing stage, this dataset has 102 attributes that are used for training the system as well as assessment. To reduce training expenses and maximize system performance, a process employed to select features in 2 states to identify and implement the fewest differentiating traits. Eliminating redundant attributes comes first. The dataset is decreased to 68. The core classification process is the Deep Belief Network (DBN), which uses a stack of Restricted Boltzmann Machines (RBM) to start unsupervised pre-training and a BPNN to fine-tune DBN parameters. After system design and deployment, metrics-based tests were run to evaluate performance. The DBN-based approach had 98.06% classification accuracy.