Adaptive IDS Concept with PRBS Multi Inputs Multi Outputs (MIMO) and Matched Filtering Algorithm

Latest development in cyber-attacks are becoming more complex and obscure. Current Deep Learning approaches in network intrusion is limited to experimental conditions, limited training on real time live traffic, manual training interventions, and highly dependent on standard dataset. Notwithstanding, adaptive system has grabbed the attention of multiple industrial players from automobiles to cybersecurity. This is due to the capability of the system to adapt to any updates, trends and changes. Hence Adaptive IDS must be studied and designed immediately to prevent further damage to cyber infrastructure and community. Adaptive IDS will automate process and contrast live traffic to spring realistic output. This work discusses the conceptual idea of Adaptive IDS in cybersecurity domain. The discussion covers three (3) parts, with initial directly related to the overview of the Adaptive IDS concept. Next part is an expanded version that discusses on the backend software that resides in the computer system and a bit of hardware outside. Finally, the third part deliberates on designing a simulated environment through PRBS. This allows synthetic composite dataset generation and detection, employing modified $\mathrm{n}=8$, giving ($255+1$) 256BitsPRBS using Embarcadero Delphi IDE VCL application development in contrast to electronic version generation using Simulink or PowerSim.