In the context of network security, when we say
adaptive network, what does it really imply? In other words, what does an adaptive network do to be adaptive? Is it similar to saying dynamic network?
To my knowledge, there is no difference in terminology between "adaptive" network security and "dynamic" network security. It is safe to believe that both reference the same thing.
Adaptive security, or real-time adaptive security, is a network security model that is able to respond to changes in and attacks against the network without human intervention. This is oftentimes implemented by utilizing communication between different security devices on the network. Here's an example.
Let's say an attacker attempts to send an exploit to a web server. An IDS on the target network is able to detect and stop the attack successfully. The IDS could then communicate with the firewall and tell it to block all incoming connections from that IP. This way, the attacker is locked out of the system completely and effectively blacklisted without the need for human intervention.
IMHO there are differences between dynamic and adaptive security. IMHO, dynamic implies a means to change policies on the fly based on a negative(pre-defined) policy set. This is like vectorizing a problem domain based on specific events. Think, why would I add policy for SMTP when the traffic is clearly HTTP. So if I detect a HTTP VERB then apply HTTP policy instead of the normal SMTP policy attached to the process.
Adaptive is Dynamic except in my experience policy decisions are backed by machine learning, Analytics and some minimal cognition. There are many companies that talk about adaptive but IMHO are really dynamic methods of the same.
Think menu driven choices for a topic that narrow as you select vs. a conversation about the same topic with inductive (some cases) and deductive reasoning driving selection.