Progress.txt

Things that will need to be fixed:

Service.py - gen_service creates vulnerabilities in a way that allows future versions to have more vulnerabilties
"When a new vulnerability is introduced for a service at a specific version, the attack complexity AC and impact I are randomly selected. This aspect was inspired
by the CVSS scoring system, which is a widely adopted metric in the cyber security industry to describe the severity of known vulnerabilities [30]. The vulnerability
will then remain in proceeding versions of the service since the vulnerability was introduced for a mean of 10 versions, at which point it was assumed that the vendor
of the service would of identified and patched the vulnerability."

Host 1 2 3 compromised, when 2 is swapped, attacker can only reach 1?

Variance of entry points 

1. Add attack scenario where targetted node
2. Reclassify diversity techniques 


Research Context Manager (Action Manager)
Research Topological sort
Combine IP and Port Scan
2 Weeks - New Attack Scenario (attacker knows the target host when neighbouring) Set the distance for target node?
2 weeks - Diversity

Things that need to be done:

Practical Applications of Shuffle Techniques
    Port Shuffle - Benchmark with ???
    Complete Topology Shuffle - Can't use EPV as every host is target? 
    Host Shuffle - ???
    OS Shuffle - ???
    Service Shuffle - Benchmark against Expected Path Variation (Need to calculate probability of each host of compromise?)
Developing Diversity MTD Techniques
    Code Diversity - Every Service has x configurations of vulnerabilities 
    Software Stack Diversity - Every Host x configurations of services it cycles through?
    Programming Language Diversity 
    Network Diversity - Deploy different network configurations (What is difference to network topology shuffle?)
Implement Dynamic Security Measures
    Attack Path Variation (APV) 
    Attack Path Exposure (APE)

RoA Thresholding?
Non-Important Fixes/Additions: 

Network.py - Different Network Generations than just Barabasi-Albert Random Graph model
Network.py - Network Generation code optimisations

OS Shuffle -> OS Diversity
Service Shuffle -> Service Diversity? (Needs to be changed)

Add attack scenario where targetted node Breakdown
	Add Constant to Host.py of Targetted State
	Network.py: set function that accepts an int < total_layer that sets a single host as "target" in selected layer
	Hacker.py: Set "target" host to top of attack list if discovered
	Scorer.py: Add Target Metrics to scorer (At time target was seen, time target was compromised) 
	Develop new metrics for effectiveness of shuffle techniques:
		IP Shuffle: Works Same as before
		Port Shuffle: Works Same as before
		Complete Topology Shuffle: 

		Host Shuffle:
			 
		OS Diversity:
		Service Diversity:



Attackers pop top from Host.stack (hacker.py)

Attacker Scan Network
	Hacker.py start_network_enum 
		Network.scan(compromised_hosts)
		*Gets list of hosts (uncompromised_hosts)
		Creates Action
			Actions.py (create Action with check_ips + check_paths)
-> Performs Port Scan 
-> Selects Host?	


When the exposed nodes are assigned/labelled -> Apply topological Sort -> Partition Network (add tags based on proportion)
		
Attacker Priority -> Target -> Tag -> Random 





1a New Attack Scenario (Target node)
1b Needs Connection + Compromise not just compromised
2. Metrics