1. Four inputs are needed: (1) Time, (2) Node i, (3) Node j, (4) Transfer value
   
   
   • Transfer value can be money, passengers, or information from i to j where i & j are numbers (1, 2, …) that distinguish the nodes, whether they are people, computers, or airports.
   • These four variables will be repeated in millions and millions of records corresponding to each data transfer
     
     
2. A window of time is defined to add up these transfers into a collection as a matrix.
   
   
   • For example, the financial transfers done between 8:00 and 9:00am, then next, those between 9:00 and 10:00am and so on.
   • The transfers are added up into a ‘connection matrix’ for each window: Cij, for that time.
   • There are a few user adjustable parameters depending on the application.
     
     
3. The software processes C over time:
   
   
   • It reduces C to a few representative values for network monitoring, and plots as a graph.
   • Entropy levels for inputs and outputs of all nodes are computed and plotted.
   • The curves are correlated over time to summarize behavior.
   • Other analysis is performed on the entropy structures

Types of Networks That Can Be Tracked and Analyzed

• Communication Networks
  • Internet Traffic
  • Phone - both wireless & wired
  • Mail, UPS, Fed-Ex
    
    
• Transportation Networks
  • Air Traffic
  • Highway Traffic
  • Waterways
  • Railroads
  • Pipelines
    
    
• Financial Networks
  • Banking Transactions
  • Accounting Flows
  • Ownership & Investment Networks
  • Input-Output Economic Flows
    

• Utility & Energy Networks
  • Electrical Grids
  • Water & Sewer Flows
  • Natural Gas & Petroleum Distribution
    
    
• Social Networks
  • Organizational & Business Structures
  • Terrorist & Criminal Networks
    
    
• Manufacturing Processes
  • Process & Manufacturing Workflow
    
    
• Biological & Ecological Networks
  
  
• Disease & Health Networks
  
  
• Electrical Networks