Multi-scale network traffic generator

1. A network traffic generation method comprising:

defining multiple, different scales in an n-state Markov modified Poisson process (MMPP) model to accommodate a full characteristic response of a modeled traffic scenario;

organizing different transition probabilities into a probability transition matrix correlating different transition probabilities with respectively different states of the MMPP model, the matrix responding to a desired traffic pattern by producing a selected one of a multiplicity of transition windows for use in generating real-time network traffic;

submitting a traffic pattern to the matrix in order to receive a corresponding one of the transition windows;

determining a state from the corresponding one of multiplicity of transition windows;

computing an inter-packet time according to the determined; and,

generating and transmitting packets utilizing the computed inter-packet time in a multi-scale, multi-window network traffic generator executing in memory by a processor of a server host and configured to generate packets emulating the self-similar nature of network traffic.

2. The method of claim 1 , wherein a transition window is established for each of the scales according to τ i =[ρ min i ,ρ max i+1 ].

3. The method of claim 2 , wherein the state is determined through the transition window for selected ones of the scales, by comparing a random probability distribution ρ i to the transition window τ i =[ρ min i ,ρ max i+1 ] determine the state according to ρ i >ρ min 1 , in consequence of which a packet is to be generated with an inter-packet time in a bursty state, ρ i >ρ max 2 in consequence of which a packet is to be generated with an inter-packet time in an idle state; and ρ max 2 <ρ i <ρ min 1 in consequence of which no packet is to be generated.

4. A network traffic generation system comprising:

a computer with memory and at least one processor;

a Markov modified Poisson process (MMPP) model stored in the memory;

a packet scheduler coupled to the MMMP model and executing in the computer;

a data store of transition windows defined for different defined scales coupled to the computer;

a probability transition matrix disposed in the data store and organizing different transition probabilities in correlation with respectively different states of the MMPP model, the matrix responding to a desired traffic pattern by producing a selected one of a multiplicity of transition windows for use in generating real-time network traffic;

traffic generation parameter computing logic stored in a memory comprising program code enabled upon execution within the memory by the processor of the computer to compute traffic generation parameters for different scales according to respective states identified within different transition windows in the data store for the different scales by submitting a traffic pattern to the matrix in order to receive a corresponding one of the transition windows, by determining a state from the corresponding one of multiplicity of transition windows, by computing an inter-packet time according to the determined state and by generating and transmitting packets utilizing the computed inter-packet time; and,

a packet transmitter coupled to the packet scheduler and also executing in the computer.

5. A computer program product comprising a non-transitory computer readable storage device including computer usable program code for network traffic generation, the computer program product including:

computer usable program code for defining multiple, different scales in an n-state Markov modified Poisson process (MMPP) model to accommodate a full characteristic response of a modeled traffic scenario;

computer usable program code for organizing different transition probabilities into a probability transition matrix correlating different transition probabilities with respectively different states of the MMPP model, the matrix responding to a desired traffic pattern by producing a selected one of a multiplicity of transition windows for use in generating real-time network traffic;

computer usable program code for submitting a traffic pattern to the matrix in order to receive a corresponding one of the transition windows;

computer usable program code for determining a state from the corresponding one of multiplicity of transition windows;

computer usable program code for computing an inter-packet time according to the determined state; and,

computer usable program code for generating and transmitting packets utilizing the computed inter-packet time in a multi-scale, multi-window network traffic generator executing in memory by a processor of a server host and configured to generate packets emulating the self-similar nature of network traffic.

6. The computer program product of claim 5 , wherein the transition window is established for each of the scales according to τ i =[ρ min i ,ρ max i+1 ].

7. The computer program product of claim 6 , wherein the state is determined through the transition window for selected ones of the scales, by comparing a random probability distribution ρ i to the transition window τ i =[ρ min i ,ρ max i+1 ] to determine the state according to ρ i >ρ min 1 in consequence of which a packet is to be generated with an inter-packet time in a bursty state, ρ i <ρ max 2 in consequence of which a packet is to be generated with an inter-packet time in an idle state; and ρ max 2 <ρ i <ρ min 1 in consequence of which no packet is to be generated.