1: \begin{abstract}

2: We define a large

3: class of continuous time multifractal random measures and processes with 

4: arbitrary log-infinitely divisible exact or asymptotic scaling law.

5: These processes generalize within a unified framework both the recently

6: defined log-normal Multifractal Random Walk (MRW) \cite{mrwepj,mrwpre} 

7: and the log-Poisson ``product of cynlindrical pulses" \cite{barral}.

8: Our construction is based on some ``continuous

9: stochastic multiplication'' (as introduced in \cite{schmitt}) 

10: from coarse to fine scales that

11: can be seen as a continuous interpolation of discrete

12: multiplicative cascades. 

13: We prove the stochastic convergence of the defined processes and

14: study their main statistical properties.

15: The question of genericity (universality) of limit

16: multifractal processes is addressed within

17: this new framework. We finally provide a method

18: for numerical simulations and discuss some specific examples.

19: \end{abstract}

20: