Dear Venkata, in multi-period deterministic inventory problems without capacity constraints stock-out will never happen, so it make no sense to calculate a service level. The optimal solution can be found thought the Wagner-Whitin algorithm, while good solutions can be found through the well-known heruistics such as Silver-Meal or PPB.

On the contrary, if capacity constraints have to be considered (i.e. each order has a higher bound that corresponds f.e. to the plant capacity) you can calculate the achieved service level as the ratio between the fulfilled demand and the total demand during a certain time horizon (e.g a year).  

Dear Venkata,

Suggestion:

Since service level  refers to the probability of being able to service incoming orders(or demand) within a period without delay from stock on hand; why not define service level(SL) as: No. of periods when customer demand is fully met(p) divided by Total Number of periods(N)

   ie.  SL = p/N