In this research, a supply chain system consisting producer, distribution centers and retailers is modeled by considering lateral transshipment between distribution centers and also using echelon-based inventory system instead of independent inventory system. The model is developed in the form of mixed integer non-linear programming (MINLP) to minimize the total location, transportation, and inventory costs of the system. The model was solved using conic programming approach and validity was examined by comparing the developed model with the basic model (i.e. the model without the contributions) in several instances with different sizes of distributors and retailers. The results represent superiority of the developed model in computational time and objective value especially in medium and large-scale problems.