Many modern companies, economic agents, and service providers, including cloud computing providers, offer increasingly sophisticated pricing plans for their services. In a competitive environment and to ensure the most efficient allocation of resources to customer needs, they offer a variety of bonuses, promotions and discounts, such as multi-tier offers, sustainability bonuses, quantity and volume discounts and many others. In addition, there are the obvious administrative and performance benefits of using multiple resources from a single cloud provider. In this paper, we describe such aggregated price and performance benefits in terms of group dependencies. For the emerging cloud resource allocation problem, we propose a dynamic programming algorithm to optimize resource selection considering resource group dependencies on price and performance criteria. The proposed approach has been integrated into CloudSim simulator and has shown its optimization efficiency against brute force, knapsack and more simplified algorithms in different simulation scenarios.