A comparative analysis of algorithms for dynamic web services composition with quality of service

• Luis Hideo Vasconcelos Nakamura USP
• Andre Luiz Verucci Cunha USP
• Júlio Cezar Estrella USP
• Marcos José Santana USP
• Regina Helena Carlucci Santana USP

Resumo

In this paper three optimization algorithms were used in order to solve the problem of dynamic composition of Web services with Quality of Service (QoS). This combinatorial optimization problem arises when multiple services, which provide subfunctions for a complete function, are aggregated in an execution flow and presented to the user as a single service. This problem has been modeled in the context of two deterministic and one stochastic algorithms, aiming to determine the best flow in terms of QoS. Besides, a performance evaluation was executed considering the composition algorithms in some scenarios and comparing them in terms of response time and quality of the obtained solution. The stochastic algorithm proved to be more advantageous for these scenarios due the deadline exploitation on the search for an optimal solution, although it does not provide guarantees of optimality.