Application of SPEA2 algorithm in Web services selection

Abstract

The problem of Web services selection based on quality of service (QoS) hasn't be essentially solved by the single objective optimal algorithm which optimizes service selection by aggregating multiple QoS parameters to form a composite objective function using weighted scoring method. This paper presents a Web services selection algorithm of QoSaware and global multi-objective optimization, termed WS-SPEA2. The essence of the proposed algorithm is that the problem of Web services selection based on QoS is transformed into a multi-objective services composition optimization problem with QoS constraints. A strength Pareto evolutionary algorithm (SPEA2) is utilized to produce a set of Pareto-optimal solutions by means of simultaneously optimizing a series of objective functions, that is, minimizing service cost(C) and service time(T) with the constaint of satifying the parameters reputation(RE), reliability(R) and availability(A). And the users can select one from these solutions for their preferences. The results of a series of simulation experiments indicate the feasibility and efficiency of this algorithm.

Advances on QoS‐aware web service selection and composition with nature‐inspired computing

“…Another well‐known multi‐objective EA, i.e. strength Pareto evolutionary algorithm (SPEA2) [100] is also tailored for web services composition problem [101]. It minimises the cost and service time of composite service with the constraints of reputation, reliability and availability.…”

Multi-objective Pointer Network for Combinatorial Optimization

“…Furthermore, individual recombination based on genetic operators and few hyperparameters make MOEA able to balance exploitation and exploration effectively. NSGAII [11] and SPEA2 [12], two well-known MOEAs, have been widely employed in many real-world applications [13], [14]. To improve the diversity of solutions when solving multi-objective optimization problems (MOOPs), evolutionary algorithms (EAs) based on decomposition have been proposed, such as NSGAIII [15] and MOEA/D [16], to maintain subpopulations that are uniformly distributed in objective space, in order to ensure that their individuals are as uniformly close to the Pareto front as possible.…”