Abstract: With the rise of cloud computing, service computing becomes the key technology of cloud computing, and academic field has new thinking about service composition. This paper proposes an adaptive service selection method, which meets atomic transactions and QoS constraints. When looking for an optimal composite path, this method considers the QoS constraints and the transaction constraints among services, and improves reliability by global optimization combined with proper coordination mode selection strategies. Firstly, this paper lays down adaptive modeling rules and QoS models, and cites other scholars’ transaction models and processing rules. Secondly, this paper analyzes the adaptive service selection method mechanism with transaction and QoS constraints. Finally, this paper evaluates the effectiveness and robustness of the proposed method by experiments.

Key words: cloud computing, service composition, QoS constraints, transaction constraints, adaptation

摘要： 随着云计算的兴起，服务计算成为其实现的一种关键技术，学术领域对服务组合有新的思考。提出了满足原子事务的服务质量（quality of service，QoS）感知的自适应服务选择方法，既考虑QoS约束，也考虑服务间的事务约束，在进行全局优化选择时结合适当的协调模式选择策略，以进一步提高可靠性。首先制定自适应规则和建立QoS模型，并引用其他学者关于事务构建和规约规则的模型。接着分析事务与QoS约束下自适应服务选择算法机制。最后通过实验评价了方法的有效性和健壮性。

关键词: 云计算, 服务组合, QoS约束, 事务约束, 自适应