Strategic customer behavior is strongly influenced by the level of information that is provided to customers. Hence, to optimize the design of queueing systems, many studies consider various versions of the same service model and compare them under different information structures. In particular, two extreme versions are usually considered and compared: the observable in which customers are informed about the number of customers in the system and the unobservable in which they are only informed about the system parameters, e.g., arrival and service rates. In the present work, we study a model that bridges these two versions. More concretely, we assume that the system alternates between observable and unobservable periods. We characterize and compute customer equilibrium joining/balking strategies and show that the present model unifies and extends existing approaches of both heterogeneously observable models and models with delayed observations. More importantly, our findings indicate that an alternating information structure implies in general higher equilibrium throughput and social welfare in comparison to both the observable and unobservable cases. We complement our results with numerical experiments and provide managerial insight on the optimal control of the system parameters.

中文翻译：

---

具有交替信息结构的排队系统中的战略客户行为

战略客户行为受到提供给客户的信息水平的强烈影响。因此，为了优化排队系统的设计，许多研究考虑了同一服务模型的不同版本，并在不同的信息结构下进行比较。特别是，通常会考虑和比较两个极端版本：可观察的，其中客户被告知系统中的客户数量，以及不可观察的，其中他们只被告知系统参数，例如到达和服务费率。在目前的工作中，我们研究了一个连接这两个版本的模型。更具体地说，我们假设系统在可观察和不可观察的时期之间交替。我们表征和计算客户均衡加入/阻止策略，并表明本模型统一和扩展了异构可观察模型和具有延迟观察的模型的现有方法。更重要的是，我们的研究结果表明，与可观察和不可观察的情况相比，交替信息结构通常意味着更高的均衡吞吐量和社会福利。我们通过数值实验补充我们的结果，并提供有关系统参数优化控制的管理见解。我们的研究结果表明，与可观察和不可观察的情况相比，交替信息结构通常意味着更高的均衡吞吐量和社会福利。我们通过数值实验补充我们的结果，并提供有关系统参数优化控制的管理见解。我们的研究结果表明，与可观察和不可观察的情况相比，交替信息结构通常意味着更高的均衡吞吐量和社会福利。我们通过数值实验补充我们的结果，并提供有关系统参数优化控制的管理见解。