Abstract: In Named Data Networking（NDN）, there is a great challenge for discovering and collecting the needful information from massive and distributed data, which needs to orchestrate complex routing and distributed processing components and adapt to network capabilities. Therefore, the paper proposes a differentiated service mechanism for NDN-based data query by following the design principle of the differentiated service in the current network. Inspired by the labeling idea of the MPLS protocol, the personalized and complex data query requirements are converted into multiple labels and their combination, which includes the interest forwarding labels and data returning labels. In the forwarding and response phase, the different processing actions of interest and response are determined by the individual labels and their combination. The NDN-based differentiated service system structure and the implementation process are proposed for personalized data query and accommodating network capabilities. Finally, the security of the labels and application scenarios of the differentiated service mechanism are discussed.

Key words: named data networking, distributed processing, data query, differentiated service, label

摘要： 在命名数据网络（NDN）中，从海量分布式数据中发现和收集所需的信息是一个巨大的挑战，需要编排复杂的路由和分布式处理组件并适应网络能力。因此，遵循当前网络中区分服务的设计原则，提出了一种基于NDN的数据查询区分服务机制。受MPLS协议标记思想的启发，将个性化、复杂的数据查询需求转换为多个标签及其组合，包括兴趣转发标签和数据返回标签。在转发和响应阶段，根据各个标签及其组合，确定不同的兴趣和响应的处理动作。提出了基于NDN的个性化数据查询和适应网络能力的区分服务系统结构及实现过程。讨论了标签的安全性和区分服务机制的应用场景。

关键词: 命名数据网络, 分布式处理, 数据查询, 区分服务, 标签