The Internet of Things (IoT) applications and services are increasingly becoming a part of daily life; from smart homes to smart cities, industry, agriculture, it is penetrating practically in every domain. Data collected over the IoT applications, mostly through the sensors connected over the devices, and with the increasing demand, it is not possible to process all the data on the devices itself. The data collected by the device sensors are in vast amount and require high-speed computation and processing, which demand advanced resources. Various applications and services that are crucial require meeting multiple performance parameters like time-sensitivity and energy efficiency, computation offloading framework comes into play to meet these performance parameters and extreme computation requirements. Computation or data offloading tasks to nearby devices or the fog or cloud structure can aid in achieving the resource requirements of IoT applications. In this paper, the role of context or situation to perform the offloading is studied and drawn to a conclusion, that to meet the performance requirements of IoT enabled services, context-based offloading can play a crucial role. Some of the existing frameworks EMCO, MobiCOP-IoT, Autonomic Management Framework, CSOS, Fog Computing Framework, based on their novelty and optimum performance are taken for implementation analysis and compared with the MAUI, AnyRun Computing (ARC), AutoScaler, Edge computing and Context-Sensitive Model for Offloading System (CoSMOS) frameworks. Based on the study of drawn results and limitations of the existing frameworks, future directions under offloading scenarios are discussed.

物联网 (IoT) 应用和服务正日益成为日常生活的一部分；从智能家居到智慧城市、工业、农业，它几乎渗透到各个领域。通过 IoT 应用程序收集的数据，主要是通过连接在设备上的传感器，随着需求的增加，不可能在设备本身上处理所有数据。设备传感器收集的数据量巨大，需要高速计算和处理，需要先进的资源。各种至关重要的应用程序和服务需要满足多个性能参数，如时间敏感性和能源效率，计算卸载框架开始发挥作用，以满足这些性能参数和极端计算要求。计算或数据卸载任务到附近的设备或雾或云结构可以帮助实现物联网应用的资源需求。在本文中，研究了上下文或情境对执行卸载的作用并得出结论，为了满足支持物联网的服务的性能要求，基于上下文的卸载可以发挥至关重要的作用。EMCO、MobiCOP-IoT、Autonomic Management Framework、CSOS、Fog Computing Framework等部分现有框架，基于其新颖性和最佳性能进行实现分析，并与MAUI、AnyRun Computing（ARC）、AutoScaler、边缘计算和卸载系统 (CoSMOS) 框架的上下文敏感模型。基于对现有框架的得出的结果和局限性的研究，