Wireless network carriers are motivated to offer more services tailored to mobile user's needs. Context awareness is one of the value-added services that have attention from both mobile users and network carriers. However, context awareness requires collecting, analyzing, and understanding of mobile user characteristics, activities, location, surrounding environment, and so on. The objective of this paper is to design a dynamic neural network based algorithm updating the context cache weights. The context cache weights are generated to reflect the mobile user's context change. In addition, a data cache scheme is proposed to handle mobile user context. Particularly, we introduce the concept of virtual flame and reference flame on which the data cache scheme is based. The user context profiles that consist of mobile user characteristics, wireless device configuration, activities, and surrounding environment classify the context entities in the cache. As such, a systematic approach can be used to handle context awareness at the client site.
