Indexing Techniques for Power Management inMulti - Attribute Data

In this paper, we discuss the power conservative indexing techniques for managing multi-attribute data broadcast on wireless channels. These indexing techniques, namely, index tree, signature and hybrid, aim at improving the battery power consumption of mobile clients. By taking into account the broadcast management factors such as clustering and scheduling, these three indexing schemes may signii-cantly reduce tune-in time while maintaining a reasonable access time. Cost models for single and multi-attribute query processing are developed. Our performance evaluation shows that the signature and hybrid methods are superior to the index tree method.