Landmark-based data storage and retrieval scheme is presented in [9]. This scheme constructs a number of globalized shortest path trees of predefined landmark nodes, each rooted at a landmark node. A data producer hashes data (according to data type) to a certain landmark node and distributes the data by using the shortest path tree, rooted at the landmark node. A data consumer queries along the same shortest path tree; it gets the data when it hits the storage path or reaches the hash node. This scheme generates storage hot spots and involves many expensive global operations. It provides no energy efficiency and lacks scalability.