Mobile Intelligence

24.4 LOAD-BALANCED MECHANISMS

The network lifetime is one of the most important issues in a WSN. However, the lifetime of a WSN highly depends on the degree of load-balance among all sensor nodes. A load-balanced WSN also implies that the WSN is energy-balanced so that the lifetime of each sensor is about the same. The main source of workloads in a WSN is the information sensing, storing, retrieving, and routing. A data-centric node that is responsible for storing a sensing information is also responsible for handling the query request for this information and, therefore, consumes more energy than the other nodes. Other than the above-mentioned basic and data-replica mechanisms, the load-balanced mechanisms aim to distribute the workloads of data storing and retrieving to all sensor nodes in a load-balanced manner. The following illustrates some important data-centric mechanisms that consider the load-balanced issue in their designs.

In a WSN, an event can be generally described as a tuple of k attribute values, A₁, A₂, …, A_k, where each attribute A_i represents a sensor reading, or some value corresponding to a detection. Li et al. developed a distributed index for multidimensional data (DIM) [9], which takes the load-balance issue into account. The main idea behind the DIM is to partition the WSN region according to a specific coding method and assign each region with a code. According to the values of the multi-attribute event, the event information can be mapped to a code that ...