In this thesis, we study the important issues in query processing in a wireless sensor system in which sensor nodes are distributed to monitor the events occurred in the environment. Each sensor node maintains a database, called sensor database [MFH02], for the sensor data of its own as well as aggregated/derived data. The processing issues of various types of queries, such as real-time queries and continuous queries, are studied. In this thesis, we focus on the correctness issues in query results and the issues of minimizing data aggregation cost in query processing. We first study the correctness issue in processing continuous monitoring queries (CMQ) on sensor data values. Sensor data values are continuously generated from a sensor node. To ensure the correctness of query results, we adopt temporal consistency as the correctness criterion in accessing sensor data values from different sensor nodes. Then we propose a sequential aggregation scheme to minimize the aggregation overhead and processing overhead for data aggregation in execution of a CMQ. To demonstrate the efficiency of the proposed sequential aggregation scheme and study the implementation issue of this scheme, we have developed a prototype system using MICA Motes in which the sequential scheme is implemented and compared with a parallel aggregation scheme. Since the main purpose of an important application of sensor query processing is to generate timely reaction to respond to the critical events occurred in the system environment, real-time query is commonly used in such environment. Each real-time query is associated with a deadline on its completion time and it is an important performance objective to meet the deadline. Otherwise, the generation of timely responses can not be guaranteed. We extend the temporal consistency model and propose the Parallel Data Shipping with Priority Transmission (PAST) to meet the deadline of query and at the same time to maintain temporal consistency in its execution with minimum data transmission cost. Finally, we study the reliability issue in aggregating data versions for execution of real-time queries in a wireless sensor network in which the probability of data loss is high due to noises. In order to increase the reliability in aggregating data versions, we extend the Parallel Data Shipping with Priority Transmission (PAST) scheme to be workload sensitive (the new algorithm is called PAST with Workload Sensitivity (PAST-WS)) in selecting the coordinator node and the paths for transmitting the data from the participating nodes to the coordinator node for execution. PAST-WS considers the workload at each relay node to minimize the total cost and delay in data transmission.
Author: Pang, Chi Wai
Source: City University of Hong Kong
Download URL 2: Visit Now