Low Latency and High Reliable Data Query Mechanism in Dynamic Wireless Sensor Networks

Abstract

As an important operation of intelligent services in large-scale wireless sensor networks, query processing can retrieve and return perceived data in the network according to user requirements. However, when wireless sensor networks are deployed in harsh environments, nodes are vulnerable to external forces, or its self- resources (energy, storage, etc.) are limited, which may lead to nodes displacement and failure, resulting in constant changes in the network topology and the failure of perceived data of some nodes. At the same time, due to the large capacity of sensor data, limited transmission bandwidth and unreliable network links, the network communication delay may be greatly increased. These factors make fast and reliable data query processing a difficult problem in wireless sensor networks. In order to solve this problem, a low latency and high reliability data query mechanism in dynamic networks is proposed. The mechanism is a non-aggregated random query method, which can realize data query by dividing sensor nodes into source nodes and query nodes. Firstly, the network is divided into several sub-regions according to the monitoring events. The source nodes in each sub-region cooperate with each other, and listen to the event information of the region in turn according to the time sequence; then, according to the estimated average node failure probability, the source nodes calculate a reasonable number of backups, and stores the source data in the neighbor nodes according to the number, so as to reduce the failure probability of the source data; secondly, in order to speed up the data query speed, the source nodes periodically encode and compress the source data block, and select multiple neighbor nodes with less residual energy and storage space as the next hop receiving nodes. These receiving nodes decide whether to receive and store the message based on the number of nodes in the local area. The above process is repeated until the compressed data is uniformly distributed in the network. On the other hand, when the query node receives query request, it also uses the method of load balancing multi-path dissemination to transmit the query request to some nodes. In order to avoid the redundant return of the target data, when the query request succeeds in querying the target data, the target node first modifies the access bit, then selects a node with the smallest distance between its neighbor nodes and the query node as the next hop receiving node, and performs the above operation iteratively until the user obtains the required event information. In the above process, in order to save the energy of the nodes, under the condition of ensuring high successful query rate, an optimization model of minimizing communication energy consumption is established, and the optimal number copies of the compressed data and the query request are calculated, then the source nodes and query nodes disseminate its replica data according to the number respectively. Finally, theoretical analysis and experimental results show that compared with other four query algorithms, the proposed query mechanism has higher successful query rate, lower communication energy consumption and communication delay.