Robust Top-k Query Evaluation in Wireless Sensor Networks

Abstract

In many applications of sensor networks including environmental monitoring and security surveillance, a large volume of sensed data generated by sensors is needed to be either collected at the base station or aggregated within the network to respond to user queries. However, due to the high failure rate of wireless communication, robust query processing is very important for some critical queries in sensor networks by dealing with the messages loss in sensor networks. In this paper, we propose a novel localized and adaptive algorithm for robust topquery evaluation in wireless sensor networks, which strives the right tradeoff between the energy consumption and the accuracy of obtained results. We finally conduct extensive experiments by simulations to evaluate the performance of the proposed algorithm on real datasets. The experimental results show that the proposed algorithm is energy-efficient when meeting the specified result accuracy requirement.

Content-Based Wake-Up for Top-k Query in Wireless Sensor Networks

“…However, EXTOK is a filtering-based data collection/aggregation to be applied to the upper layer of protocol stack, and no wake-up and medium access control is considered. Local interactions among the sensors and/or spatial correlation among the readings are utilized in [27] [28]. Priority-based top-k monitoring (PRIM) [29] schedules the transmission of sensing data in the order of sensor readings and suspends transmissions after the data satisfies the top-k condition.…”

Segment linear network coding in wireless sensor networks

“…These sensor nodes are distributed and self-organized, working together for a common task under a limited wireless infrastructure [1] [2]. WSN has attracted a lot of interest for its wide range of uses such as military and medical applications [3] [4]. Most works in WSN focus on energy efficient strategies for transmissions between sensor nodes.…”

Secure distributed top-k aggregation

“…The top-k aggregation problem is relevant in many fields: sensor networks [8], [9], distributed databases [10], distributed systems [1], [11], [12], cloud computing [35] and metadata directed search in the network of information [13] to name a few. Security-wise, there is extensive literature on privacy-preserving top-k aggregation [36], [37] which we have also considered in previous work [38].…”

“…One can achieve sub-linear scalability properties when the aggregate is computed in-network by a network of cooperating nodes [1]- [7]. Here, we focus on the TOP−K distributed aggregation function which has numerous applications in sensor networks [8], [9], distributed databases [10] and distributed systems [1], [11], [12]. Our work is motivated by applications that include decentralized web site ranking and meta-data search in information networks [13].…”