Sensor Field Resource Management for Sensor Network Data Mining

This research is motivated by data mining for wireless sensor network applications. The authors consider applications where data is acquired in real-time, and thus data mining is performed on live streams of data rather than on stored databases. One challenge in supporting such applications is that sensor node power is a precious resource that needs to be managed as such. To conserve energy in the sensor field, the authors propose and evaluate several approaches to acquiring, and then caching data in a sensor field data server. The authors show that for true real-time applications, for which response time dictates data quality, policies that emulate cache hits by computing and returning approximate values for sensor data yield a simultaneous quality improvement and cost saving. This “win-win” is because when data acquisition response time is sufficiently important, the decrease in resource consumption and increase in data quality achieved by using approximate values outweighs the negative impact on data accuracy due to the approximation. In contrast, when data accuracy drives quality, a linear trade-off between resource consumption and data accuracy emerges. The authors then identify caching and lookup policies for which the sensor field query rate is bounded when servicing an arbitrary workload of user queries. This upper bound is achieved by having multiple user queries share the cost of a sensor field query. Finally, the authors discuss the challenges facing sensor network data mining applications in terms of data collection, warehousing, and mining techniques.

Knowledge Discovery for Sensor Network Comprehension

In this chapter we explore different characteristics of sensor networks which define new requirements for knowledge discovery, with the common goal of extracting some kind of comprehension about sensor data and sensor networks, focusing on clustering techniques which provide useful information about sensor networks as it represents the interactions between sensors. This network comprehension ability is related with sensor data clustering and clustering of the data streams produced by the sensors. A wide range of techniques already exists to assess these interactions in centralized scenarios, but the seizable processing abilities of sensors in distributed algorithms present several benefits that shall be considered in future designs. Also, sensors produce data at high rate. Often, human experts need to inspect these data streams visually in order to decide on some corrective or proactive operations (Rodrigues & Gama, 2008). Visualization of data streams, and of data mining results, is therefore extremely relevant to sensor data management, and can enhance sensor network comprehension, and should be addressed in future works.

Anomaly Detection in Streaming Sensor Data

This chapter considers a cell phone network as a set of automatically deployed sensors that records movement and interaction patterns of the population. The authors discuss methods for detecting anomalies in the streaming data produced by the cell phone network. The authors motivate this discussion by describing the Wireless Phone Based Emergency Response (WIPER) system, a proof-of-concept decision support system for emergency response managers. This chapter also discusses some of the scientific work enabled by this type of sensor data and the related privacy issues. The authors describe scientific studies that use the cell phone data set and steps we have taken to ensure the security of the data. The authors also describe the overall decision support system and discuss three methods of anomaly detection that they have applied to the data.

A Survey of Dynamic Key Management Schemes in Sensor Networks

In sensor networks, the large numbers of tiny sensor nodes communicate remotely or locally among themselves to accomplish a wide range of applications. However, such a network poses serious security protocol design challenges due to ad hoc nature of the communication and the presence of constraints such as limited energy, slower processor speed and small memory size. To secure such a wireless network, the efficient key management techniques are important as existing techniques from mobile ad hoc networks assume resource-equipped nodes. There are some recent security protocols that have been proposed for sensor networks and some of them have also been implemented in a real environment. This chapter provides an overview of research in the area of key management for sensor networks mainly focused on using a cluster head based architecture. First we provide a review of the existing security protocols based on private/public key cryptography, Kerberos, Digital signatures and IP security. Next, the authors investigate some of the existing work on key management protocols for sensor networks along with their advantages and disadvantages. Finally, some new approaches for providing key management, cluster head security and dynamic key computations are explored.

Intelligent Querying Techniques for Sensor Data Fusion

Sensor data fusion imposes a number of novel requirements on query languages and query processing techniques. A spatial/temporal query language called SQL has been proposed to support the retrieval of multimedia information from multiple sources and databases. This chapter investigates intelligent querying techniques including fusion techniques, multimedia data transformations, interactive progressive query building and SQL query processing techniques using sensor data fusion. The authors illustrate and discuss tasks and query patterns for information fusion, provide a number of examples of iterative queries and show the effectiveness of SQL in a command-action scenario.

Query Optimisation for Data Mining in Peer-to-Peer Sensor Networks

One of the more recent sources of large volumes of generated data is sensor devices, where dedicated sensing equipment is used to monitor events and happenings in a wide range of domains, including monitoring human biometrics and behaviour. This chapter proposes an approach and an implementation of semi-automated enrichment of raw sensor data, where the sensor data can come from a wide variety of sources. The authors extract semantics from the sensor data using their XSENSE processing architecture in a multi-stage analysis. The net result is that sensor data values are transformed into XML data so that well-established XML querying via XPATH and similar techniques can be followed. The authors then propose to distribute the XML data on a peer-to-peer configuration and show, through simulations, what the computational costs of executing queries on this P2P network, will be. This approach is validated approach through the use of an array of sensor data readings taken from a range of biometric sensor devices, fitted to movie-watchers as they watched Hollywood movies. These readings were synchronised with video and audio analysis of the actual movies themselves, where we automatically detect movie highlights, which the authors try to correlate with observed human reactions. The XSENSE architecture is used to semantically enrich both the biometric sensor readings and the outputs of video analysis, into one large sensor database. This chapter thus presents and validates a scalable means of semi-automating the semantic enrichment of sensor data, thereby providing a means of large-scale sensor data management which is a necessary step in supporting data mining from sensor networks.

Integrated Intelligence

Warehousing and analytics of sensor network data is an area growing in relevance as more and more sensor data are collected and made available for analysis. Applications that involve processing of streaming sensor data require efficient storage, analysis, and monitoring of data streams. Traditionally, in these applications, RDBMSs have been confined to the storage stage. While contemporary RDBMSs were not designed to handle stream-like data, the tight integration of sophisticated analytic capabilities into the core database engine offers a powerful infrastructure that can more broadly support sensor network applications. Other useful components found in RDBMs include: extraction, transformation and load (ETL), centralized data warehousing, and automated alert capabilities. The combination of these components addresses significant challenges in sensor data applications such as data transformations, feature extraction, mining model build and deployment, distributed model scoring, and alerting/messaging infrastructure. This chapter discusses the usage of existing RDBMS functionality in the context of sensor network applications.

Event/Stream Processing for Advanced Applications

Event processing in the form of ECA rules has been researched extensively from the situation monitoring viewpoint to detect changes in a timely manner and to take appropriate actions. Several event specification languages and processing models have been developed, analyzed, and implemented. More recently, data stream processing has been receiving a lot of attention to deal with applications that generate large amounts of data in real-time at varying input rates and to compute functions over multiple streams that satisfy quality of service (QoS) requirements. A few systems based on the data stream processing model have been proposed to deal with change detection and situation monitoring. However, current data stream processing models lack the notion of composite event specification and computation, and they cannot be readily combined with event detection and rule specification, which are necessary and important for many applications. This chapter discusses a couple of representative scenarios that require both stream and event processing. The authors then summarize the similarities and differences between the event and data stream processing models. The comparison clearly indicates that for most of the applications considered for stream processing, event component is needed and is not currently supported. And conversely, earlier event processing systems assumed primitive (or simple) events triggered by DBMS and other applications and did not consider computed events. By synthesizing these two and combining their strengths, the authors present an integrated model – one that will be better than the sum of its parts. The authors discuss the notion of a semantic window, which extends the current window concept for continuous queries, and stream modifiers in order to extend current stream computation model for complicated change detection. They further discuss the extension of event specification to include continuous queries. Finally, the authors demonstrate how one of the scenarios discussed earlier can be elegantly and effectively modeled using the integrated approach.

Peer-to-Peer Data Clustering in Self-Organizing Sensor Networks

This work proposes and evaluates distributed algorithms for data clustering in self-organizing ad-hoc sensor networks with computational, connectivity, and power constraints. Self-organization is essential in environments with a large number of devices, because the resulting system cannot be configured and maintained by specific human adjustments on its single components. One of the benefits of in-network data clustering algorithms is the capability of the network to transmit only relevant, high level information, namely models, instead of large amounts of raw data, also reducing drastically energy consumption. For instance, a sensor network could directly identify or anticipate extreme environmental events such as tsunami, tornado or volcanic eruptions notifying only the alarm or its probability, rather than transmitting via satellite each single normal wave motion. The efficiency and efficacy of the methods is evaluated by simulation measuring network traffic, and comparing the generated models with ideal results returned by density-based clustering algorithms for centralized systems.

Why General Outlier Detection Techniques Do Not Suffice for Wireless Sensor Networks

Raw data collected in wireless sensor networks are often unreliable and inaccurate due to noise, faulty sensors and harsh environmental effects. Sensor data that significantly deviate from normal pattern of sensed data are often called outliers. Outlier detection in wireless sensor networks aims at identifying such readings, which represent either measurement errors or interesting events. Due to numerous shortcomings, commonly used outlier detection techniques for general data seem not to be directly applicable to outlier detection in wireless sensor networks. In this chapter, the authors report on the current stateof- the-art on outlier detection techniques for general data, provide a comprehensive technique-based taxonomy for these techniques, and highlight their characteristics in a comparative view. Furthermore, the authors address challenges of outlier detection in wireless sensor networks, provide a guideline on requirements that suitable outlier detection techniques for wireless sensor networks should meet, and will explain why general outlier detection techniques do not suffice.