Abstract

Intrusion detection is a surveillance problem of practical import that is well suited to wireless sensor networks. In this paper, we study the application of sensor networks to the intrusion detection problem and the related problems of classifying and tracking targets. Our approach is based on a dense, distributed, wireless network of multi-modal resource-poor sensors combined into loosely coherent sensor arrays that perform in situ detection, estimation, compression, and exfiltration. We ground our study in the context of a security scenario called “A Line in the Sand” and accordingly define the target, system, environment, and fault models. Based on the performance requirements of the scenario and the sensing, communication, energy, and computation ability of the sensor network, we explore the design space of sensors, signal processing algorithms, communications, networking, and middleware services. We introduce the influence field, which can be estimated from a network of binary sensors, as the basis for a novel classifier. A contribution of our work is that we do not assume a reliable network; on the contrary, we quantitatively analyze the effects of network unreliability on application performance. Our work includes multiple experimental deployments of over 90 sensor nodes at MacDill Air Force Base in Tampa, FL, as well as other field experiments of comparable scale. Based on these experiences, we identify a set of key lessons and articulate a few of the challenges facing extreme scaling to tens or hundreds of thousands of sensor nodes.

Keywords: Wireless sensor networks; Smart dust; Target classification and tracking; Reliability; Stabilization

Article Outline

1. Introduction

2. Related work

3. Problem formulation

3.1. User requirements and performance metrics

3.2. Target models

3.3. System model

3.4. Environment model

3.5. Fault model

4. Design considerations

4.1. Reliability

4.2. Energy

4.3. Complexity

5. Sensing

5.1. Phenomenology

5.2. Sensing options

5.3. Sensor selection

6. Signal detection and parameter estimation

6.1. Magnetic

6.2. Radar

7. Classification

7.1. The influence field as a spatial statistic

7.2. Classifier design

7.3. Validation

8. Tracking

9. Time synchronization

9.1. Algorithm

9.2. Accuracy

9.3. Results

10. Communications and networking

10.1. Routing algorithm

10.2. Reliable communications

10.3. Experimental results

11. Implementation

11.1. System architecture

11.2. Network nodes

11.3. Sensor boards

11.4. Packaging

11.5. Visualization

12. Key lessons

12.1. Impact of network unreliability

12.2. Testing at scale

12.3. Unanticipated faulty behavior

12.4. Quality of network reprogramming

13. Conclusions and future work

References

Vitae