EyeSec: A Retrofittable Augmented Reality Tool for Troubleshooting Wireless Sensor Networks in the Field

Similar to enterprise-class IT devices, the WSN is usually monitored by capturing network traffic at the gateway and displaying node and link status to the network operator at a central control terminal [12,29]. However, after a device malfunction has been detected, situations are different. While enterprise-class IT devices can be managed actively using e.g. Simple Network Management Protocol (SNMP), sensor node behavior can only be observed passively. Sensor nodes typically lack network management protocols, because flash memory is scarce, firmware must be sleek and additional network traffic, which could interfere with WSN operations, must be avoided. Hence, if sensor node failure is reported by the c Martin Striegel et al. This paper has been published in the Proceedings of the 2019 International Conference on Embedded Wireless Systems and Networks, EWSN 2019, Beijing, China, February 25-27. 2019 central monitoring system, an operator is sent into the field to pinpoint and fix the problem on-site. To be able to provide the operator with information on the network at the location of deployment, Turon et al. and Bokde et al. describe hand-held devices [30,11]. Those devices obtain information from the gateway. This is disadvantageous, as the hand-held devices require a permanent connection to the gateway and their operativeness depends on those of the gateway. To become independent of the gateway as network traffic source, separate and passive capture devices have been proposed [23,22,32]. Consisting of mobile capture or sniffer nodes, they can be deployed temporarily to overhear all WSN traffic.

Combining mobile capture networks and hand-held devices for visualization permits the operator to work independently of existing infrastructure. However, by just shifting the visualization of network data from a central terminal to a hand-held device does not tackle specific problems encountered by the operator. While a complete view on the WSN such as text-based traffic statistics and large network graphs can be displayed at a hand-held device, the operator needs only limited information targeted at solving a specific problem. Instead, the operator needs to map digital device representation, i.e. network addresses, and the visual device representation of the physical sensor node in front of him. This permits to identify the physical device causing problems in the network. In turn, while manipulating a physical device, the operator needs to see the effects on the digital world, e.g. if a network connection can be restored.

Currently, there exists no tool designed to provide the operator in the field with just the information needed to debug wireless sensor networks. To fill this gap, we propose EyeSec, a tool utilizing Augmented Reality (AR) and tailored towards the operator’s needs. EyeSec includes Augmented Reality Devices (AR Devices), which detect and identifiy physical sensor nodes using Quick Response (QR) code markers. Portable and extensible Sniffer Units capture network traffic and extract digital information. Data from the visual and digital domain are merged and stored at a portable Backend. An AR Device obtains consolidated information from the Backend and superimposes physical sensor nodes with this information. Data flows and connectivity between sensor nodes are visualized. Unlike centralized network visualization solutions, which display the full network, we exploit the operator’s physical proximity to sensor nodes to limit displayed information to exactly those he is interested in. EyeSec is designed such that Sniffer Units and Backend can be installed ad-hoc at a WSN deployment site and removed after troubleshooting is finished, neither requiring changes to sensor node firmware nor interfering with WSN operations at any time. Our main contributions are:

• The first completely mobile WSN monitoring system, which permits the operator to work independently of any existing infrastructure, called EyeSec.

• Passive operation, i.e. no modifications to sensor node or gateway firmware need to be made. This permits EyeSec to be retrofitted to already deployed WSNs without introducing additional sources of error.

• Modular and protocol-agnostic design, so extra transmission protocols can be added easily. EyeSec utilizes off-the-shelf radio transceivers and Android OS smart phones, which are available widely and cheaply. The remainder of this paper is structured as follows: In Section 2, we detail requirements and design considerations for building a retrofittable and protocol-agnostic AR system. In Section 3, we show how the design has been implemented in hardware and software. In S

…(Full text truncated)…