Research Article: Jammer Localization Using Wireless Devices with Mitigation by Self-Configuration

Date Published: September 1, 2016

Publisher: Public Library of Science

Author(s): Qazi Mamoon Ashraf, Mohamed Hadi Habaebi, Md. Rafiqul Islam, Matthias Dehmer.


Communication abilities of a wireless network decrease significantly in the presence of a jammer. This paper presents a reactive technique, to detect and locate the position of a jammer using a distributed collection of wireless sensor devices. We employ the theory of autonomic computing as a framework to design the same. Upon detection of a jammer, the affected nodes self-configure their power consumption which stops unnecessary waste of battery resources. The scheme then proceeds to determine the approximate location of the jammer by analysing the location of active nodes as well as the affected nodes. This is done by employing a circular curve fitting algorithm. Results indicate a high degree of accuracy in localizing a jammer has been achieved.

Partial Text

Wireless networks, in general, and wireless sensor networks (WSNs), in particular, are increasing in popularity due to the recent rapid decline of costs and the constant availability of new technology. Much research is also being done in terms of designing green protocols to make wireless networks energy efficient. As a result, wireless networks are now being extensively deployed from home automation systems to industrial communication systems. As these networks gain in scalability and size, the dependence on human manual intervention has to be minimized [1]. The networks have to be self-sufficient in all concerns especially in the aspects of self-configuration, self-organization, self-awareness, self-recovery and self-protection [2].

In this section, we highlight our evaluation results on the performance of the proposed localization algorithm. The scheme was simulated on COOJA simulator in Contiki OS using Zolertia Z1 nodes. In our simulation test, the communication between the Zolertia Z1 nodes was executed on ContikiMAC protocol with default contention parameters that are provided in the simulation environment. The choice of the underlying MAC protocol and the associated performance plays an important role in reducing the number of collisions as well as inefficient message delivery. However, studying the effect of the underlying MAC protocol on timing and collision is out of scope for this research, as the results primarily focus on the accuracy of detection of the jammer using the proposed scheme. The Contiki OS uses ContikiMAC protocol as the default mechanism for effective power-saving (used for execution in self-configuration) in which the sensor nodes can make their radio listening ability to be turned off when not required. Although the ContikiMAC is a power-efficient protocol, it still requires accurate timing and contention between successive transmissions. The ContikiMAC protocol makes use of Clear Channel Assessment (CCA) which measures the Received Signal Strength Indicator (RSSI) of the radio transceiver to alert the availability of the radio channel. ContikiMAC allows decreasing the power consumption of the sensor nodes by switching off the radio when not necessary. A proper duty cycling algorithm becomes the key for power-efficiency, the functionality of which we exploit in the self-configuration part in the proposed scheme.

Assuming a simple circular areas for wireless coverage is an acceptable starting point, however, actual wireless communications never match such ideal cases due to many factors such as antennae and environmental factors. The sensing and transmission range, (Rt) of a node is much shorter than the communication and interference range (Ri). Since the localization scheme makes use of Rt, therefore, it can be argued that the scheme is also expected to provide an accurate localization for an irregular radio pattern such as illustrated in Fig 9. The ideal radio pattern for a jammer was presented in Fig 2. In our simulation study using the Contiki OS, we selected ‘Unit Disk Graph Medium (UDGM): Distance Loss’ model as the radio model. In this model, only the transmission distance of the nodes as well as the transmission success ratio can be configured.

This paper proposed a scheme to locate a jammer, and consequently to save energy of the affected nodes in the jammer affected areas. The jammer made use of the autonomic control loop, by monitoring and analyzing the locations of the affected nodes. The autonomic components of monitor, plan, analyze and execute were utilized for the design of the scheme. Constant monitoring of the device state through routine communication allows the nodes and the jammer to be mobile. Analysis of the communication pattern allows the central node to understand the presence of a jammer using a circular curve fitting geometric algorithm. We observed that 95% localization accuracy was achieved for a set of 20 nodes, when the transmission distance and actual distance of a jammer from the central node are similar. In the meantime, the affected nodes also monitor the area and execute an energy saving process upon failure of communication. Having planned the self-configuration scheme, the nodes can execute a lower duty cycle if the presence of a jammer has been established.