Interview

Abstract

Cristian Muller from the Federal University of Pampa in Brazil talks to us about the work behind the paper ‘An Improved Solution for Node Location Multilateration Algorithms in Wireless Sensor Network’, page 1179 Cristian Muller My interest in this field came from working within the Microelectronics’ Group as an undergraduate student. The initial focus of the work was not on multilateration itself, but on a framework to simulate wireless sensor network; mainly localisation algorithms. I was studying the implementation of some classical location algorithms and was trying to improve them; looking for ways to reduce complexity without significantly compromising the location precision. Initially it was only a promising idea, but it gave me the motivation to start studying multilateration algorithms and its mathematical background to better understand the properties of our improved algorithm and to compare it with the other algorithms. A wireless sensor network consists of a network formed of many tiny sensor nodes deployed on a specific site at performing tasks like monitoring. These tiny nodes have sensing, processing, and communicating capabilities, and they are powered generally by batteries. Therefore, wireless sensor networks rely on technological advances in sensors, semiconductor devices, networking protocols and energy storage/generation. There is a wide range of possible applications, such as health, military, industrial, home, environment and traffic monitoring. As a simple example, a wireless sensor network could be deployed in a dense forest to detect smoke that could be an indication of fire. The first reason is that in many applications the node's location information is a part of the task performed by the wireless sensor network. An example of this case is environment monitoring, where we want to know where the sensed data came from. Location information is important as it can be used to improve security and to optimise the routing protocols of a network, which may result in energy saving. In the real world, the presence of significant range error is expected. In such scenarios, the proposed solution has lower computational complexity and better convergence performance than the multilateration method based on the Gauss-Newton, which is commonly used for this purpose. The solution proposed in our Letter can be useful for multilateration tasks for node location in real applications based on wireless sensor networks that have hardware and/or power constraints, especially when there are batteries that can neither be replaced nor recharged. Thus, our solution contributes to reduced power consumption, extending the nodes lifetime, while maintaining a good location precision. As localisation is of paramount importance for many applications, I believe this technique could be inserted as a location engine, either in software or hardware, in nodes in sensor networks. Since it is relatively simple, it can be used in applications with hardware with power constraints, like those envisioned by the emerging field of Internet of Things. There have been significant advances in the last few years in robotics, such as drones, for instance. These autonomous devices can aid localisation in networks where nodes move with actuators or without guidance. There is great potential for localisation schemes regarding indoor applications using RFID devices. Power supply: It seems obvious but the bottleneck of many applications, including those versioned by the IoT (Internet of Things) field, may be overcome by advances in energy harvesting, eliminating traditional batteries.