Wireless sensor networks (WSN) have the potential to become the third wireless revolution after wireless voice networks in the 80s and wireless data networks in the late 90s. This revolution will finally connect together the physical world of the human and the virtual world of the electronic devices. Though in the recent years large progress in power consumption reduction has been made in the wireless arena in order to increase the battery life, this is still not enough to achieve a wide adoption of this technology. Indeed, while nowadays consumers are used to charge batteries in laptops, mobile phones and other high-tech products, this operation becomes infeasible when scaled up to large industrial, enterprise or home networks composed of thousands of wireless nodes. Wireless sensor networks come as a new way to connect electronic equipments reducing, in this way, the costs associated with the installation and maintenance of large wired networks. To accomplish this task, it is necessary to reduce the energy consumption of the wireless node to a point where energy harvesting becomes feasible and the node energy autonomy exceeds the life time of the wireless node itself. This thesis focuses on the radio design, which is the backbone of any wireless node. A common approach to radio design for WSNs is to start from a very simple radio (like an RFID) adding more functionalities up to the point in which the power budget is reached. In this way, the robustness of the wireless link is traded off for power reducing the range of applications that can draw benefit form a WSN. In this thesis, we propose a novel approach to the radio design for WSNs. We started from a proven architecture like Bluetooth, and progressively we removed all the functionalities that are not required for WSNs. The robustness of the wireless link is guaranteed by using a fast frequency hopping spread spectrum technique while the power budget is achieved by optimizing the radio architecture and the frequency hopping synthesizer Two different radio architectures and a novel fast frequency hopping synthesizer are proposed that cover the large space of applications for WSNs. The two architectures make use of the peculiarities of each scenario and, together with a novel fast frequency hopping synthesizer, proved that spread spectrum techniques can be used also in severely power constrained scenarios like WSNs. This solution opens a new window toward a radio design, which ultimately trades off flexibility, rather than robustness, for power consumption. In this way, we broadened the range of applications for WSNs to areas in which security and reliability of the communication link are mandatory.
|Qualification||Doctor of Philosophy|
|Award date||20 Jan 2010|
|Place of Publication||Eindhoven|
|Publication status||Published - 2010|