The main considerations that drive the market within wireless sensor networks (WSNs) are cost, size and power consumption. Within research, the latter is considered to be an important factor as it essentially determines the size of a battery that should be used to power the sensor node and accordingly the overall size of the node itself. Additionally, it drastically influences the lifetime of the sensor nodes whose batteries cannot be recharged or even replaced when they are depleted as it would be cost-inefficient for large networks. Moreover, due to size constraints in some specific WSN applications, sensor nodes cannot use batteries at all but solely depend on energy scavenging from the surrounding environment (vibrations, heat, light). In order to make such applications feasible and to extend the node lifetime as much as possible (up to several years), ultra-low-power (ULP) design is highly desirable. Otherwise many WSN applications would still be prevented from widespread adoption.
Conventional approaches for ULP design in WSNs, such as duty cycling, wake-up receivers and ultra-wideband (UWB) technology, have their own limitations or weaknesses. For instance, UWB is limited to a short range of few meters due to emission regulations. On the other hand, wake-up receivers are prone to interference and duty-cycling needs strict timing/rendezvous which becomes extremely difficult to achieve in case of hundreds or thousands of nodes.
The Slow Wireless project is targeting ULP and low data rate (100 bit/s) WSN applications operating in the 2.4-GHz ISM band with much more focus on robustness to interference. The key concept behind the project is to utilize very narrowband (VNB) signaling (orders of kHz). This is based on the fact that the majority of WSN applications is characterized by low data rate as the traffic generated by WSN nodes is extremely low. Additionally, the dominant interference within the 2.4-GHz band is from WiFi signals which have a wideband nature. Therefore, by following continuous transmission (i.e. very relaxed duty cycle) instead of heavy duty-cycling, we are able to follow the VNB scheme, and accordingly, only a small portion of the interference signal will enter the VNB receiver. Another advantage is the relaxation in the transmitter power and/or the receiver noise figure.
AIM OF THE PROJECT
The aim of the Slow Wireless project is to investigate—through developing a prototype node—the feasibility of the VNB scheme as an ULP WSN technology that is invulnerable to the existing wideband interference in the 2.4-GHz ISM band. Three research groups from three neighboring but distinct areas are participating in the project: telecommunication engineering (TE), integrated circuit design (ICD), and computer architecture and embedded systems (CAES). The TE group will be investigating the physical layer aspects of the VNB scheme with respect to link behavior, fading, interference and frequency control. The formulated research questions for the group are
- carrier phase noise;
- frequency management amongst the sensor nodes;
- fading and interference modelling;
- optimal design parameters and adaptation.
The ICD and CAES groups are aiming at the investigation and low-power implementation in the analog (RF-frontend) and digital (baseband processing) domains, respectively, including related system architecture design trade-offs.
PARTICIPATING UT RESEARCH GROUPS
- Telecommunication Engineering Group (TE)
- Integrated Circuit Design Group (ICD)
- Computer Architecture and Embedded Systems (CAES)