Having the ability to interact with the physical environment and wireless communication capability, WSNs are a powerful platform for supporting many different applications and they have been employed in a wide range of real-world scenarios including scientific research, health care, environmental monitoring, public utilities, industry domain, security and surveillance, sustainable technologies, and military applications.
While WSNs are already shown to be capable of supporting and facilitating many applications, they also represent a diverse and challenging research area . In many applications, each WSN sensor node is required to support various sensors for different measurements.
For example, in a climate monitoring experiment, the sensor node may need to support temperature sensing, relative humidity sensing, barometric pressure sensing, ambient light sensing, etc. And in some applications, new sensors must be added to the existing system when additional measurement functions are required.
Nowadays, analogue sensors are still the most common kind in use because they are low cost, robust and reliable, and they can survive in harsh environments. In most cases the inherent characteristics of WSNs require sensor nodes to be low power, low cost, and small size.
However due to the variety of characteristics of analogue sensors and their non-standardised output signal, how to support or integrate different types of sensors within each sensor node of the WSN in a flexible and scalable way while not compromising power consumption, size and cost, becomes a significant challenge.
The traditional approach to support a number of sensors is to employ different types of signal conditioning circuits and ADCs. But this will significantly increase the complexity and power consumption as well as the size and the cost of the sensor nodes, and this type of system can lack flexibility and scalability when new or additional sensors are required.
In the analogue domain, the two major programmable hardware technologies are the field programmable analogue arrays (FPAA) and programmable mixed signal system on chip technology (e.g. PSoC).
Both technologies are capable of providing all the common components for supporting different analogue sensors, but compared to FPAA, PSoC possesses attributes which ideally suitable for WSN applications.
To address these issues in our reconfigurable adaptive wireless sensor (RAWS) program we chose to use PSoC technology for the following reasons:
i. Far lower power consumption than FPAA.
ii. A PSoC has a built-in CPU subsystem with SRAM, EEPROM, and flash memory in addition to configurable analogue blocks while FPAA requires an external microcontroller. This should significantly reduce the complexity of the sensor node as well as the overall cost and power consumption.
iii. PSoC contains programmable digital blocks that can be reconfigured into different digital and communication peripheral functions such as timers, PWMs, I2C, SPI, UART, etc. This enables better flexibility, scalability and applicability of the system.
The prototype uses the 8051-based Cypress PSoC CY8C29466. It contains not only a reconfigurable analogue block array, but also a CPU subsystem with memory, and a digital block array. This PSoC chip also has the necessary low power consumption. The typical supply current is 2mA in active mode and 3µA in sleep mode.
The ZigBee transceiver selected for the RAWS node prototype is the XBee ZB RF module. This module operates in 2.4GHz band and has raw data rate of 250Kbps. It has a low transmit power of 1.25mW and sleep mode current of 1.25µA, and the transmission range is 40m indoor and 120m in line- of-sight conditions. A ZigBee gateway is set up using the same XBee ZB module.
The gateway will be responsible for initialising and establishing a ZigBee network, collecting data from the RAWS node and then transmitting to the computer through the USB port. In the RAWS node side, the PSoC microcontroller interfaces with the ZigBee module via UART. The PSoC controls the ZigBee module by AT command to connect to the wireless network, and sends the sensor measurement reading to the network gateway.
The RAWS node prototype has been developed to realise and prove the concept of RAWS node technology. For the future, more sensors will be included and tested to further develop the RAWS node platform. New technology features will also be incorporated such as single cycle high level instruction execution (e.g. from C source code) and deep sleep modes for ultra-low-power operation and power consumption tests will be carried out.
To read this external content in full, download the complete paper from the open online archives at The Callan Institute at Waterford.