Abstract
Wireless sensor networks (WSN) are fast becoming extremely popular with usage in a variety of applications starting from mobile communication to home security. One of the main constraints faced during the design of the wireless radio sensors is power consumption. Most of the sensors are battery operated and many of them cannot operate without power for over a day. Transceiver is one of the main power guzzlers. Several techniques are used for reducing the transceiver power consumption in order to increase the battery life of the unit. Wake up radio receivers are offering a new way of operating the WSNs. Wake up radio receivers use power in the range of Nano watts. Also, by switching off the main WSN unit when it is not used, wake up radio receiver helps in prolonging the battery life substantially. In fact, further researches are taking place in this area for the development of energy sufficient wake up radio receivers that require no energy from outside sources for a prolonged period. A slew of challenges are there in implementing wake up radio receivers as doing so may introduce delay, latency, quick wear and tear due to frequent switching and several other relevant issues. However, research suggests that those can be easily overcome in the coming days with different design technique modifications in the WSN design. Wake up radio receivers are a big step towards designing energy neutral WSNs.
Introduction
The usage of wireless sensor networks is on the rise in recent years. WSNs are used increasingly in environmental monitoring, industrial domains, and healthcare and security systems. This is an emerging technology in use currently for examining the physical world to obtain information on a wide variety of applications. Wireless sensor networks can be seen as an assembly of many types of small electronic units (energy units, computational units and memory units) that are working together autonomously to accomplish a specific type of task (Umbdenstock et. al. 2013). WSNs are becoming so popular because they are easy to deploy and require no cables, wire or other infrastructure to start functioning. Wireless sensor networks are indeed self-supportive and can perform a host of operations such as processing, sensing, storage and communication from the time of installation without any extra hardware or software support (Magno et. al., 2014). One of the major issues with WSN is their energy consumption. Therefore, these units either require regular charging of the battery or require a cable that connects the system with a power source. This defeats the purpose of a wireless sensor network. There are a lot of researches being carried out in both hardware as well as software side of WSNs to reduce the power consumption of these units. Wake up radio receiver is one hardware unit that reduces the overall power consumption by switching on the main unit only when it requires to perform a task. This essay makes an analysis of wake-up radio receiver and how it can reduce the energy consumption of WSNs significantly.
Literature Review
WSNs are a combination of different types of circuits. Memory, computation, transceiver (receiver-transmitter) and power unit are the main components of a WSN. Almost all the WSNs are battery operated. Therefore, the amount of energy available for a WSN is constrained by the small device (battery) form factor (Huo, 2014). Maximum power is consumed in the transmitter and receiver unit (often called the transceiver unit).
Figure 1: Typical power consumption in a WSN unit (Umbdenstock et. al., 2013)
It can be seen from the above figure that transceiver is the most power hungry unit of the WSN. Currently, there are several techniques being used reduce the overall consumption of power (Huo, 2014).
Software Based Approach
Low duty cycle is theoretically the only way to reduce power in a given design of WSN. This low duty cycle can be implemented by a periodic wake up scheme using a software programming. Every node within the WSN switches between a wake up state (action state) and a sleep state (power efficient state) irrespective of whether the WSN unit is communicating, transmitting, receiving or computing (Umbdenstock et. al. 2013). This low duty software protocols are built on top of these periodic wake up cycles to enable communication even if a node is in sleep mode. The main problem with this approach is that very rarely the application requirement and the periodic wake up design scheme match, and in such cases, performance may significantly be compromised. Software based approach cannot address the issue of uncertainty associated with inbound communication (Umbdenstock et. al. 2013).
Hardware Based Approach
Figure 2: Longer delay, lower energy consumption but higher latency (Huo, 2014)
Apart from energy, latency is another parameter extremely important in the design of a WSN. In the case of a duty cycle, the energy characteristics of the WSN improve but the latency of the circuit suffers proportionately. It is a tradeoff between energy and latency and also this approach does not take into account the actual real life uncertainties of the incoming signal.
Energy Consumption and Wake-up Radio Receivers
In recent years, there has been a lot of research in the WSN circuit to minimize the energy consumption without compromising the latency through the input signal timing. Current synchronization process used in WSNs is periodic and real life signals are often asynchronous (Umbdenstock et. al. 2013). Wake-up transceiver provides a solution that reduces energy requirement without compromising on latency. Wake-up transceiver works purely based on-demand approach and thus improves the latency. Wake-up transceivers never waste energy in idle listening. These transceivers are an additional circuitry on the main WSN circuit that continuously monitors the incoming communication.
Types of Transceivers
There are several types Wake-up transceivers (WUT). The most researched WUTs are active transceivers that fulfill the energy requirement from the WSN battery circuit. Passive WUTs are transceivers that generate their own energy for operation from the incoming communication signals (Magno et. al., 2014). They do not require any external power. WUTs can also be categorized on the basis of input signal types. Some of the WUTs work on radio based frequencies and some of the WUTs work only on acoustic signals (Umbdenstock et. al. 2013). There are other types of transceivers that operate on the basis of wind, water and heat.
Operation of a WUT
Figure 3: Wakeup process in a WSN circuit (Magno et. al., 2014)
It can be seen in the above figure that the WUT unit acts as the asynchronous sampler unit in a WSN circuit. When it receives an incoming communication (through radio frequency or acoustics signals), it sends a signal to the clock and enables it, which in turn switches on the power hungry WSN main circuit. The clock also enables the data and communication to reach the proper modules in the synchronous blocks. In the present circuits, clocks are always turned on, which drains a lot of power. In a wake-up receiver circuit, even the clock is triggered only through a 16 bit encoder/decoder (wake-up unit). As an additional circuit is added to the WSN, in active mode, the energy consumption goes up by 50-70%. However, in idle mode, since only the wake-up circuit is on and the rest of the circuit is off, the power consumption can go down as low as 1000% (Magno et. al., 2014).
Design of a Passive WUT circuit
The design of WUT for active circuits is easier as it has no constraints on input power and draws power from the battery of the WSN circuit. Active WUTs are better for usage if the idle to active time of the WSN is 20:1. If active time is more than 5% of the total time of operation, then active WUT circuits end up consuming more power than the current architecture (Huo, 2014). That is why current research is focused on inventing passive WUT circuits, which are energy neutral. If that can be achieved then the WSN energy usage will see a huge improvement even if the idle time is as low as 20%.
Figure 4: A Typical Passive Wake-up Receiver Design (Huo, 2014)
The above circuit is the fundamental circuit of a passive wake-up receiver. Current circuits under research are more complex and sophisticated but the basic building blocks of the passive wake-up receiver circuit are same as the above figure. In this passive wake-up receiver, the antenna receives the signal (EM wave) and generates an input voltage, which is then stepped up by the transformer and fed into the main wake-up receiver circuit. Diode is used to create an output voltage, which is fed into the clock circuit of WSN to bring it back to active mode from sleep mode. The signal, which is received via the antenna, helps build a voltage across the capacitor. The basic equations for designing the elements of the circuit are
Figure 5: Governing equations of the passive wake-up receivers (Huo, 2014)
E is the total amount of energy collected by the circuit, Pr is the received power by the wake-up receiver, t is the time required to accumulate energy and ef is the efficiency of the energy harvesting. In the second equation, Cse is the rating of the capacitor. These circuits are currently capable of achieving a sensitivity of -29.3 dBm and can achieve an output voltage of 0.7 V at 868 Mhz.
Advanced Designs
The simple passive wake-up receiver design is robust but may require a big length of antenna or lacks range. If the sensitivity of the circuit is increased, then it tends to pick up noise and cannot separate out input signals from noise and may switch on the circuit even if there is no actual input communication, resulting in more energy consumption (Huo, 2014). To make the circuit more robust and energy efficient, recent researches are using CMOS to design the wake-up receivers (Huo, 2014).
Figure 6: Multi pass and Multiband Wake-up Receiver (Magno et. al., 2014)
Multi-band and multi pass filters with CMOS sensor bases designs give superior switching and signal detection, but power consumption of those circuits is still high and future researches are required to improve the circuit designs (Magno et. al., 2014).
Challenges
The main problem in the wake-up filter remains the lack of focus in the research community on this issue. In fact, WSNs are in use for decades now and during this period, the primary focus was given on reliability, cost, quality and size of the WSN circuits (Huo, 2014). Relatively less importance was given to power. However, only the last few years have seen some real inertia in this type of research. Industrial WSN units are still heavily focused on performance and reliability, but mainly health monitoring and medical procedures are the main area creating requirement for the ultra-low power wake-up radio receivers.
Also, simple wake-up filters are unable to produce good and consistent results. In many industrial setups, there are many types of frequencies that may be picked up by a simple wake-up filter (Magno et. al., 2014). Even in almost no cases, it can send an acknowledge signal as simple WUTs have no information about the sending system. More complex designs can identify the source signals more reliably and can send back acknowledge signal but more complex designs consume more power and add more latency to the circuit (Huo, 2014). More future work in the area is required before ultra-low power wake-up radio receivers make their way into industrial WSN circuits.
Conclusion
WSNs will replace many wired devices in the coming days because of their ease of installation and use. In fact, there is no alternative for WSNs in many industrial applications and security systems. However, the main problem with WSNs is that they are battery operated. Like any other battery operated device, WSNs also need to be charged periodically. Often regular charging of the WSN systems in the industrial and other operational setup is very time consuming and cumbersome. Major research focus is dedicated to the area of WSNs energy consumption. Transceiver is the unit that consumes most energy in the WSN circuit. Current software and hardware methods use periodic on-off of the main circuit. This reduces the overall energy consumption of the circuit but introduces latency in response. Also, as these are predetermined on-off periods, they often are unable to match the real life input signals. To overcome those issues, a new generation of research is taking place in the area of wake-up radio receivers that are able to switch the WSNs circuits with asynchronous switching based on input signal. This is helping in the reduction of power drastically as the circuit is running only on demand basis. Although the extra circuitry and extra power requirement are not ideal in the active mode, but in sleep mode, the power requirement can drastically go down. In fact, the future research is concentrating on building power neutral wake-up radio receivers. In fact, many researchers are also trying to build a circuit that can supply a small amount of power through the wake-up circuit to the WSNs.
References
Magno, M., Marinkovic, S., Srbinovski, B. and Popovici, E.M. (2014). Wake-up radio receiver based power minimization techniques for wireless sensor networks: Microelectronics Journal. 45(12), 1627-1633, doi:10.1016/j.mejo.2014.08.010
Hutu, F., Khoumeri, A., Villemaud, G. and Gorce, J. (2014). A new wake-up radio architecture for wireless sensor networks. EURASIP Journal on Wireless Communications and Networking. Retrieved on 20 April 2015, from <http://jwcn.eurasipjournals.com/content/2014/1/177>
Pletcher, N. and Rabaey, J.M. (2008). Ultra-Low Power Wake-Up Receivers for Wireless Sensor Networks. Electrical Engineering and Computer Sciences, University of California at Berkeley. Retrieved on 20 April 2015, from <http://www.eecs.berkeley.edu/Pubs/TechRpts/2008/EECS-2008-59.pdf>
Huo, L. (2014). A Comprehensive Study of Passive Wake-up Radio in Wireless Sensor Networks. Delft University of Technology. Retrieved on 20 April 2015, from <http://www.es.ewi.tudelft.nl/msc-theses/2014-Huo.pdf>
Umbdenstock, E., Schafer, F., Kleinsteuber, M. and Meyer, Hagen. (2013). Wake-Up-Receiver in energy efficient Wireless Sensor Networks for security applications. Technische Universität München. Retrieved on 20 April 2015, from <http://www.agence-nationale-recherche.fr/Colloques/WISG2013/articles/Article_Umbdenstock.pdf>