Energy harvesting Wheel Speed Sensor 
Master of Science Thesis 

 

DHASARATHY PARTHASARATHY 

Department of Microtechnology and Nanoscience 

CHALMERS UNIVERSITY OF TECHNOLOGY 

Göteborg, Sweden, 2012 

  



1 

 

  



2 

 

 

 

 

 

 

 

 

 

Energy harvesting Wheel Speed Sensor 

DHASARATHY PARTHASARTHY 

 

 

 

 

 

Department of Microtechnology and Nanoscience 

CHALMERS UNIVERSITY OF TECHNOLOGY 

Göteborg, Sweden 2012 

 
 



3 

 

Energy harvesting Wheel Speed Sensor 

DHASARATHY PARTHASARATHY 

 

© DHASARATHY PARTHASARATHY, 2012 

 

Department of Microtechnology and Nanoscience, 

Chalmers University of Technology, 

SE-412 96 Göteborg 

Sweden 

Telephone +46 (0)31-772 1000 
  



4 

 

Energy harvesting Wheel Speed Sensor 

DHASARATHY PARTHASARATHY 

Department of Microtechnology and Nanoscience, 

Chalmers University of Technology 

 

 

Abstract: 
 

This thesis presents a prototype energy harvesting autonomous sensor, called the Autonomous Wheel 

Speed Sensor (AWSS), that is targeted for operation in the Electronic Braking System (EBS) of 

vehicles. In order to monitor the rotational state of a wheel, the EBS currently uses a passive Wheel 

Speed Sensor (WSS) which is a variable reluctance electromagnetic transducer. In the existing EBS 

setup, one WSS is used per wheel, each of which is connected to the EBS system using extensive 

cabling. 

This project presents the first successful attempt at converting the WSS into an energy harvesting 

wireless wheel speed sensor or in other words an Autonomous Wheel Speed Sensor (AWSS), which 

provides information about the rotational state of the wheel to the EBS system over a wireless link. 

Unlike most wireless sensors which use batteries, the AWSS employs energy harvesting to power 

itself by simultaneously using the WSS electromagnetic transducer as an energy harvester as well as a 

sensor. By thus making an autonomous WSS, the amount of cables in the existing WSS assembly can 

be reduced, leading to savings in material, assembly and maintenance costs.  

The self-powered prototype AWSS successfully implements periodic wireless transmission of the 

wheel speed along with near real-time wheel lock detection, at a duty cycle of less than  The 

AWSS has been built using readily available COTS components and uses a proprietary low-power 

standard for wireless communication. 

The prototype AWSS implemented in this project successfully demonstrates that the WSS is capable 

of being used as an energy harvesting transducer. In the experimental setup used in this project, the 

WSS yields harvestable power of  for speed ranges of . This opens 

up the possibility of using the variable reluctance sensor setup to harvest energy from any rotational 

assembly and use this harvested energy to power autonomous sensors. 

This prototype system is intended for operation in AB Volvo vehicle applications and the project is a 

partnership between Chalmers University of Technology and Volvo Group Trucks Technology – 

Advanced Technology and Research, Göteborg, Sweden. 

 

Keywords: energy harvesting sensor, wheel speed sensor, variable reluctance sensor 

  



5 

 

Acknowledgements 
 

It is not often that one is given five months of time, an Aladdin’s cave of a lab, the indulgence and 

guidance to build something as fun as an autonomous sensor right up from scratch. Sure enough, it’s 

been the most amazing five month period of learning, and now that the project is done, I turn my 

attention to the most important part: a shout-out of 'tack så mycket!' to everybody involved with the 

project.  

I begin with thanking my thesis examiner Per Lundgren at Chalmers. Thanks Per, for the opportunity, 

the advice and the tough reviews which made sure that my attention to detail wasn't wavering.  

I would then like to thank Alejandro Cortes at Volvo Technology (VTEC) for giving me this 

incredible opportunity to work with the state of the art of sensor technology. 

I then thank Roy Johansson at VTEC for his invaluable guidance. Thanks Roy, first for thinking up 

this project, then for all the early morning chats at 0730 where we always put my design through 

scrutiny and most important of all, for helping me (an automotive noob) by answering my never 

ending questions about trucks, though I think you've to work more on your fake Aussie accent! 

I save my biggest thanks for my supervisor Mikaela Öhman at VTEC. I couldn't thank you enough, 

Mikaela, for giving me this opportunity along a simple mandate to knock myself out! Thanks for being 

a great mentor and for your confidence in letting me freely frame the contents of my thesis. 

  



6 

 

Abbreviations: 

ABS Anti-lock Braking System 

ALL Application Logic Layer 

AP Access Point (of a sensor network) 

API Application Programming Interface 

ASR Anti-slip Regulation 

AWSS Autonomous Wheel Speed Sensor 

AWSSN Autonomous Wheel Speed Sensor Network 

DAQ Data Acquisition 

DSP Digital Signal Processing/Processor 

EBS Electronic Braking System 

ECU Electronic Control Unit 

ED End Device (of a sensor network) 

EHU Energy Harvesting Unit 

EMC Electromagnetic Compatibility 

ez430 TI ez430-RF2500 development tool 

FSM Finite State Machine 

GSM Global System for Mobile communications 

HAL Hardware Abstraction Layer 

IDE Integrated Development Environment 

ISR Interrupt Service Routine 

JTAG Joint Test Action Group 

LAN Local Area Network 

LPM Low Power Mode (MSP430 operational mode) 

LPM Low Power Mode 

MCU Microcontroller Unit 

MPPT Maximum Power Point Tracking 

PAN Personal Area Network 

PHY Physical layer of a communication protocol stack 

PM Pseudo Modulator 

PP Peak to Peak 

RMS Root Mean Square 

RPM Revolutions Per Minute 

RT Random Telegraph process 

RTOS Real-Time Operating System 

SoC System on Chip 

SPI Serial Peripheral Interface bus 

TDMA Time Division Multiple Access 

TI Texas Instruments Inc. 

TPMS Tire Pressure Monitoring System 

UART Universal Asynchronous Receiver/Transmitter 

WER Wheel Emulation Rig 

WISCON Wireless Sensor Concept Node 

VLSI Very Large Scale Integration 

VR Variable Reluctance 

WSN Wireless Sensor Network 

WSS Wheel Speed Sensor 

WSU Wireless Sensing Unit 

 

  



7 

 

Contents 

Abstract: ..................................................................................................................................... 4 

Acknowledgements .................................................................................................................... 5 

Abbreviations: ............................................................................................................................ 6 

Contents ..................................................................................................................................... 7 

List of Figures: .......................................................................................................................... 9 

List of Tables: .......................................................................................................................... 11 

Chapter 1 - Introduction ......................................................................................................... 12 

1.1 Summary of key contributions of this project: ............................................................................. 12 

1.2 Background of this project: ......................................................................................................... 13 

1.3 A guide to the contents of this report: ......................................................................................... 13 

Chapter 2 - Autonomous sensors in vehicles ......................................................................... 14 

2.1 Autonomous sensors: ................................................................................................................... 14 

2.2 The Wireless Sensor Network (WSN) paradigm: ........................................................................ 14 

2.3 Energy autonomy using Energy Harvesting:............................................................................... 15 

2.4 Autonomous Wireless Sensor Networks (WSN) in vehicles: ....................................................... 16 

Chapter 3 - The Wheel Speed Sensor (WSS) and its applications ......................................... 18 

3.1 The Wheel Speed Sensor (WSS): ................................................................................................. 18 

3.2 The WSS as an energy source: .................................................................................................... 21 

3.3 Primary application of the WSS - The Electronic Braking System (EBS): .................................. 23 

3.4 The role of the WSS in the EBS: .................................................................................................. 23 

3.5 WSS installation as part of the EBS: ........................................................................................... 24 

3.6 Other Variable Reluctance (VR) sensor applications in a vehicle: ............................................. 25 

Chapter 4 - The Autonomous Wheel Speed Sensor (AWSS) ................................................. 27 

4.1 The functional architecture of an AWSS node: ........................................................................... 27 

4.2 The Autonomous Wheel Speed Sensor Network (AWSSN): ......................................................... 28 

4.3 Challenges in implementing the AWSSN: .................................................................................... 28 

4.4 Potential benefits: ........................................................................................................................ 30 

Chapter 5 - Prototype specification ......................................................................................... 32 

5.1 Prototype network architecture:.................................................................................................. 32 

5.2 Prototype functionality: ............................................................................................................... 33 

5.3 Parametric model: ....................................................................................................................... 33 

5.4 General design principles for increasing availability of autonomous sensors: .......................... 44 



8 

 

Chapter 6 - Prototype development setup ............................................................................... 47 

6.1 The Wheel Emulation Rig (WER): ............................................................................................... 47 

6.2 Hardware choices for the ED and AP:........................................................................................ 48 

6.3 Putting it all together - the hardware development bench: ......................................................... 51 

6.4 Software development: ................................................................................................................ 52 

6.5 System modeling: ......................................................................................................................... 52 

Chapter 7 - Prototype implementation .................................................................................... 53 

7.1 Modular stack definition: ............................................................................................................ 53 

7.2 Hardware Layer: ......................................................................................................................... 53 

7.3 The software layers: .................................................................................................................... 55 

7.4 The AWSSN cluster library: ........................................................................................................ 59 

Chapter 8 - Prototype characterization ................................................................................... 60 

8.1 Characterizing the Wheel Emulation Rig (WER): ....................................................................... 60 

8.2 Characterizing the ED/AWSS node: ............................................................................................ 66 

8.3 Illustrative end-to-end application scenario: .............................................................................. 79 

Chapter 9 – Conclusions ......................................................................................................... 81 

9.1 Summary of results: ..................................................................................................................... 81 

9.2 Suggestions for further development ........................................................................................... 82 

Bibliography ............................................................................................................................ 84 

Appendix A - The Random Telegraph (RT) process .............................................................. 88 

Appendix B - Hardware Abstraction Layer (HAL) user/programmer guide ........................ 90 

Appendix C - Application Logic Layer (ALL) programmer/user guide ................................ 92 

Appendix D - AWSSN Cluster definition ............................................................................... 94 

Appendix E - Parametric register ........................................................................................... 95 

Appendix F – Prototype characterization test cases ............................................................... 98 

Appendix G – Confidential information ............................................................................... 101 

  



9 

 

List of Figures: 
Figure 1 : A WSN mesh network [9] ..................................................................................................................... 15 
Figure 2 : Architecture of an energy harvesting wireless sensor .......................................................................... 16 
Figure 3: An illustration of WSS assembly near the wheel of a truck [15] ........................................................... 18 
Figure 4 : The WABCO WSS and a depiction of its internal structure ................................................................. 19 
Figure 5 : The WSS along with a pole wheel ........................................................................................................ 19 
Figure 6 : Power and efficiency of an electrical circuit as a function of the load [16] ........................................ 21 
Figure 7 : Knorr-Bremse EBS used in some Volvo trucks [27] ............................................................................ 23 
Figure 8 : WSS in the EBS setup [27] ................................................................................................................... 24 
Figure 9 : WSS mounting on the front axle of Volvo D68 - with deflector and protection shield [15] ................. 25 
Figure 10 : WSS harness mounting procedure for the front axle [29] .................................................................. 25 
Figure 11 : A camshaft (left) and a crankshaft (right) .......................................................................................... 26 
Figure 12 : The Autonomous Wheel Speed Sensor (AWSS) .................................................................................. 27 
Figure 13 : The Autonomous Wheel Speed Sensor Network (AWSSN) ................................................................. 28 
Figure 14: Conceptual depiction of the targeted prototype - a two node AWSSN ................................................ 32 
Figure 15 : Targeted prototype AWSSN setup ...................................................................................................... 32 
Figure 16 : The 'leaky bucket' model of the energy harvesting ED node .............................................................. 36 
Figure 17 : Classification of WSU functionality ................................................................................................... 39 
Figure 18 : Design principles for increasing availability of autonomous sensing ............................................... 46 
Figure 19 : Experimental setup for the WSS - (a) the WER and (b) & (c) a closer look at the pole wheel and VR 

sensor setup ........................................................................................................................................................... 47 
Figure 20 : CBC-Eval-09 functional description due to [44] ............................................................................... 48 
Figure 21 : SimpliciTI communication stack [49] ................................................................................................ 50 
Figure 22 : Topologies supported by SimpliciTI [49] ........................................................................................... 50 
Figure 23 :  Prototype hardware development and test bench ............................................................................. 51 
Figure 24 : Modular definition of an AWSS node ................................................................................................. 53 
Figure 25 : Hardware Layer of the ED node ........................................................................................................ 54 
Figure 26 : AWSSN program structure ................................................................................................................. 55 
Figure 27 : State machine implementation of ED ALL ......................................................................................... 57 
Figure 28 : Flow diagram depicting AP ALL logic .............................................................................................. 59 
Figure 29 : WER transfer characteristics - signal voltage ................................................................................... 61 
Figure 30 : WER transfer characteristics - signal frequency ............................................................................... 62 
Figure 31 :  WSS COMSOL 2D model .................................................................................................................. 64 
Figure 32 : Simulated result of the magnetic flux density ..................................................................................... 65 
Figure 33 : WSS COMSOL model simulated results simulation results ............................................................... 66 
Figure 34: WSS signal distortion due to loading .................................................................................................. 66 
Figure 35 :  WSS signal pulse conversion – high frequency case ......................................................................... 67 
Figure 36 : WSS signal pulse conversion - low frequency case ............................................................................ 67 
Figure 37 : WSS unlock event - high speed ........................................................................................................... 68 
Figure 38 : WSS lock event - high speed braking ................................................................................................. 69 
Figure 39 : WSS lock event - low speed braking ................................................................................................... 69 
Figure 40 : ED startup with average current draw of  and discharge capacity of  .............. 70 
Figure 41 : ED event sensing and transmission .................................................................................................... 71 
Figure 42 : Wheel unlock event transmission with average current draw of  and a discharge capacity 

of  .......................................................................................................................................................... 71 
Figure 43: ED attribute sensing and transmission ............................................................................................... 72 
Figure 44 : Attribute sensing and transmission in a high speed case with an average current of  and a 

discharge capacity of  .......................................................................................................................... 73 
Figure 45 : Attribute sensing and transmission in a low speed case with an average current of  and a 

discharge capacity of  .......................................................................................................................... 73 
Figure 46 : Variation of with  for  .............................................................. 74 
Figure 47 : Plot of duty cycle as a function of attribute transmission interval with  ........ 75 



10 

 

Figure 48: Effect of event transitions on the duty cycle of operations .................................................................. 76 
Figure 49 : Average discharge due to the ED as a function of attribute transmission interval ............................ 76 
Figure 50 : Variation of average power consumption with attribute transmission interval ................................. 77 
Figure 51 : ED cold start example ........................................................................................................................ 78 
Figure 52 : AWSSN end-to-end illustration .......................................................................................................... 80 
Figure 53 : RT process for  ................................................................................................................. 88 
Figure 54 : RT process for  ................................................................................................................. 89 

 

  



11 

 

List of Tables: 
Table 1 : Wireless networks for automotive applications due to [13] .................................................................. 17 
Table 2 : WSS functional parameters .................................................................................................................... 34 
Table 3: EHU parametric specification ................................................................................................................ 37 
Table 4: WSS pulse converter truth table .............................................................................................................. 40 
Table 5 : WSU wheel speed attribute sensing specification .................................................................................. 40 
Table 6 : WSS rotational state sensor truth table ................................................................................................. 41 
Table 7 : WSU event sensing specification ........................................................................................................... 41 
Table 8 : WSU RF communication specification .................................................................................................. 41 
Table 9 : WSU duty cycle specification ................................................................................................................. 42 
Table 10 : ED power consumption parameters ..................................................................................................... 43 
Table 11 : EHU parameters as implemented by the CBC-Eval-09 ....................................................................... 49 
Table 12 :  WSU wheel speed attribute sensing parameters as implemented by ez430 (MSP430F2274) ............. 49 
Table 13 : ED parameters implemented by the Hardware Layer ......................................................................... 54 
Table 14 : ED parameters implemented by the HAL ............................................................................................ 56 
Table 15 : ED parameters implemented by the ALL ............................................................................................. 58 
Table 16 : WSS parameters as implemented by the WER ..................................................................................... 63 
Table 17 : WSS COMSOL model parameters ....................................................................................................... 65 
Table 18 : Experimental measurement of Hardware Layer parameters ............................................................... 70 
Table 19 : Experimental measurement of event sensing parameters .................................................................... 72 
Table 20 : Experimental measurement of attribute sensing parameters ............................................................... 74 

  



12 

 

Chapter 1 - Introduction 
 

"In the 1980s, the PC revolution put computing at our fingertips. 

In the 1990s, the Internet revolution connected us to an information web that spans the planet. 

And now the next revolution is connecting the Internet back to the physical world we live in-in effect, 

giving that world its first electronic nervous system. 

Call it the Sensor Revolution: an outpouring of devices that monitor our surroundings in ways we 

could barely imagine a few years ago. Some of it is already here. The rest is coming soon." 

- US National Science Foundation 'The Sensor Revolution - A special report' [1]. 

Sensors and sensor networks are heralding an age of astounding technological possibilities, glimpses 

of which are already being seen. Machines equipped with sensors that are completely in tune with the 

ambient environment are fast breaking out of the realm of science fiction and becoming commonplace. 

This race for a technological fantasy world depends upon the mastering of a wide range of 

technologies, among which ranks the technology of Energy harvesting. This technology, which 

imparts intelligent embedded sensor devices the ability to scavenge energy from the ambient 

environment, gives a special meaning to the term ‘ambient intelligence’. Energy harvesting, along 

with other key enabling technologies such as microelectronics, nanotechnology, signal processing and 

wireless communications, have already set the pace for rapid developments in this ongoing sensor 

revolution. The automotive electronics scene has also been an active participant in the sensor 

revolution, where new sensing technologies are continuously redefining the ways and means of 

controlling transportation. Energy harvesting sensors have now gained traction in the automotive 

industry as it has been identified as a technology that has real potential to broaden horizons. 

This chapter presents a summary of conclusions and contributions of this thesis project in section 1.1 

and a guide to the contents of the report in section 1.2. 

1.1 Summary of key contributions of this project: 

The first contribution of this project is proving the concept that the Wheel Speed Sensor (WSS) can be 

used as an energy harvesting transducer to power autonomous sensors. In the experimental setup used 

in this project, the WSS has been measured to produce harvestable power in the range of 

 for a wheel speed range of . These power levels have been shown to sustain 

carefully engineered wireless sensor operations. 

The next contribution is the development of a prototype Autonomous WSS (AWSS) which uses the 

power harvested from the WSS, in addition to simultaneously using the WSS as a sensor. The 

prototype AWSS is tested to be capable of comfortably transmitting wheel speed information at 

intervals of  while consuming less than  (under certain modeling assumptions) of 

power on average with a duty cycle of less than  for a wheel that keeps rotating for  of the 

time. It is also capable of communicating wheel start/stop events with a worst case delay of , a 

time interval which can be easily reduced in future iterations.  

Based on the design of the prototype AWSS, a set of general design principles have been derived, 

which would help in engineering energy harvesting sensors with real-time capabilities. These design 



13 

 

principles, which aim at increasing the availability of the autonomous sensor using highly abstracted 

passive stimulation, is the next contribution of this project.  

This project also recommends the serious consideration of the Variable Reluctance sensing assembly 

as a platform for harvesting energy from any rotating assembly. This harvested energy can be used to 

power sensors nearby. For example, by adapting the pole-wheel and VR sensor combination to fit 

inside the tire, enough energy can be harvested to power the Tire Pressure Monitoring System 

(TPMS).  

The entire project was conducted with an eye on the future and the prototype was designed and 

developed in such a way that it lays down the framework for further development. A basic 

mathematical model of the autonomous sensor has been developed, with an emphasis on incorporating 

stochastic models. A modular stack definition for operations in each sensor node has also been 

developed along with software platforms for each layer of the stack. 

1.2 Background of this project: 

This thesis project is a spinoff of the Wireless Sensor Concept Node (WISCON) project that is jointly 

being undertaken by Volvo Technology Corporation, Chalmers University of Technology and 

Halmstad University. The aim of WISCON is the survey of technology and development of 

prototypes, in order to build a knowledge base for energy harvesting in automotive applications. The 

project is in its second year of activity and this author, as part of WISCON, has participated in 

compilation of surveys of energy harvesting technologies and applications [2], and energy harvesting 

power electronics [3]. These technology surveys form the technical basis of this thesis project. 

1.3 A guide to the contents of this report: 

The idea of autonomous sensors and their potential application in automotive control systems is 

presented in Chapter 2. The Wheel Speed Sensor (WSS) which is targeted for rendering autonomy, its 

capabilities as a sensor and as an energy harvester and its application in the Electronic Braking System 

(EBS) of a truck among others, are presented in Chapter 3. The idea of an Autonomous Wheel Speed 

Sensor (AWSS), which combines the usage of the WSS as a signal and energy source, its 

implementation challenges and benefits are presented in Chapter 4. A minimal 

behavioral/mathematical model of the AWSS system that is geared towards the prototype 

implementation is presented in Chapter 5 along with system level functionality implemented in the 

prototype. The lab development setup is presented in Chapter 6 which describes the hardware and 

wireless communication standard chosen for implementing the prototype. The modular stack 

definition of an AWSS node along with a description of the implementation of each layer of the stack 

is presented in Chapter 7. The characterization of the prototype system and the results of 

characterization are presented in Chapter 8. Finally, a summary of results and suggestions for 

enhancements is presented are Chapter 9.  



14 

 

Chapter 2 - Autonomous sensors in vehicles 
 

Sensors are pervasive in today’s vehicles providing data for performance, safety, and convenience and 

comfort functions. The data provided by these sensors are made available to one of the many 

electronic control units (ECU) in the vehicle, which process this data and perform control actions. The 

sensors used in vehicles are of incredible variety ranging from the Oxygen sensor for regulating 

particulate matter in the exhaust, the Manifold Air Pressure (MAP) sensor to measure the mass air 

flow rate in the engine to decide fuel injection, right up to climate sensors for regulating the climate in 

the cab. Both the number and the variety of sensors used in a vehicle have been continuously 

increasing, a trend which is set to continue in the future. The main drivers behind such ever increasing 

sensor deployments in vehicles are increased operational efficiency on one hand and 

regulation/legislation on the other hand which mandate the deployment of sensors. The usage of 

sensors in vehicles is expected to expand over the years with the increase in ‘X-by-wire’, the 

electrical/electronic control of mechanical systems. A combination these and many more factors have 

made the automotive sensors market, according to one market research report [4], a $14.5billion 

industry in 2011 and projected to increase to $20billion in 2016. The average number of sensors in a 

vehicle, according to a review paper submitted in 2008 [5], is 40 in the North American market which 

is projected to increase to 70 in 2013. 

This chapter presents the idea of autonomous sensors in section 2.1, the Wireless Sensor Networks 

(WSN) paradigm that is used to realize this idea in 2.2. The well-established technology for providing 

energy autonomy to these autonomous sensors is presented in 2.3. Finally the extent of autonomous 

sensors in vehicles is presented in 2.4. 

2.1 Autonomous sensors: 

While sensor applications have been steadily increasing in vehicles, the world of sensors itself has 

been experiencing exciting developments due to breakthroughs in the areas microelectronics, micro- 

and nanotechnology. The integration of solid-state sensors with associated electronics and embedded 

networking services had given rise to autonomous sensing devices, i.e. independent embedded systems 

communicating data and events for machine consumption. The idea of autonomous sensors has been 

around for quite some time and gained traction with the definition of the Sensor Web by Delin et al. 

[6] at NASA's Jet Propulsion Laboratory (JPL). Paraphrasing from their definition, an autonomous 

sensor has at the least: 

1. Processing autonomy - the sensor has its own CPU in the form of a 

Microcontroller/ASIC/FPGA 

2. Communication autonomy - the sensor has means of communicating wirelessly 

3. Energy autonomy - the sensor has means of powering itself 

 

In very simple terms, autonomous sensors are smart embedded devices, which are not wired to 

anything else, and have capabilities to operate independently. Such autonomous devices can be widely 

distributed spatially and deployed in locations that are inaccessible. 

2.2 The Wireless Sensor Network (WSN) paradigm: 

The idea of networked sensors is an inevitable consequence of spatial distribution of autonomous 

sensors. This idea was born quite early with the first sensor network going operational in the mid-

1970s. The history of sensor networks is quite fascinating, beginning with the Sound Surveillance 

System (SOSUS) deployed by the US Navy on the Atlantic seabed to detect Soviet submarines [7]. 



15 

 

Academic interest in sensor networks began with the definition of the paradigm known as 'Smart dust' 

in a research proposal submitted by the University of California, Berkeley (UCB) in 1997 [8]. The 

term Sensor Network or Wireless Sensor Network (WSN) is commonplace today to describe this 

paradigm of spatially distributed networked embedded sensor devices, each of which is called as a 

sensor node or a mote. Nodes in a sensor network communicate by transmitting a message to another 

node in its neighborhood, which in turn passes on the message towards the destination. 

 

Figure 1 : A WSN mesh network [9] 

WSN are capable of deploying sensors covering huge geographical areas, reaching inaccessible places 

thereby providing remote sensing capabilities of far-reaching proportions. On the other hand it is also 

suitable for deployment in small physical areas such as a house or a room for Information Technology 

(IT)-based utilization of resources in the target environment. WSN represents the ultimate deployment 

model for sensors where each sensor is tasked with data collection and reporting in a completely 

autonomous manner. WSN communication protocols available today, such as Zigbee and Bluetooth, 

are highly capable and accomplish functionality specified in all the seven layers of the ISO Open 

Systems Interconnection (OSI) standard. 

2.3 Energy autonomy using Energy Harvesting: 

For too long a time batteries have been (and still are) at the forefront of providing energy autonomy. 

Frequent battery replacements and environmentally safe disposal concerns, however, have led to the 

evolution of techniques for recycling/scavenging or harvesting energy from the ambient environment, 

an approach that is known as energy harvesting. 

Energy harvesting (or ambient energy harvesting to be more accurate), refers to scavenging small 

amounts of power from the ambient environment to power autonomous sensors. Harvested power 

levels of  are typical, since CMOS VLSI based ultra-low power systems can 

implement a wide range of functions using such small amounts of power. In order to achieve complete 

autonomy using energy harvesting, a wireless sensor node usually has the following architecture: 



16 

 

 

Figure 2 : Architecture of an energy harvesting wireless sensor 

Ambient sources of energy include ambient vibrations, temperature gradients and solar energy, among 

many more. In most cases, a dedicated energy harvesting transducer is utilized to convert a chosen 

form of ambient energy into raw electrical power. A comprehensive study of energy harvesting 

research and development can be found, from among a plethora of sources, in [10] and [11]. Another 

major area of energy harvesting research is power management whose role is to convert the raw 

harvested electrical energy into usable DC power. Power electronics for energy harvesting is quite a 

mature area with many COTS offerings, while simultaneously being a field of active research, whose 

state of the art can be gleaned from [12]. With energy harvesting and wireless communication, a 

sensor node taking the above architectural form becomes autonomous. 

As mentioned in section 1.2, a survey of energy of energy harvesting technologies and energy 

harvesting power management have been compiled as part of WISCON in [2] and [3] respectively. 

2.4 Autonomous Wireless Sensor Networks (WSN) in vehicles: 

Electronic Control Units (ECUs) in vehicles have paid most attention to processing autonomy, with 

wired networked embedded devices being the order of the day. In a typical architecture of automotive 

control systems, ECUs with autonomous processing capabilities are placed on communication buses 

such as Controller Area Network (CAN) or Local Interconnect Network (LIN) buses. While at a 

system level the control units are more or less independent of each other, at each control unit the 

functionality is quite centralized. Also, since sensors are connected to individual ECUs, the idea of 

'Sensing as a service' and networked sensors is still at its infancy. The 'universe' of a sensor in a 

vehicle currently extends only as much as the ECU it serves.  

 

Communication autonomy is also in its infancy, reflected by the fact that the number of wireless 

sensors in a vehicle is currently very low. The newly introduced TPMS is often the only wireless 

sensor in a vehicle. This trend is set to reverse as communication autonomy has the potential to 

introduce significant cost savings by the removal of electrical cables, which are often several 

kilometers in aggregated length, along with savings in the associated assembly effort. In addition, new 

wireless sensors could be deployed in areas which were impossible to reach using cables, while they 

can also be used for applications which are too trivial to justify wiring costs. 

 

The idea of wireless sensors and WSNs in vehicles does provide tantalizing prospects and has already 

been the focus of significant amount of research. In a review by D'Orazio et al. [13] an extensive 

classification of in-vehicle communication networks based on application and data rate requirements is 



17 

 

provided. According to this review, current wireless technologies that are suitable for automotive 

applications are: 

 

Wireless networking standard Representative data rate Applications 

Ultra-Wide Band (UWB) 100Mbps Infotainment, body electronics 

WLAN/Wi-Fi 1Mbps Infotainment 

Bluetooth 500Kbps Infotainment 

Zigbee 150Kbps Body electronics 

Table 1 : Wireless networks for automotive applications due to [13] 

From this classification, it can be seen that wireless sensors and networks are currently being 

considered only for infotainment applications, where wireless LAN technologies are becoming 

popular. PAN technologies are being advocated for infotainment and body electronics applications. 

However according to this review, wireless sensors are not considered for safety and 'X-by-wire' 

applications.  

Energy autonomy for automotive sensors is slightly more mature compared to communication 

autonomy. While the automotive industry has pioneered technologies which recycle wasted energy, 

for example - waste heat recovery from vehicle exhaust using thermoelectric materials, ambient 

energy harvesting for automotive sensors has only begun to start getting attention. The automobile is 

quite rich in ambient sources of energy such as ambient vibrations, heat, solar and electromagnetic 

energy. A review of possible energy harvesting sources is presented in [14] and the list of possible 

sources is definitely not limited to this. 

This project aims to develop a prototype for proving the concept of using autonomous sensors in an 

automotive safety application. The prototype autonomous sensor would include provision autonomy in 

all three areas of processing, power and communication. 

  



18 

 

Chapter 3 - The Wheel Speed Sensor (WSS) and its applications 
 

The Wheel Speed Sensor (WSS) is the automotive sensor that has been chosen to be made 

autonomous in this project. The WSS transducer has certain capabilities, which are suitable for using it 

simultaneously as a sensor and an energy harvester. 

This chapter introduces the WSS in section 3.1 and the possibility of using the WSS as an energy 

harvester along with a review of comparable harvesting techniques is presented in section 3.2. The 

primary application of the WSS in trucks is presented in sections 3.3, 3.4 and 3.5 while other 

applications are presented in section 3.6. 

3.1 The Wheel Speed Sensor (WSS): 

The WSS is used to sense the state of rotation of a wheel in a contactless manner, using a magnetic 

sensor and a pole wheel that rotates along with the vehicle wheel. A sinusoidal signal produced by the 

WSS is a direct measure of the rotational speed of the wheel. In a vehicle, one WSS is provided for 

each wheel in a deployment that is illustrated by the following figure: 

 

Figure 3: An illustration of WSS assembly near the wheel of a truck [15] 

The WSS falls under the category of Variable Reluctance (VR) sensors, a class of electromagnetic 

sensors that is popular for sensing displacement. A typical VR sensor consists of a permanent magnet 

to which a ferromagnetic core (or pole piece) is attached, with a coil wound around the core. A WSS 

procured from WABCO (http://www.wabco-auto.com/) is used in the project, whose conceptual 

description can be seen below: 

http://www.wabco-auto.com/


19 

 

 

Figure 4 : The WABCO WSS and a depiction of its internal structure 

The permanent magnet acts as a source of magnetic flux, which when interrupted, induces a voltage in 

the coil. The sensor, sometimes called magnetic pickup, is therefore placed near a flux path whose 

reluctance (resistance to the flow of magnetic flux) varies in proportion to the feature that is being 

sensed. In this case, the variable flux path is provided with a gear wheel also called the pole-wheel, 

tooth-wheel or tone-wheel. The pole wheel, in the case of wheel speed measurement, is mounted on 

the axle of the wheel and therefore has the same RPM speed as the wheel. The WSS along with the 

pole wheel it is used with for RPM measurements is shown below: 

 

 

Figure 5 : The WSS along with a pole wheel 

The principle of operation of the WSS is to vary the reluctance of the flux path, depending upon speed 

of rotation, which in turn varies the flux that is linked by the coil. This induces a voltage proportional 

to the reluctance of the flux path, in accordance with the Maxwell-Faraday law. The flux of the 

permanent magnet placed near the wheel is variably linked by the teeth in the case of radial pole wheel 

and by the gaps in case of the axial pole wheel. The induced voltage is therefore proportional to the 

rate of change of flux as well as the distance between the VR sensor and the pole wheel. 



20 

 

3.1.1 A note on WSS terminology: 

The terminology for referencing the WSS setup and its constituent parts are not usually consistent and 

therefore the terminology used in this report will be clarified here. 

1. VR sensor/VR magnetic sensor/VR magnetic pickup/Magnetic pickup - this refers to the 

magnetic element shown in Figure 4. 

2. WSS - this term will be used to describe the complete assembly of the pole wheel and the VR 

magnetic sensor. However in certain places WSS will also refer to the VR sensor, a reflection 

of the common usage of the term WSS to refer to the magnetic sensor alone. 

3.1.2 WSS signal voltage: 

While the WSS operation is based on the Maxwell-Faraday law, a mathematical model for the output 

signal voltage of the WABCO WSS has not been developed because of the reluctance of the 

manufacturer to share the architectural details of the sensor in the duration of the project. Therefore in 

this project, empirical data is used to assess the signal voltage of the WSS. 

The speed of rotation of the pole-wheel primarily determines the amplitude of the WSS output signal 

voltage. Apart from the speed of rotation of the wheel, the WSS output signal amplitude also depends 

upon a number of parameters such as intensity of the magnetic field, number of turns in the induction 

coil around the magnet, magnetic properties of the pole wheel, the air gap between the pole wheel and 

the sensor, etc. Though the output voltage depends upon such an extensive list of parameters, most of 

them are typically set during installation and not controllable later. For most practical purposes the 

voltage depends only upon the speed of rotation of the pole-wheel.  

The main feature of interest from a sensing point of view is the frequency of the induced voltage 

which is proportional to the wheel rotation speed. While it is the frequency and not the amplitude of 

the induced voltage that is of primary interest for sensing purposes, it is advantageous (for reasons 

described in section 3.2.2) to tweak the assembly to get as strong a signal as possible. An illustration 

of the signal voltage from sample measurements provided by WABCO is shown in Appendix G. The 

figure presents output signal amplitude as a function of the air gap for different WSS. For an  

output it can be seen that the signal voltage has peak-to-peak amplitude of about . This data is 

simply an illustration because, as explained earlier, the signal characteristics depend upon the 

assembly. A detailed presentation of the WSS signal characteristics along with empirical data for the 

experimental assembly used in this project is presented in section 8.1. 

3.1.3 WSS signal power: 

The WSS has constant, mostly resistive (in the sub-kHz range), source impedance ( ) which, if 

exploited properly, makes it possible to draw significant amounts of power from it. The maximum 

power transfer theorem in circuit theory states that maximum power is drawn from a circuit if the load 

impedance matches the source impedance. However, the disadvantage of impedance matching is that 

the efficiency of power extraction is . This scenario can be represented by the graph below: 



21 

 

 

Figure 6 : Power and efficiency of an electrical circuit as a function of the load [16] 

Under conditions of maximum power transfer, the WSS average signal power is given by:  

  (1) 

 

Here  is the average signal voltage under a certain speed of rotation of the pole wheel. Assuming 

a source resistance of about , for the voltage produced as shown in Appendix G, it is possible to 

extract about  from the WSS which may be sufficient for powering certain applications. Here 

again this data is only an illustration of maximum power extraction and the empirical power data 

measured on the experimental setup used in this project is presented in section 8.1. 

3.2 The WSS as an energy source: 

While the application of the WSS as a signal source, i.e. a sensor, is fairly common, the main research 

angle of this project is determining the capabilities of the WSS as a power source. Electromagnetic 

generators based on the Maxwell-Faraday law are widely used for generating electrical power. This 

principle which is applied in the AC motor is also equally applicable to electromagnetic meso/micro-

energy harvesters of various types, which generate power using induction. 

It has been established in section 3.1.1 that the WSS is just another electromagnetic generator with a 

permanent magnet acting as the flux source and the pole wheel varying the reluctance. An illustrative 

case of the power that is available for extraction from this signal is provided in section 3.1.3. The 

primary source of this extractable energy is the kinetic energy of the rotating wheel and therefore the 

WSS is an electromagnetic energy harvester extracting energy from ambient rotational kinetic energy. 

This combination of sensing and power harvesting capabilities creates a unique convergence, which 

opens up the possibility of using the WSS both as the signal source as well as the energy source. 

In order to place the WSS in the context of electromagnetic energy harvesters a review of research in 

the area of electromagnetic energy harvesting and rotational kinetic harvesting is presented below. 

3.2.1 A review of electromagnetic and rotational kinetic energy harvesters: 

The WSS uses ambient rotational kinetic energy as its energy source and generates power by 

electromagnetic transduction. This represents the convergence of the areas of electromagnetic energy 

harvesting and rotational energy harvesting. 

Electromagnetic transducers for ambient kinetic energy harvesting have been widely discussed in 

literature. In most of these cases, the ambient kinetic energy is in the form of linear acceleration in one 

to three dimensions. A comprehensive review of electromagnetic transducers for ambient vibration 



22 

 

harvesting has been presented in [17]. The authors conclude from their survey that in its current state 

of research, electromagnetic vibration harvesters are in the range of   in size and 

 in output power. Most of the harvesters use the resonance phenomenon and 

produce highest amount of energy in a narrow range of mechanical vibrations. 

When it comes to rotational kinetic energy harvesting a variety of designs has been reported in 

literature. Among the piezoelectric transduction techniques, Manla et al. [18] have reported a non-

resonant piezoelectric transduction mechanism where the centripetal force due to the rotation of a 

wheel is used to make a ball bearing impact a piezoelectric material. The authors have reported 

measured power of  at a rotational speed of . Hu et al. [19] have proposed flexible 

piezo generators to harvest energy from the deformation of tires during rotation, meant for powering a 

Tire Pressure Monitoring System (TPMS) node. Khameneifar et al. [20] have proposed a piezoelectric 

cantilever with a tip mass that is attached to a rotating hub with a simulated output power of . 

A similar structure using a magneto-electric transducer which is a combined 

magnetostrictive/piezoelectric laminate, producing  has been reported by Wen et al. [21]. 

Gu and Livermore [22] have reported a radially oriented self-tuning piezoelectric cantilever beam. 

Rotational harvesting designs involving electromagnetic transduction of rotation include Conrad [23] 

who has reported pendulum based designs for harvesting power from both linear and rotational inertia 

of a ship’s propeller. The reported device is designed to be embedded in the propeller and is reported 

to produce an output power of  in the rotational speed range of . Toh et 

al. [24] have reported a modification of a conventional generator by attaching an unbalanced mass to 

the rotor of the DC generator. The design reportedly produces harvested power up to  at a 

relatively high rotational speed of . Wang et al. [25] have reported a well-weighted 

pendulum, i.e. a pendulum with one or more weights which adjusts the oscillation frequency to meet 

the rotational frequency and uses electromagnetic transduction to convert this oscillation to electrical 

power. The authors have explored a deployment in the wheel of a vehicle in order to power a TPMS 

node and have reported measured power in the range of  depending upon the speed of 

rotation.  

3.2.2 The WSS as an energy harvester: 

The WSS, like many of the reviewed examples, uses the stator-rotor combination for inertial 

harvesting. The meso-scale dimensions of the WSS is quite advantageous because, as shown in [26], 

scaling laws for electromagnetic transduction is such that higher open circuit voltages can be produced 

by meso scale wire wound coils. The WSS is easily capable of producing open circuit voltages, as will 

be shown in section 8.1, in the  range. This relatively high range of open circuit voltage is its 

biggest advantage because with a source impedance of  at maximum power transfer, this open 

circuit voltage provides a power of . This is comparable to the range of extractable 

power reported in literature. 

Structurally, the WSS system needs an elaborate assembly to fix the pole wheel to the axle and hold 

the VR sensor close to the pole wheel. The structural assembly of the WSS in a truck is presented in 

section 3.5 while the experimental WSS setup is presented in section 6.1. The elaborate assembly 

procedure makes the WSS less integrated as a structure when compared to the reviewed harvesters. 

However the advantage of using the WSS as an energy harvester is its wide acceptance in rotation 

sensing applications. The pole-wheel and VR sensor assembly can be used in any wheel-axle assembly 

such as those found in engines, generators and motors and similar machinery. The meso-scale WSS 

setup however restricts its use to meso-scale machinery, where size is not a concern. The Variable 



23 

 

Reluctance (VR) transduction technology is quite mature and can be used directly, without any 

modifications, as an energy harvester.  

3.3 Primary application of the WSS - The Electronic Braking System (EBS): 

The EBS acts as the platform for providing electronic control of braking services in vehicles. The 

primary function of EBS is the electronic actuation of pneumatic systems to engage brakes in vehicles. 

In addition to the core braking function, the EBS continuously monitors the condition of the brakes in 

order to detect malfunctions. The EBS platform also integrates a number of associated functionality 

such as Anti-lock Braking System (ABS), Traction Control System (TCS), Electronic Stability Control 

(ESC), etc. The figure below depicts an EBS that is installed in some Volvo trucks. 

 

Figure 7 : Knorr-Bremse EBS used in some Volvo trucks [27] 

3.4 The role of the WSS in the EBS: 

The WSS provides real-time information about the rotational speed of the wheel to which it is 

connected, and therefore acts as a primary signal source for the EBS Channel Module (otherwise 

known as the EBS Modulator). The Channel Module processes the WSS signal and produces the 

instantaneous wheel speed data which is made use of for the following control functions: 

1. Vehicle speed indicator – The speed data that is displayed on the dashboard is derived from 

the WSS signal 

2. Gear selection - In the I-Shift intelligent gear shift system found in Volvo trucks [28] uses the 

instantaneous vehicle speed along with other characteristic data to decide the gear that has to 

be engaged. 

3. Anti-lock Braking System (ABS) – Whenever brakes are applied the WSS signal is used to 

detect if the wheel is rotating/unlocked or not rotating/locked. If a wheel is locked while the 

brake is applied, there are chances that the wheel is sliding instead of rotating which causes 

maneuvering problems. Detecting and controlling undesirable wheel locking falls under the 

purview of the ABS. 



24 

 

4. Anti-slip Regulation (ASR)/Traction control – This is a secondary function of the ABS 

system where the intention is to prevent loss of traction. Possible differences in traction, i.e. 

the grip of the wheel on the road, between different wheels are derived from the instantaneous 

wheel speeds (available from the WSS). Corrective action is taken by a combination of 

throttling and applying brakes on selected wheels to equalize traction. 

5. Electronic Stability Control (ESC) – When a vehicle is in operation there is a possibility of 

loss of steering control. This happens where the actual direction of the vehicle does not match 

the intended direction of motion set using the steering wheel, which is possible during evasive 

swerves, under steering or over steering. In such cases, corrective action is taken to restore 

stability by applying brakes to counter-steer. The actual direction of the vehicle is calculated 

using measured lateral acceleration, the vehicle yaw and the wheel speeds (using the WSS). 

The instantaneous wheel speed is quite an important piece of information and therefore its usage may 

not be limited to the above list of applications.  

 

Figure 8 : WSS in the EBS setup [27] 

3.5 WSS installation as part of the EBS: 

WSS installation in Volvo trucks can be broadly divided into two phases which are 

1. WSS mounting on axle - the installation of the sensor close to the wheel on the axle which is 

usually done by the axle manufacturer 

2. WSS harness mounting - the installation of cables connecting the sensor mounted on the axle 

to the EBS modulator which is done by the vehicle manufacturer 

The WSS installation on an axle is an elaborate process which varies depending upon the type of the 

axle and the type of the brake used. A detailed presentation of WSS installation in different axles can 

be found in the Volvo assembly document [15], out of which the mounting process for a front wheel 

with a disc brake is presented here for illustrative purposes. The WSS assembly on the front axle of a 

Volvo D68 truck is shown below. 



25 

 

 

Figure 9 : WSS mounting on the front axle of Volvo D68 - with deflector and protection shield [15] 

Once the WSS is mounted on the axle, the next assembly step is to mount the harness connecting the 

WSS to the EBS modulator. The harness mounting procedure, just like the WSS mounting on the axle, 

depends upon the axle. Harness mounting procedures for different axles are described in the Volvo 

installation requirements document [29]. The assembly of the harness typically involves connecting 

the routing cables, by securely fastening them using clamps at different points, to ensure that the 

electrical connection is not degraded or lost.  

 

Figure 10 : WSS harness mounting procedure for the front axle [29] 

The figure above illustrates the harness clamping process for connecting the WSS to the chassis 

harness connector. The red loops in the picture indicate clamping points, one of which is marked, 

which are affixed manually costing assembly effort. The total cost for the harness material and 

assembly, as shown in Appendix G is about . A cost analysis of the WSS axle assembly has 

not been presented because it is the harness assembly that stands to be replaced as explained further 

on. 

3.6 Other Variable Reluctance (VR) sensor applications in a vehicle: 

The same pole wheel and VR sensor setup used in the WSS is reused for similar applications in the 

vehicle. For example, it is used for a number of position and speed sensing operations in the engine 

such as the crankshaft position sensor and the camshaft position sensor. An illustration of the camshaft 

and the crankshaft is shown below. 



26 

 

 

Figure 11 : A camshaft (left) and a crankshaft (right) 

By using VR sensors, the positions of the camshaft and crankshaft can be sensed which is then used to 

determine the cylinder position and the engine stroke which in turn plays an important role in fuel 

injection. This sensing mechanism can be extrapolated and the VR sensor can be used for 

position/speed sensing in any rotational frame of reference.   



27 

 

Chapter 4 - The Autonomous Wheel Speed Sensor (AWSS) 
 

The harness assembly cost for the WSS, shown in section 3.5, in combination with the energy 

harvesting capabilities of the WSS, shown in section 3.2, makes the WSS a reasonable target for 

rendering autonomy. The synergy of sensing and energy harvesting capabilities of the WSS sensor to 

create an Autonomous WSS (AWSS) is a notion that is quite attractive. 

This chapter presents the functional architecture of the proposed AWSS node and its subsequent 

integration in a sensor network in sections 4.1 and 4.2. The challenges of implementing such a system 

and the potential benefits offered by the system are presented in sections 4.3 and 4.4. 

4.1 The functional architecture of an AWSS node: 

Synergizing the sensing and energy harvesting capabilities of the VR magnetic sensor lies at the heart 

of converting the passive WSS into an autonomous WSS, and the architecture of the AWSS node 

reflects this idea as shown below.  

 

Figure 12 : The Autonomous Wheel Speed Sensor (AWSS) 

By integrating the WSS with an energy harvesting mechanism and a wireless sensing platform, which 

includes a microcontroller and RF transceiver, the new embedded device called the AWSS is created. 

Because of the wireless communication of the AWSS, the harness assembly of the passive WSS stands 

to be eliminated. A brief description of the constituent parts of the AWSS follows. 

The Energy Harvesting Unit (EHU): The EHU would be responsible for converting the raw 

sinusoidal signal output of the WSS into usable DC power. As pointed out in section 2.3 the research 

area of energy harvesting power management is purely dedicated towards accomplishing this process 

chain efficiently. The EHU in combination with the VR sensor act as the power source for the AWSS 

and therefore provides power autonomy. 

The Wireless Sensing Unit (WSU): The WSU is responsible for implementing the node level 

management, sensing and messaging operations of the AWSS. It is therefore responsible for providing 



28 

 

the AWSS with processing and communication autonomy. Powered by the EHU, the WSU senses and 

processes the WSS signal output. Based on the processed results it makes determinations and 

communicates them, some in real time. The WSU must perform all these operations, while consuming 

the least amount of power possible. This precludes the usage of ultra-low power hardware processing 

platforms. 

4.2 The Autonomous Wheel Speed Sensor Network (AWSSN): 

Reflecting the WSS deployment, one AWSS node needs to be deployed per wheel, and the data from 

each AWSS has to be collected and provided to the EBS system. In order to integrate each individual 

AWSS into the EBS, a WSN infrastructure is required. A WSN would take care of setting up and 

managing the network of AWSS nodes while seamlessly routing data between them and the EBS. A 

potential AWSSN could take the following form where one AWSS node or an End Device (ED) is 

attached to every wheel in the vehicle and all of them communicate with an Access Point (AP). 

 

Figure 13 : The Autonomous Wheel Speed Sensor Network (AWSSN) 

The ED, which is the AWSS node, has already been described, while the AP is a typical WSN sink 

node, which acts as a bridge between the EBS and the WSN. It routes the data between each ED and 

the EBS while in addition creating and maintaining the network of autonomous nodes. While there is 

no question that such a WSN would be quite a challenge to implement, the technology for such a 

network is already available and ripe for prototyping. 

4.3 Challenges in implementing the AWSSN: 

The application of autonomous sensors, which use energy harvesting and wireless communications, in 

a critical real-time control system like the EBS raise a lot of concerns. These concerns, along with a 

few more challenges, which have to be overcome to realize the network of autonomous sensors, are as 

follows. 

1. Availability concerns: The biggest concern of an autonomous sensor is whether it is available 

at a certain point of time to perform an operation. Here the term ‘availability’ is defined, by 

the field of reliability engineering, as the proportion of time for which a system is functioning. 

If a system experiences a failure, which renders it incapable of performing an operation, it is 

in a state on non-availability. Availability is heavily used for evaluating the performance of 



29 

 

equipment in the telecom and IT industries. In energy harvesting and in wireless sensors, ‘high 

availability’ is a derivative definition where, for example, efficiently harvesting power and 

judicious usage of this harvested power ultimately contributes towards high availability. High 

availability of autonomous sensors is a definition that is commonly used in this report. 

One way of increasing availability is to increase the amount of energy that is being harvested. 

Increasing the efficiency of transduction and power extraction is hotly being pursued as [10], 

[11], [12] and [30] would show. Another approach is the hybrid energy harvesting technique 

where multiple ambient sources of energy are tapped to increase the harvested energy levels. 

Examples include the combination of kinetic and RF energy harvesting using a single coil 

proposed in [31] and the hybrid ambient light and thermal harvester proposed in [32]. 

Another way of increasing availability is judicious usage of the harvested power, i.e. to reduce 

the duty cycle of operation. A set of design principles have been proposed in section 5.4, 

which calls for abstracted passive sensing to reduce duty cycle and power consumption. Ways 

of improving availability are not limited to the options described here and there is little doubt 

that as the technologies related to autonomous sensors mature, the availability levels would 

improve. 

2. Real-time task execution concerns: As pointed out in [13], wireless sensors for critical 

automotive applications in the areas of safety and X-by-wire are currently not being seriously 

considered due to concerns ranging from latency to interference. In addition to being wireless, 

an autonomous node is also powered by energy harvesting which only adds up to the real-time 

execution concerns.  

Real-time scheduling in energy harvesting sensors is in a state of infancy and a sample of the 

research activity in this area is presented here. Scheduling mechanisms for recurrent tasks 

have been proposed by Audet et al. [33] and lazy scheduling of tasks with non-deterministic 

arrival times proposed by Moser et al. in [34]. Gu and He propose a mechanism for bounding 

delays in communication in [35] while Liu et al. propose the adaptation of the well-known 

Dynamic Voltage and Frequency Scaling (DVFS) technique for energy harvesting sensors in 

[36] as do Dehghan and Kargahi in [37]. 

Though it is currently not a widespread application area of WSN, real time wireless sensing 

and control is certainly not new. Implementation of real-time scheduling for wireless control 

and sensing, especially in the industrial process control area, is actively being researched. The 

WirelessHART open communication standard seems to be a popular choice in this area as 

reported in [38], [39] and [40]. WirelessHART uses a TDMA based transmission mechanism, 

more akin to voice communication standards like GSM, which makes real-time operations 

possible. Though WirelessHART is not used in this project, the intention here is to illustrate 

that real-time wireless sensing and control is definitely achievable.  

3. Integration concerns: A significant area of concern is the integration of the newly added 

electronics to the WSS assembly. A suitable mounting point for the autonomous sensor and 

suitable packaging must be chosen to integrate the CMOS electronics so that the node 

electronics is able to withstand the rigorous environment close to the wheel. The mounting 

point, in many ways, decides whether this application is feasible or not since the new 

assembly could turn out to be more elaborate than the existing assembly. 

 



30 

 

4. New assembly procedures: Procedures have to be developed to program and assemble these 

new ECUs in the vehicle. This would present a number of challenges, for example, a 

particular AWSS node has to be coupled both physically and in program with a particular 

wheel so that the modulator can recognize the exact wheel to which a particular message 

applies. This is not a major technical challenge but it does require intensive data collection and 

management for maintaining node level information. 

 

5. EMC concerns: With the introduction of new wireless sensors, extensive evaluation needs to 

be done to ensure that the addition of this sensor does not affect any of the electronics in the 

neighborhood. The radio emissions must be balanced with signal power requirements of the 

wireless communication standard being used. 

4.4 Potential benefits: 

While the development of the AWSSN presents many challenges, there are potential gains from such 

an application which are listed below. 

1. Assembly cost savings: The AWSSN could potentially eliminate all cable and connector costs 

listed in Appendix G and save a maximum of  per sensor. However, net savings 

depend very much upon the chosen method of integration and assembly of the proposed 

AWSS node. If the WSS mounting point on the axle is determined to be too hostile for the 

newly added electronics, then a mounting point that is further away has to be chosen, which 

incurs associated cabling and assembly costs. While the determination of net cost savings 

would certainly be non-trivial, a cost of more than  for mounting the WSS in the current 

form does warrant a serious consideration of this alternative method.  

 

2. Warranty savings: The use of the autonomous WSS instead of the wired version has the 

potential to reduce aftermarket maintenance. Like any other harness, the WSS harness is 

responsible for a majority of the reported aftermarket WSS failures. These harness 

connectivity problems and associated failure can be potentially reduced using real-time 

wireless communication links. In addition, diagnosing problems in a workshop becomes much 

simpler as there is no harness to trace faults. 

 

3. Customization savings: Due to cable length variations for different axles, there is a 

considerable customization effort where, for every axle, a separate WSS configuration is 

developed, stocked and maintained. Each WSS, with a particular harness configuration 

constitutes a ‘part’, which has to be individually developed and stocked. Because of the 

wireless link, the AWSS ‘part’ can be made agnostic to the axle on which it is mounted and 

this can drastically reduce the customization effort. 

 

4. Possible savings from EBS architectural changes: There could be several changes in the 

EBS system architecture, due to the deployment of new wireless sensors, which could lead to 

savings. For example, if all the nodes communicate wirelessly with the modulator, one or two 

access points could replace the speed sensing electronics that is present in each modulator. 

The modulator performs actuation functions which can possibly be re-architected to make use 

of the intelligent sensors deployed near the wheel. 

 

5. New EBS applications: The AWSS node increases the level of distribution of intelligence in 

the EBS system which opens up the possibilities of new applications. For example, it would 



31 

 

be possible to sense various wheel events such as wheel lock, wheel slip, etc. at the wheel and 

control them using this new distributed system.  

 

6. New wheel management applications: While the AWSS concentrates specifically on wheel 

speed monitoring, the idea of using the VR sensor as an energy source need not be restricted 

to this particular application. If the entire pole wheel and VR sensor setup is moved to the 

wheel hub or even inside the tires, then it would be possible to use the harvested energy to 

power any number of wheel monitoring sensors. This broadens the horizons of the VR sensor 

which can basically act as an energy harvesting transducer platform for any wheel-axle 

assembly. 

 

One such wheel management application is the TPMS where a pressure sensor is mounted 

inside the wheel to monitor air pressure. Because of the inaccessible nature of the sensor, a 

significant amount of research is being directed towards finding autonomous energy sources 

for this sensor. In-tire vibration and strain energy harvesters seem to be popular choices, as 

pointed out by a few examples such as [19], [41] and [42]. Alternatively some of the proposals 

target RF energy harvesting as seen in [43]. Being a wheel-axle assembly, the VR sensor and 

pole wheel combination can be adapted to fit inside the tire to power the TPMS. Tire pressure 

monitoring may not have as critical real-time execution requirements as the EBS system and 

therefore energy harvesting sensors can be used with a higher amount of confidence in such an 

application.   



32 

 

Chapter 5 - Prototype specification 
 

As a starting point, the simplest version of the AWSSN, as described in section 4.2, is targeted for 

implementation in this project. While the prototype cannot target all of the perceived challenges listed 

in section 4.3, it will target the most important operational concerns of availability and capability for 

real-time task execution. 

 

The architecture of the targeted prototype sensor network is presented in section 5.1 and the functions 

chosen to implement in this prototype system is listed in section 5.2. A systematic parametric 

specification of the prototype system is specified in section 5.3 and a set of general design principles 

for implementing any real time energy harvesting autonomous system is presented in section 5.4. 

5.1 Prototype network architecture: 

The AWSSN in its simplest form consists of one AWSS node acting as the ED and one AP which 

receives messages from the ED and relays them to the EBS modulator. An AWSSN consisting of only 

two nodes is the target of implementation. This is depicted by the following figure. 

 

Figure 14: Conceptual depiction of the targeted prototype - a two node AWSSN 

However the EBS modulator in its current form is designed to operate only with the raw WSS signal 

and cannot process discrete messages. Since modification of the modulator is not possible in the 

limited time that is available for this project, an additional simplification is done by implementing a 

Matlab based Pseudo Modulator (PM). The PM simply displays messages relayed by the AP, in a 

setup that is depicted by the following figure. 

 

Figure 15 : Targeted prototype AWSSN setup 



33 

 

The choice of the two-node WSN abstracts away many challenges in real-time wireless scheduling 

such as Media Access Control (MAC), network level routing and many more challenges posed by a 

multiple node WSN. All these challenges are left to be tackled in future iterations of the project. 

5.2 Prototype functionality: 

In the EBS system deployed in trucks the WSS signal data is used for a number of applications, as 

described in section 3.4. Among these applications the ABS is targeted and the prototype AWSSN 

system implements a representative sample of functions intended for the ABS application. From the 

performance of these representative functions, results can be drawn to evaluate applicability to other 

EBS applications. It must be noted at this point that no concern apart from high availability and real-

time task execution are considered in the prototype implementation. Keeping in mind the time-limited 

nature of the project, the following functionality were chosen for implementation as a representative 

set for the ABS. 

 

1. Periodic wheel speed sensing - the ED shall sense and transmit the speed of the wheel 

periodically to the AP which relays the message to the PM.  

2. Wheel rotational state sensing - the ED shall sense the starting/stopping of the wheel and 

instantaneously transmit an event to the AP which is relayed to the PM. This message would 

signal that the wheel has started/stopped rotating. 

The periodic wheel speed sensing and transmission represents the data stream which operates at a far 

lower duty cycle than the passive WSS and the rotational state sensing represents the event stream.  

An emphasis on event-driven communication and control, in place of data-driven communication, in a 

network of autonomous sensors is one of the key recommendations of the project. Continuously 

transmitting wheel speed data makes poor use of the processing capabilities of the AWSS and the 

distribution of intelligence. While data and events can be combined, the best case scenario of operation 

with the AWSS node would be a completely event-driven model where the node would handle most of 

the management functions on-site and communicate only important events up the hierarchy.  

The main reasoning behind this idea is the fact that abstracted events are likely to occur less frequently 

in time and therefore can potentially reduce the duty cycle of operation, consequently improving the 

availability. This principle motivates most of the design choices of the prototype. Extrapolating this 

idea, when it comes to applications such as ASR, ESC, etc. the recommended course of operation is to 

detect events such as loss of control and loss of stability at lower levels of system hierarchy and 

communicate them to points of control/for actuation.  

5.3 Parametric model: 

A systematic specification of the targeted AWSSN, as shown in Figure 15, is the first step in choosing 

the hardware, designing and eventually developing the prototype. The parametric definitions in this 

section are quite extensive and for ease of reference, a table of all parameters (the parametric register) 

has been compiled and presented in Appendix E. The notations of the different parameters are used 

consistently throughout the report and in the source code of the software. 

5.3.1 Specifying the WSS: 

As noted earlier, the WSS is the primary object of interest for sensing and energy harvesting and in 

many ways dictates the requirements and limitations of the system. The WSS generates a sinusoidal 



34 

 

voltage signal at its output whose frequency depends upon the speed of rotation of the pole wheel. The 

WSS can be specified using the following parameters. 

Notation Description 

 Air gap between the pole wheel and VR sensor 

(m) 

 Diameter of the wheel (m) 

 Number of teeth in the pole wheel 

 Minimum wheel speed (Kmph) 

 Maximum wheel speed (Kmph) 

 Sensor output voltage for 1Hz (V/Hz) 

 Minimum pole wheel teeth frequency (Hz) 

 Maximum pole wheel teeth frequency (Hz) 

  

  

 Sensor output voltage at  

 Sensor output voltage at  

 Source impedance of the WSS (Ω) 

 Minimum average signal power with matched 

impedance (W) 

 Maximum average signal power with matched 

impedance (W) 

 Mean average signal power with matched 

impedance (W) 

 The average number of rotational transitions per 

hour 

 Proportion of time for which the wheel is locked 

i.e. not rotating 

Table 2 : WSS functional parameters 

The speed at which the wheel of a truck rotates is a logical place to begin modeling as it is this 

rotational speed that is represented by the WSS output. The conversion from speed in Kmph to 

frequency in Hz can be done using the conversion factor 

  ( 2 ) 

 

Using this, the frequency parameters  and  can be derived from  and  using 

  ( 3 ) 

 

It should be noted here that  because the WSS is not capable of detecting a speed of . Since 

the relation between the frequency and the induced voltage in a VR sensor is linear, according to the 

Maxwell-Faraday law, the voltage parameters   and  can be derived from the frequency 

parameters  and  and  using 

  ( 4 ) 

 

The output power limits  and  can be derived from the voltage limits  and  using 



35 

 

  ( 5 ) 

 

As the wheel is put through its paces, its speed converges to a statistical average . If the 

instantaneous wheel speeds are uniformly distributed, then this mean average power is given by 

  ( 6 ) 

 

It must be noted that the choice of a uniform distribution is only for illustration and the idea is valid 

with any other statistical distribution. In practical cases, the distribution of wheel speeds is highly 

unlikely to be uniform, but the distribution is chosen nevertheless for simplicity. The average power in 

( 6 ) is applicable only if the wheel is rotating continuously. If the proportion of time for which the 

wheel is not rotating (  is considered then the average power is given by 

  ( 7 ) 

 

The above reasoning follows elementary probability theory and is generic enough to be applied with 

any statistical model of wheel rotation.  

As an illustration, the transition of the rotational state of the wheel can be modeled using the well-

known Random Telegraph (RT) process, described in Appendix A. The RT process is a readymade bi-

stable random process where the transition between the states follows a Poisson distribution. Since the 

state probabilities converge, as shown in Figure 53 and Figure 54, the RT process model is modeled in 

a way that is sensitive to knowledge of the initial state. As the wheel of a truck is meant to be rotating 

most of the time, the assumption is made that the wheel starts from an unlocked position. According to 

the RT process model in an interval of time  we have 

  ( 8 ) 

 

The choice of  turns out to be necessary because, from ( 41 ) and ( 44 ) , we have 

 

 

This shows that a choice of  is quite restrictive for modeling the wheel rotation because it is 

necessary for  to be less than . The sensitivity of the knowledge of initial state however 

reduces exponentially with the Poisson parameter as shown in ( 46 ). 

The parameter  in ( 8 ), is the average number of rotational state transitions per hour. Using 

( 8 ) in ( 7 ), 

  ( 9 ) 

 



36 

 

The choice of the RT process may not be optimal for modeling wheel rotation because of the 

convergence of state occupation probabilities. As time progresses, in the RT process model, both 

states become equally likely to be occupied as shown in Figure 53. The fitness of the model can 

therefore be evaluated only after careful examination of wheel speed data. Another simplification in 

the above model is that the distribution of wheel speeds and the locking of the wheel has been 

modeled independently, which is also not true in reality. The main intention here is to illustrate and 

promote stochastic modeling of the wheel speed for analyzing the available power as well as the 

sensing activity.  

5.3.2 Modeling the AWSS node using the 'leaky bucket' analogy: 

The ED is the primary object of interest in this project and represents a major part of the 

implementation effort. The operation of the energy harvesting node ED can be described as using 

water from a 'leaky bucket'. Here the water in the bucket represents the energy that is available in the 

EHU at the disposal of the WSU and the leak represents the actual usage of energy from the bucket. 

The EHU always tries to replenish the leak of energy that is drained by the WSU. The bucket here is 

normally a capacitor, or in some cases a supercapacitor, capable of providing power in bursts. The 

EHU could either replenish the leak directly from the harvester input or use a bucket of its own, for 

example a rechargeable battery. The analogy is depicted by the following figure: 

 

Figure 16 : The 'leaky bucket' model of the energy harvesting ED node 

For sustainable operation the refill rate must be able to sustain the leak rate, which in effect captures 

the essence of operation of an energy harvesting system. The leaky bucket analogy has been used to 

model the operation of the ED node and design its functionality. In the above figure, the harvester, 

rechargeable battery and capacitor are part of the EHU and the WSU acts as the load. 

5.3.3 Specifying the EHU of the ED:  

The power that is available from the WSS in an average sense is modeled by ( 7 ) and in the specific 

case of an RT process model by ( 9 ). The EHU must be able to successfully harvest this energy from 

the WSS output signal and supply a regulated DC voltage. The EHU parameters can be listed as 

follows: 



37 

 

Notation Description 

 Minimum harvestable voltage at the energy 

harvester input (V) 

 Maximum harvestable voltage at the energy 

harvester input (V) 

 Minimum allowed transducer source impedance 

(Ω) 

 Maximum allowed transducer source impedance 

(Ω) 

 The capacity of the rechargeable battery (µAh) 

 Effective series resistance of the battery(Ω) 

 Output capacitor (F) 

 Instantaneous supply voltage across  (V) 

 Instantaneous discharge from  (µAh) 

 Maximum supply voltage across  
 (V) 

 Minimum allowable voltage to which  can 

discharge (V) 

 The charging voltage of  (V) 

 Maximum allowable peak discharge (µAh) 

Table 3: EHU parametric specification 

With the conditions 

  ( 10 ) 

 

  ( 11 ) 

 

The capacitor  provides the output power to the load which in turn must draw the power in bursts. 

In one burst the load draws a current of  in time  and discharges  from  

to . The discharge of the capacitor is modeled as follows. 

  ( 12 ) 

 

Though this discharge is exponential, if the discharge magnitude is small enough, it can be 

approximated as a linear discharge. In such a case 

  ( 13 ) 

 

Using ( 13 ) in ( 12 ) we have 

  ( 14 ) 

 



38 

 

The product of the average current drawn and the time for which it draws this current gives the charge 

that is drawn out from the buffer capacitor. The expression in ( 14 ) expresses the discharge in units of 

 which is converted into units of  as follows 

  ( 15 ) 

 

Expressing the discharge in terms of a current-time product is very useful for practical embedded 

system development where the current that is drawn and the time for which is drawn are easily 

relatable in software and hardware.  

Even though  can go down close to , this is not desirable because most CMOS circuits need a 

supply voltage of at least  to operate. In addition, too much fluctuation in  may also not be 

desirable because there may be circuit applications which depend upon the magnitude of the supply 

voltage. This implies that the discharge of  should be carefully regulated to ensure that  stays 

within well-defined limits. In this model these limits are set by  and , with preferably 

a small difference between them. These voltage limits can be converted into an equivalent discharge 

limit which can be expressed as follows. 

  ( 16 ) 

 

The WSU must make sure to not draw a charge that is higher than  so that the voltage across 

the capacitor is maintained between  and .  

Having modeled the discharge, attention can now be turned to modeling the charging of  by the 

rechargeable battery. As shown in the leaky bucket analogy, the leak of charge from  due to 

consumption by the load is replenished by the rechargeable battery. When  has been discharged 

to , the time taken for the battery to charge it back to  is given by 

  ( 17 ) 

 

Therefore after discharging  to , the WSU must necessarily refrain from further discharging 

the capacitor for at least  amount of time, a necessity which arises out of the indefinite nature 

of periodic transmission. 

While being nearly accurate models of the charge and discharge processes, ( 14 ) and ( 17 ) need to be 

simplified further to be practically usable. For this, another approximation is made that the discharge 

from  to   is quick enough such that the average value of the drop is . While this 

discharge profile does not reflect the exponential discharge it may just work if the discharge is below, 

for example,  of . In such a case, we have 

  ( 18 ) 

 



39 

 

The above equation can equivalently expressed, based on ( 16 ), as 

  ( 19 ) 

 

While being approximate, the expressions in ( 17 ) and  ( 19 ) are more practically usable as the only 

unknowns would be  and  while all other values can either be measured or set. The value of 

  effectively sets the limits on the periodic transmission period and therefore the duty cycle of 

operation of the WSU. 

5.3.4 Specifying the WSU of the ED  

The operation of the WSU can be described as follows. 

1. The WSU must be able to sense and process sinusoidal waves ranging from  

to .  

2. As long as the WSS output signal frequency , the WSU must be able to sense if  drops 

to zero (or below a certain threshold) indicative of the fact that the wheel is locked and 

immediately transmit a message indicating this event. 

3. As long as  (or below a certain threshold), the WSU must be able to sense if  crosses 

zero or the threshold indicating that the wheel has unlocked and transmit a message 

communicating this event. 

4. Sensed results should be transmitted on a standards based RF link 

5. The WSU must be able to save energy by having a duty cycle of <1 

Based on the above statements, a classification of ED functionality can readily be formulated as 

follows.  

 

Figure 17 : Classification of WSU functionality 

Here attribute sensing refers to the sensing of transient attributes in the WSS system, which in this 

case is the periodic speed sensing and transmission. 

WSU sensing model: The attribute sensing functional requirement implies that the WSU should have 

a frequency sensor. Time period sensing was chosen to implement frequency sensing because of the 

ready availability of digital time period sensing techniques and the ability to realize these techniques 

without consuming too much power. Here, time period sensing uses a counter to determine the time 

elapsed between successive rising/falling edges of the WSS signal output.  

It has to be noted that the WSS signal varies not only in frequency but also in amplitude. For pure 

frequency sensing purposes, it is advantageous to abstract away these variations in amplitude and 



40 

 

present a consistent signal that varies only in frequency. In order to perform this conversion a specific 

interface circuit called the ‘WSS pulse converter’ is now defined which implements the logic: 

WSS signal input WSS Pulse converter output 

  

  

Table 4: WSS pulse converter truth table 

Here  is the instantaneous voltage of the WSS output signal. Typically a switching threshold of 

 is not possible due to voltage hysteresis, and is usually around a few hundred millivolts. The time 

period sensing functionality, as implemented by the pulse converter and the time period sensor, can be 

specified as follows. 

Notation Description 

 Sensing window (number of rising edges to sense 

for calculation and averaging) 

 Timer word length (bits) 

 Timer clock rate 

 Time taken to sense and process an attribute (s) 

 Minimum interval between successive 

transmissions of the attribute (s) 

Table 5 : WSU wheel speed attribute sensing specification 

So that we have conditions: 

  ( 20 ) 

 

  ( 21 ) 

 

  ( 22 ) 

 

Since attribute sensing senses the time period of a pulse wave, the sensing duration depends upon the 

instantaneous frequency of the signal. Keeping up the assumption of uniformly distributed speeds, the 

attribute sensing time can be expressed using the expected value of this distribution. 

  ( 23 ) 

 

The attribute transmission interval is chosen to be fixed in the prototype implementation and must 

necessarily be longer than the longest attribute sensing period. 

  ( 24 ) 

 

When it comes to event sensing, the WSU must be able to sense the event, i.e. the locked or unlocked 

state of the wheel, and communicate it. Though the lock status can be sensed from the periodic speed 

attribute, such sensing does not satisfy real-time event sensing needs and calls for a dedicated 



41 

 

rotational state sensor. The rotational state sensor is a bi-stable multivibrator, whose two states 

represent the rotational states, as depicted by the following truth table. 

WSS sinusoidal input Rotational state sensor output 

  

  

Table 6 : WSS rotational state sensor truth table 

The transitions of the rotational state sensor output are indicative of the change of state from rotating 

to stationary and these transitions are primary objects of interest. The event sensing functionality can 

be specified by: 

Notation Description 

 Lock threshold frequency (Hz) 

 The rise time of the output signal indicating an 

unlock event (s) 

 The fall time of the output signal indicating a 

lock event (s) 

 Time taken to sense a transition event (s) 

 The average time between two transition events 

(s) 

Table 7 : WSU event sensing specification 

When it comes to modeling , one has to consider the fact that the events occur in a non-

deterministic manner. Therefore for highest reliability of event detection, the event sensing period 

must be indefinite. 

  ( 25 ) 

 

The average duration between transitions is given by: 

  ( 26 ) 

 

WSU communication model: After sensing and processing the WSS signal, the WSU must then 

communicate results to the AP. The communication functionality can be specified as follows. 

Notation Description 

 PHY level packet length (bits) 

 Data rate of transmission (kbps) 

 The time taken for transmitting  
 (s) 

Table 8 : WSU RF communication specification 

Here,  would include the attribute/event payload and the overhead that is added to it by the WSN 

communication standard. Parameters such as the PHY level overhead and the data rate are well 

specified by the WSN communication standard.  



42 

 

  ( 27 ) 

 

Apart from communicating attributes and events the WSU must also be able to maintain the link 

between the ED and the AP according to the WSN standards, but a detailed specification of WSN 

management is beyond the scope of this project. 

WSU duty cycle model: The WSU and hence the ED consumes energy when it senses and transmits 

the attributes and events. This system, or indeed any energy harvesting system, is better off if it has an 

operating duty cycle of , so that it is active for a significantly lower amount of time 

compared to the time when it is not. This is mainly because of the intermittent availability of the 

ambient energy and finite EHU capacity, which precludes the sensor from operating continuously. 

New parameters can now be defined to define the duty cycle of the ED: 

Notation Description 

 The active period for attribute sensing and 

transmission (s) 

 The active period for event sensing and 

transmission (s) 

 Duty cycle of operation 

Table 9 : WSU duty cycle specification 

In the attribute sensing case, the active period is quite straightforward as the sensing time  is 

determinable. Therefore we have 

  ( 28 ) 

 

When it comes to event sensing, we have a sensing period that is indefinite, which in practical cases 

could last several minutes. If an active sensor is used, then this indefinite active period - which could 

draw steady power for many minutes - would potentially violate the discharge limits set by ( 14 ).  

In order for the rotational state sensor, or indeed any sensor in an energy harvesting system, to be 

indefinitely active, it must necessarily use passive means of sensing. 

Now if the rotational state sensor is passive, an indefinite active period has no impact on the active 

duty period. The event sensing function is now considered active only when a transition occurs which 

triggers a radio transmission. 

  ( 29 ) 

 

Since in most practical cases  and  , duty cycle is 

given by: 

  ( 30 ) 

 

The above equation can be equivalently expressed using ( 26 ) as 



43 

 

  ( 31 ) 

 

With event sensing and transmission now in place, it is not necessary to transmit the attributes in the 

duration the wheel is locked. This means that the sensor is not active for the entire period when the 

wheel is locked and this significantly reduces the duty cycle as seen below. 

  ( 32 ) 

 

In ( 32 ) the advantage of event sensing and transmission becomes apparent because it decreases the 

most influential term of the duty cycle by a factor of . 

It is not surprising to note that passive sensing is better in every case which is shown by the fact that if 

the pulse converter is passive,  is essentially  so that  

  ( 33 ) 

 

This reduction in  further brings down the duty cycle of operations. Therefore, while passive 

sensors are best to have for both deterministic and non-deterministic sensing, it is an absolute 

requirement for non-deterministic sensing in autonomous sensors. 

WSU average discharge and power consumption model: The ED functionality has now been broken 

down into a set of discrete operations, namely event and attributes sensing. The average discharge and 

power consumption, during the operational lifetime of the ED, can be calculated if the discharge and 

the current drawn due to each individual operation are known. Defining new parameters: 

Notation Description 

 RMS current drawn during attribute sensing and 

transmission (A) 

 Buffer capacitor discharge due to attribute 

sensing (µAh) 

 RMS current drawn during event sensing and 

transmission (A) 

 Buffer capacitor discharge due to event sensing 

(µAh) 

 Average discharge (µAh) 

 Average current consumption (A) 

 Average power consumption (W) 

Table 10 : ED power consumption parameters 

The average discharge is only a slight modification of the duty cycle expression ( 32 ) and is given by 

 ( 34 ) 

 

Similarly the average current that is drawn is given by 



44 

 

 ( 35 ) 

 

Assuming the theoretical discharge model specified in ( 18 ), the average power consumption is given 

by 

  ( 36 ) 

 

5.3.5 Specifying the AP and the PM: 

The AP and the PM parts of the AWSSN are not systematically specified as part of this