computer programming MATLAB

Low-Cost Short –Range Wireless Optical FSK Modem for Swimmers Feedback

Rabee M. Hagem1, David V. Thiel1,2*, Steven G. O’Keefe1

Thomas Fickenscher3

Andrew Wixted1,2

3Chair, High-Frequency Engineering, Helmut Schmidt

1Centre for Wireless Monitoring and Applications

University

2Centre for Excellence in Applied Sports Research

University of the Federal Armed Forces

Queensland Academy of Sport

Hamburg, Germany

Griffith University

Abstract —This paper reports 3 axis accelerometer data transfer over a one meter underwater path at 10 cm depth using a 2400 bps optical wireless frequency shift keying (FSK) at very low frequency (VLF). The modulation frequencies used were 10 and 12 KHz. The prototype modem was designed and implemented for real time feedback for swimmers in the pool. The optical transmitter included an accelerometer unit with a microcontroller, the modulator and a detector circuit based on an integrated detector preamplifier (IDP). The cost of the components for the optical transmitter and receiver was less than AU$25. Range experiments were performed in air and underwater, with and without bubbles. The received data was error free for 1.3 m in air and for more than 1.1 m underwater without bubbles. The underwater range decreased to 70 cm with bubbles. The availability of the link between the wrist and head of a swimmer was approximately 50% and varied with the position of the wrist. This enables stroke rate data to be presented to the swimmer via a goggle mounted display.

I. INTRODUCTION

The evaluation of swimmers can be performed wirelessly using a small portable accelerometer/gyroscope unit with data capture. Post processing allows interpretation of the swimming data [1]. No previous work has been reported for optical real time swimmer feedback. In training and longer swim events, feedback to the swimmer using sensors can improve training and performance by pre-setting the stroke rate and lap times and providing the swimmer with visual information about their current performance. The challenge is to achieve sufficient communication distance underwater between the sensor unit and a display unit mounted on the goggles.

II. LITERATURE REVIEW

Wireless communications between motion sensors placed on various parts of the body of a swimmer can be used to provide real time feedback through a heads-up display on the swimmer’s goggles. A wrist-mounted accelerometer can provide data such as stroke rate and lap time which can be used to improve compliance with swimming strategies and training regimes. The communications system between the wrist and the head must achieve a maximum distance of

approximately 1 m. Radio frequency suffers from severe attenuation in water and the antenna size is relatively large. Acoustic communications has the disadvantages of relative low speed and multipath problems. An optical wireless link can provide a relatively high speed data rate with low attenuation in the visible part of the spectrum. In particular green light has the minimum attenuation through clear, still water [2]. The design goals of the optical system in this paper were low cost, short range and a low data rate. FSK with a VLF carrier frequency at 10 and 12 KHz were designed and implemented.

High power consumption and cost were reported in most optical wireless communications systems reported in the literature. Anguita et al [3] developed a point to point optical wireless transceiver system based on using a blue LED as transmitter and a photodiode as a receiver. A single board node was used in an underwater wireless sensor network (UWSN) for communications between the sensor nodes. Two Spartan-3 boards were used. A transmission distance of 1.8 m was achieved with a data rate of 100 kbps.

Vasilescu et al [4] presented system hardware and software for underwater wireless sensor networks using a mixture of optical and acoustic communications to monitor coral reefs and fisheries. The cost of an optical communication board was about $50 per node while the acoustic modem was about $3000 per node. For the optical node, the transmitter was Luxeon 5 LXHL-PM02 with 532 nm green LED with about 700 mW radiated power while consuming 6 W of the input power. The receiver was a high speed PIN photodiode PDB -C156 with 8 mm2 surface area. The range achieved for the optical system was about 2.2 m point to point with a cone of 30 degree with data rate of 320 kbps and 8 m with a lens to concentrate the light beam.

Lu et al [5] designed and implemented an underwater optical wireless communication system with a range of between 5 to 10 m. Inexpensive components were used for the implementation of the communication system with complex detection algorithms including signal detection and clock synchronization. The cost of this system was below

$ 15. The LED was the RL5-G13008 Super-Green LED with a 520 nm wavelength and the maximum operation power was 12 mW. The photodiode was a Silonex SLD-70 BG2A with a maximum sensitivity at wavelength 550 nm and a sensitivity spectral range from 400 nm to 700 nm. In addition, a BG filter for infrared rejection was included with the photodiode. The radiation beam half angle for the LED and the photodiode were 45 and 60 degrees respectively. The experimental results showed that for 7 m, the detection probability was 100% while at 10 m this reduced to 80%. The data rate was restricted by the digital signal processing (DSP) board and was approximately 310 bps.

Schill et al [6] designed a small size optical communication transceiver for a swarm of submersible robots. The combination of the IrDA physical layer with a 3 W high power green and blue LED in the visible spectrum was introduced. The transmitter was a Luxeon III Emitter and the receiver was a SLD-70BG2A which is sensitive to the wavelength range 400-700 nm. The IRDA chip MCP2120 which includes encoder and decoder was used at the transmitter and the receiver, while a MAX3120 chip was used for amplification and filtering. The communications was point to point and the cost of the transmitter was about AU$45 per unit. An air and underwater experiment was performed to assess the viability of the link at different wavelengths of optical radiation. The maximum range achieved in air with the cyan emitter was 2.02 m followed by blue 1.71 m and green with 1.49 m. In water the range was reduced to 1.7 m.

The optical system presented in this paper will be used to give real time feedback to a swimmer about swim performance data including stroke rate and lap times.

III. SYSTEM DESIGN

An underwater optical communications link budget is dependent on the range, the attenuation through the water, the orientation and directivity of the transmitter, the orientation and directivity of the optical detectors, the transmitted power, the receiver sensitivity and the effect of the ambient light in the pool [2].

The optical transmitter was a superflux green LED (λ = 520 nm) with a +35o divergence angle and a 9500 mcd light intensity which is equivalent to a radiated power of 1.5 mW [2]. XR-2206 was used as the FSK modulator at the transmitter. The optical detector (transimpedance amplifier with a pin photodetector) had a 0.3AW-1 responsivity at λ = 520 nm, and provides an output voltage proportional to the incident optical power. XR-2211 was used as a phase locked loop (PLL) and FSK demodulator at the receiver. The receiver optical filter was the cokin P004 centered on 510 nm in order to reduce the effect of the ambient light [2]. The wireless sensor used to generate the acceleration data was the nCore 2.0 designed by Davey et al [1]. Some hardware and software modification for this sensor were required in order to connect it to the optical system.

The FSK optical link with a 3 axis accelerometer running at 50 samples per second was designed and tested in the pool. In water without bubbles, the distance achieved was greater than 1.1 m and in bubbled water this distance decreased to 70 cm. This range is sufficient for communications between the wrist and the head of a swimmer allowing an optical display of the processed data using a multi-colored LED mounted in the goggles. Fig. 1 shows the block diagram for the transmitter and fig. 2 shows the block diagram of the receiver.

Figure 1. Block diagram of FSK optical transmitter.

Figure 2. Block diagram of FSK receiver circuit. The RGB LED is attached to the swimmer goggles.

IV. FREESTYLE SWIM CYCLE

Typically the stroke patterns for the freestyle swim stroke are described in terms of six phases. Fig. 3 shows a side view for these phases as determined by the position of the hand. The first phase from 1-2 is the entry and stretch, the second is 2-3 phase called downsweep to catch. The third phase 3 is called catch and from 3-4 this is called the insweep. The next phase is the upsweep from 4-5. The phase between positions 5-6 is called the release and exit [7].

6

2 1

5

3

Figure 3. Side view of the phases of the stroke in freestyle swimming.

V. EXPERIMENTAL DESIGN AND RESULTS

A number of range measurements were undertaken in order to characterize the optical link before attaching the circuit to a swimmer. The first experiment was conducted in air and the distance achieved was greater than 1.3 m. The second experiment was conducted in still water (i.e. clear water without bubbles) with the transmitter and receiver both located 10 cm below the water surface. The effect of total internal reflection on the propagation path was clearly evident in the results. This experiment achieved a link distance greater than 1.1 m. Fig. 4 shows the circuit and calibrated support frame at the side of a swimming pool.

Figure 4. The experimental measurement system showing the rigid mounting frame in a swimming pool used for range measurements.

Range measurements in water with intense aeration were conducted in a jet spa. This bubbled water experiment revealed that the range decreased to 70 cm [2]. In this case the effect of total internal reflection from the water surface is not significant as the water surface is highly perturbed. This range is sufficient for communications between the wrist and the head of a swimmer. It was thought that the bubble density created by a swimmer would be much less than that observed in the spa and so the attenuation along the propagation path would be smaller. The underwater observation of a freestyle swimmer shows bubble formation, but the bubble density and bubble size distribution is very dependent on the speed and style of the swimmer.

The FSK VLF optical circuit was tested on a swimmer in air and in water to check the availability of optical link in different positions for the swimmer’s hand. The swimmer was asked to lie face -down on the side of a bench and move his head and arms in a manner which resembled freestyle swimming (sometimes referred to as dry-land swimming). Fig. 5 shows the y-axis of the acceleration data sent and received in real time in air from the optical transmitter attached to a swimmers hand to the optical receiver attached to his/her head. The acceleration data was recorded in the wrist mounted device and also in the head mounted device. The two recordings were matched to deduce the periods of link failure. This is plotted as zero in the dashed (red) line in Figure 5. The acceleration data clearly shows the variation in acceleration due to the stroke cycle. This characteristic variation in the acceleration can be used to determine the time between successive strokes.

The percentage of time when the data was received was calculated to be approximately 50% for this dry-land swimming situation. Table 1 shows the percentage of received data for different positions of the stroke in air.

Figure 5. Real time transmitted (continuous line/blue) and received acceleration data (dashed line/red) from the wrist for dry-land freestyle swimming. The propagation path was totally air. The acceleration is relative to the earth’s gravitational acceleration (g’s).

TABLE 1.Percentage of data received for different stroke positions given in Fig. 3.

Positions

Definition

Percentage of

data received

1-2

Entry and

100%

stretch

2-3

Downsweep to

100%

catch

3

Catch

0%

3-4

Insweep

0%

4-5

Upsweep

0%

5-6

Release and

100%

exit

A recreational swimmer was asked to swim freestyle in the pool. Fig. 6 shows the real time transmitted and received data for one acceleration axis on the wrist. The location of the transmitter on the wrist and the receiver on the head can be seen in Figure 7. The stroke characteristics are clearly evident but with more variability between strokes when compared to dry-land swimming. The optical path is broken more frequently when compared to the dry-land swimming. This is thought to be the result of the effect of the roughness of the

water surface and the possibility that at some times the transmitter and receiver are on opposite sides of the water surface. The overall result however was that the reliability of the communications link during freestyle swimming in the pool was approximately 50%.

Figure 6. Real time transmitted (continuous line/blue) and received acceleration data (dashed line/red) during freestyle swimming in a swimming pool. The acceleration is normalized to the earth’s gravitational acceleration (g’s).

Receiver

Transmitter

& sensor

Figure 7. Optical link trial in the swimming pool showing the swimmer wearing the transmitter and the receiver circuits.

VI. CONCLUSIONS

A low cost, short range optical wireless communications system using a green LED transmitter and IDP was designed and implemented based on FSK modulation with VLF carrier signal at 10 and 12 KHz. The optical link was tested and the stroke phases for freestyle swimming were investigated in order to check the percentage of received data for different arm positions. The results showed that the link was error free for approximately 50% of the time.

The prototype system described can be improved through a reduction in the size of both the transmitter and the receiver. Future work will be directed towards the design of the goggles feedback system in order to give a real time feedback to a swimmer. An investigation of the communications from a swimmer to pool side is important to allow interactions from the coach to the swimmer. The deployment of more than one movement sensor located in different places on the body (eg wrist, sacrum and ankle etc) requires a network of sensors. This will give additional information about the swimmer coordination, movement and speed. The optical link is suitable for a body-centric wireless sensor network.

ACKNOWLEDGMENTS

This work was conducted as part of Rabee Hagem’s Ph.D. program. He is supported by the MHED scholarship granted by Iraqi government. The authors also wish to thank the Queensland Academy of Sport for the facilities provided for the various experiments, and Dr. Daniel A. James for helping during these tests. This work has been supported by a research grant from the Australian Research Council. This research was conducted under Griffith University Ethics Protocol number ENG 05 10 HREC.

REFERENCES

• N. Davey, D. James, A. Wixted, Y. Ohgi, "A low cost self contained platform for human motion analysis," in The Impact of Technology on Sport II, F. K. Fuss, et al., Eds., London: Taylor & Francis, 2008, pp. 101-111.

• R. Hagem, D. Thiel, S. O'Keefe, T. Fickenscher, "The effect of air bubbles on an underwater optical communications system for wireless sensor network applications", Microwave & Optical Tech. Letters. submitted, 2011.

• D. Anguita, D. Brizzolara, G. Parodi "Building an Underwater Wireless Sensor Network Based on Optical: Communication: Research Challenges and Current Results," in Sensor Technologies and Applications, 2009. SENSORCOMM '09. Third International Conference on Sensor Technologies and Applications, 2009, pp. 476-479.

• I. Vasilescu, K. Kotay, D. Rus, M. Dunbabin, P. Corke "Data collection, storage and retrieval with an underwater sensor network," presented at the Proceedings of the 3rd ACM international conference on embedded networked sensor systems, San Diego, California, USA, 2005.

• F. Lu, S. Lee, J. Mounzer, C. Schurgers "Low-cost medium-range optical underwater modem: short paper," presented at the Proceedings of the Fourth ACM International Workshop on UnderWater Networks, Berkeley, California, 2009.

[6] F. Schill, U. Zimmer, J. Trumpf "Visible Spectrum Optical Communication and Distance Sensing for Underwater Applications,"

Proc. Australasian Conf. Robotics & Automation, 2004.

• E. W. Maglischo, Swimming fastest, Human Kinetics: Champaign, IL, 2003.