Self Contained Adaptable Optical Wireless Communications System ...

3144

IEEE SENSORS JOURNAL, VOL. 13, NO. 8, AUGUST 2013

Self Contained Adaptable Optical Wireless Communications System for Stroke Rate During Swimming Rabee M. Hagem, Steven G. O’Keefe, Thomas Fickenscher, and David Victor Thiel

Index Terms— Underwater optical wireless communication, real time swimmers feedback, visible light communication, stroke rate, accelerometer.

on the goggles allows visual signals to be given to the athlete during swimming. As the maximum distance between a swimmer’s wrist and goggles is approximately 1m the communications system must achieve this distance or more. The range varies because of the movement of the swimmer’s arms and head. The paper is organized as follows. In Section II, the importance of real time feedback to a swimmer is discussed as motivation for the work. Section III describes some of the design challenges encountered for such a system. Most previous underwater communication does not require a consideration of surface reflection or the effect of air bubbles on the transmission path. Section IV gives a brief review of some related work in swimming. The hardware electronics and the modulation techniques are discussed in Section V. The algorithms to determine the stroke rate and goggles display decision are presented in Section VI. The stroke phases for the free style swimming and the experimental results are discussed in Section VII. Finally, in sections VIII and IX we draw conclusions and discuss the direction of future work.

I. I NTRODUCTION

II. W HY IS R EAL T IME F EEDBACK I MPORTANT TO A S WIMMER ?

Abstract— The velocity of a swimmer can be determined from the stroke rate and the stroke length or by integrating the forward acceleration. In competitive swimming, these parameters are very important for race planning. This paper presents a wrist mounted accelerometer and optical wireless communications to display goggles to give real time feedback to a swimmer during swimming. The system data rate is 2.4 kbps ON-OFF keying modulation for the optical wireless signal. The system uses visible light communication in the green-blue wavelength. Design challenges include interference from bubbles and strong background light. The final device is low cost with low power consumption and small size. Intra-stroke transmit times are scheduled using the acceleration sensor data. Experiments are conducted in air and under water for this system to optimize the link availability. Algorithms for finding the absolute maximum of the y-axis acceleration for each stroke cycle and the goggles display decision are implemented at the transmitter and the receiver, respectively. Hardware, software, and implementation modifications to improve the system are successfully tested.

A

THLETE monitoring using movement sensors can provide important information. For example swimming performance can be improved and injury recovery and training progress tracked using a motion sensor system. Most swimming analysis based on sensors requires post processing of the swimming data after finishing the swim. This paper describes a new system to provide a swimmer with real time feedback to adjust stroke rate during swimming. A wrist-mounted accelerometer with a communications link to a receiver located

Manuscript received March 3, 2013; revised April 4, 2013; accepted May 7, 2013. Date of publication May 16, 2013; date of current version July 10, 2013. This work was supported in part by the MHED scholarship granted by Iraqi government, the Australian Research Council, and the Griffith University Ethics protocol under Grant ENG 05 10 HREC. This is an expanded paper from the IEEE SENSORS 2012 Conference. The associate editor coordinating the review of this paper and approving it for publication was Dr. Francis P. Hindle. R. M. Hagem, S. O’Keefe, and D. V. Thiel are with Centre for Wireless Monitoring and Applications, School of Engineering, Griffith University, Nathan 4111, Australia (e-mail: [email protected]; [email protected]; [email protected]). T. Fickenscher is with the Department of Electrical Engineering, Helmut Schmidt University, University of the Federal Armed Force Hamburg, Hamburg 22043, Germany (e-mail: [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JSEN.2013.2262933

Recreational and competitive athletes are becoming more reliant on the use of technology in order to improve their performance. This is supported by the evolution of technology including wireless communications, microcontrollers and wireless for smart sensors. In sports such as swimming, the performance is strongly correlated with technique and swimmers make adjustment to reach the optimal performance. Currently, swimmers depend on coaches and video to improve their technique. Wireless motion sensors provide information for post processing. However, there is no real time feedback in such systems. Real time feedback is very useful in swimming because some aspects of swimming are opposite to human intuition for example, a swimmer can put in extra effort but the results can be a slower velocity because of poor technique. A real time system can provide swimmers with information during swimming in order to adjust instantly and they can be informed of the effect of their changes according to the feedback. In addition, from the stroke rate and the stroke length, the velocity of a swimmer can be determined. Therefore, monitoring stroke rate can provide information to reach the optimal performance. Such a system can be used during training in order for a swimmer to find the most effective techniques to improve performance [1].

1530-437X/$31.00 © 2013 IEEE

HAGEM et al.: SELF CONTAINED ADAPTABLE OPTICAL WIRELESS COMMUNICATIONS SYSTEM

3145

III. C HALLENGES AND C ONDITIONS Wireless communications faces challenges in difficult environments such as water. For swimming applications, the transmitter and the receiver experience two different environments, air and water during a single stroke. This is because the hand and the head of a swimmer are sometimes in air and sometimes in water. Therefore, the transmitter and the receiver should perform well in both environments. These challenges for the system presented in this section include: A. Communication Type There are three methods of wireless communications including: radio frequency (RF) by sending electromagnetic waves, acoustic communication by sending sound waves and optical wireless communication by sending light waves. Radio waves suffer from a high attenuation in water [2]–[5] and at low frequencies require large antennas and high transmission power. This is not practical to use underwater [6]. The underwater sound propagation is influenced by high noise levels, propagation delay, path loss, multi-path, high bit error rate and limited bandwidth which depends on range and frequency [6]. In addition, acoustic waves could distress swimmers. Therefore, acoustic communication is not the best to use underwater with applications that need high data rates [7]. Underwater optical wireless communication is a feasible solution because of the high data rate and high bandwidth. However, absorption and scattering have an effect on the optical propagation path [7]–[9].

Trade off between the water attenuation and the spectral sensitivity of the receiver is important. Fig. 1 shows the absorption coefficient for pure water [12].

B. Optical Transmitter

E. Modulation Type

An optical light source is characterized by viewing angle, wavelength, luminous intensity and current. A Light Emitting Diode (LED) was chosen because it radiates over a wide angle range. This is important because of the movement of the transmitter. The LED converts an electrical signal to an optical signal and has a low transmit power, (less hazard than a laser) and can be used for indoor and outdoor short distance applications [10].

Intensity modulation and direct detection have been used in the underwater environment. Modeling and simulation results for different modulation techniques for underwater optical wireless communications were reported in [13]. The techniques are usually classified into two types (a) Coherent detection of base band signal such as frequency shift keying (FSK) and Phase shift keying (PSK) or (b) Non-coherent detection of base band signal when no information about the phase is available such as amplitude shift keying (ASK), on-off keying (OOK) and (PPM). It has been proven in [13] that PPM is better for low power underwater applications while PSK gives good performance in terms of bandwidth and error performance but with poor power efficiency. In addition, OOK and PPM are usually used in a simple direct detection system with lower complexity while FSK and PSK have high complexity. FSK and OOK systems were the compromise chosen for this system. Table I shows the difference between different modulation techniques.

C. Optical Receiver Photodetector technologies include photoresisters, phototransistors, avalanche photodidodes and pin photodiodes. A photodiode converts an optical signal to an electrical signal or current. The factors taken into account for choosing a suitable photodiode include sensitivity, dynamic range, size and simplicity. The p.i.n. photodiode was the most suitable choice for the swimmers feedback system because of the linearity, size and reliability [11].

Fig. 1.

The absorption coefficient of pure water [12]. TABLE I

C OMPARISON B ETWEEN D IFFERENT M ODULATION T ECHNIQUES W HERE P R EPRESENTS THE S MALLEST P ULSE W IDTH [13] OOK

FSK

DPSK

4-PPM

8-PPM

Transmit power

Middle

Higher

Highest

Low

Lowest

Maximum rate

1/(2P)

1/(2P)

1/(2P)

(1/2P)

(3/8P)

Modulation complexity

Low

Higher

Highest

Lower

Lower

F. Bubbles Effect D. Optical Wavelength Choosing the most suitable wavelength is important for the optical wireless system that performs in water. There is a strong relationship between wavelength and water attenuation. The minimum absorption and scattering coefficients occurs in the blue-green wavelength range between (450–570) nm.

As swimmers are located relatively close to the surface of the water, arm and leg movements can generate a significant number of air bubbles. Total internal reflection from the surface and attenuation due to bubbles may be a major problem for communications. The effect of air bubbles on the optical signal was reported previously [14]. Experiments were

3146


Water Surface Optical Receiver Optical Transmitter

Fig. 2. The cloud of bubbles generated by a swimmer during swimming in freestyle are evident near head, arms and legs. Fig. 3. Implementing the wireless optical system on a swimmer’s body. The optical transmitter on the wrist and the optical receiver on the goggles.

conducted in air and water with and without bubbles. The results showed that the attenuation coefficient increased from 0.09 to 2.9 m−1 when a spa jet was turned on. Fig. 2 shows free style swimming with the cloud of bubbles generated by arms and legs clearly visible. G. Strong Background Light Strong background light from the lighting system and the sun causes optical interference and can saturate the receiver. One of the effective solutions is to use a narrow band pass filter (NBPF) in order to reduce the effect of the background light by allowing only the desired wavelength to proceed the detector. H. Link Budget The communication link between the transmitter and the receiver was sometimes line of sight (LOS) and sometimes non line of sight (NLOS). All calculations and the simulations related to the link budget were discussed by the authors [15] based on using two different LEDs in order to check the success of a communication link and eventually recommending the suitable components. The higher the power received and the greater the receiver sensitivity, the higher will be the reliability of the communications system at the expense of battery power [16]. The simulation results showed that the HPBW (Half Power Beamwidth) and the bubbles were important factors for the communication link availability. The superflux LED and the opt101 which is an IDP (Integrated detector preamplifier) achieved 1 m communication link. This combination was reported in [17] for a low cost optical link. I. Electronics Requirements Small size, water proof and battery powered are very important requirements for this system. The small size makes it easy to wear by a swimmer. Small batteries reduce the weight so that the swimmer’s performance is not affected. Inexpensive devices are required in this system. The power consumption was also important to make the system run as long as possible. There are some additional challenges considered such as: swim speed, position of the transmitter and the receiver on the swimmer’s body which means mounting accuracy for the optical transmitter and receiver (position as well as look angle),

swim style and the individual swimmer. Most of these challenges can be solved by making an automated self calibration system (smart sensor system) which determines the optimum link availability depending on the swimmer’s arm position based on the acceleration data. This can be achieved by proper design of an optical transceiver including an embedded algorithm for self calibration. IV. R ELATED S WIM S YSTEMS The basic idea of the real time swimmers feedback system is shown in fig. 3. Most of the previous work in swim monitoring depends on video analysis or sensors (including accelerometer and gyroscope). However, all depend on processing the data after a swimmer has finished swimming [18]. There is no real time feedback [19], [20]. A wearable assistance device for swimmers known as swimmaster [21] has two independent parts. The first was acceleration sensors with a microcontroller and memory for data recording and the second was to give a different pre-programmed feedback (audio, tactile and visual). The number of strokes, body orientation and body balance as well as swimmer velocity were extracted after a swimmer finished swimming. The system was based on three different signals (i) continuous signal, (ii) two short signal events and (iii) four different signals to investigate the reaction of swimmers. The swimmer was required to change swimming according to these signals. In [1] a feedback system for swimmers was designed as a wearable device. Forward speed and body orientation information were provided to the swimmer. The system was based on a three-axis accelerometer, optionally three axis gyroscopes, memory and a microcontroller to process the data of a swimmer’s performance and display feedback. The feedback could be a digital display in the goggle or an audio response via an earpiece. The communication link was wire. In [22] a system to determine the average stroke interval and average right and left stroke acceleration was presented. The acceleration data was stored in the sensor board for transmission wirelessly to a computer after the swimmer left the pool. In [23] the roll and pitch angles were used to recognize the type of stroke performed based on an inertial sensor mounted on the swimmer’s goggles. The sensor


contains a three axis accelerometer, 2.4 GHz RF transmitter, 2 MB on board memory and it was water proof. The sample rate was 50 Hz and the sensor either saved data on board or sent it in real time to a base station. There was feedback for the coach because the system has the ability of sending real time data to a base station outside the pool. In [24] three different feedback systems were evaluated; visual, audio and haptic. Experiments were conducted to check the recognition rate and the reaction time with these three different feedbacks. The visual and the haptic changes were recognized 70%–100% of the time and the reaction of the subjects was in the range of 1.25 to 2.25 s while the response for audio feedback was less than 70%. The reaction time was twice as long compared to visual or haptic feedback. These results showed that the audio system is not appropriate while swimming. In summary, no portable system has been reported that gives real time feedback to the swimmers while swimming. In conclusion, there is no optical real time feedback in all the systems mentioned. This paper reports on the use of an LED in an optical communications system.

3147

Fig. 4. Electronics for the visible light communication system showing the optical transmitter (Tx), optical receiver (Rx), the nCore 2.0 wireless sensor, the goggles with the RGB LED implemented and the LEDs ring [25].

V. S YSTEM D ESIGN V ISIBLE L IGHT C OMMUNICATIONS A green superflux LEDs at 520 nm wavelength was used as an optical transmitter with a 70° viewing angle and 9500 mcd luminous intensity. XR-2206 was used as the FSK (Frequency Shift Keying) modulator at the transmitter. The optical detector (transimpedance amplifier with a pin photodetector) had a responsivity of 0.3AW−1 at 520 nm, and provides an output current 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. All calculations related to the smart sensor and the LED ring are reported in [25]. Some important factors were considered in the design of the omnidirectional system based on the ring. These factors are the diameter of the ring supporting the LEDs and the field of view of the LEDs. Eight green LEDs were located on a circular ring of 7.62 cm diameter and length of 24 cm. The wireless sensor used to generate the acceleration data was the nCore 2.0. This device includes 3-axis accelerometer, a microcontroller with a memory in order to save the movement data that can be downloaded later for further analysis by a coach or a swimming expert. Fig. 4 shows the visible light communication system including the optical transmitter, optical receiver, the nCore 2.0 wireless sensor, the goggles with the RGB (Red, Green, Blue) LED implemented and the LED ring. The goggles LED provides minimal information (red = too slow, green = correct pace and blue = too fast). The swimmer needs only to response to the color change. It is difficult to concentrate on swimming technique while continuously reading a more complex goggles display. This figure shows the positions of the eight LEDs on the ring. The LEDs ring was worn by a swimmer and LED1 was beneath the thumb. Fig. 5 and 6 show the block diagram for the transmitter and the receiver respectively. Fig. 7 shows testing the system in a swimming pool. The mounting of the wireless sensor is

Fig. 5.

Fig. 6.

Block diagram of the transmitter.

Block diagram of the receiver.

shown in fig. 8(a) including the orientation of the three axis accelerometer. The X-axis represents the direction from the wrist to elbow joint, distal-proximal direction. Y-axis represents the direction from left hand little finger to thumb, defined as ulnar-radial direction. Z-axis represents the direction from palm to the back of hand, planar-dosal direction [26]. The position of LED4 and LED6 are shown in fig. 8(b). LED1 is beneath the thumb. VI. S TROKE R ATE AND E MBEDDED A LGORITHMS Important parameters such as stroke rate, stroke length and swimmer velocity can be extracted from the acceleration data of the wrist provided by the swimmer. Two algorithms to find the stroke rate were considered. The first was the absolute

3148


Fig. 7. Testing the optical wireless system for real time swimmers feedback [25].

Fig. 10. Receiver flowchart chart showing the embedded algorithm for the real time swimmers feedback.

Fig. 8. (a) The wireless sensor orientation on a swimmer’s wrist. (b) The positions of LED4 and LED6 on the ring.

the maximum acceleration, and the time difference between strokes [27]. The stroke rate can be readily obtained from these data. The minimum peaks of the y axis are most likely caused by the impact force of the left hand entry. It repeats cyclically. The raw acceleration data of this axis was chosen because every stroke duration can be calculated exactly by looking to the minimum or maximum peaks [26]. This is similar to running systems when the foot impact dominates the accelerometer output. Both algorithms were implemented on the microcontroller of the nCore 2.0 wireless sensor. Fig. 10 shows the receiver embedded algorithm for the real time swimmers feedback. Fig. 11 shows the flow chart for the stroke time difference algorithm calculation at the transmitter side. VII. E XPERIMENTS

Fig. 9. The 3 axis wrist acceleration data and the locations of the maximum acceleration after applying the maximum algorithm on the y-axis [27]. The sample rate is 100 samples per second.

maximum or minimum algorithm of the y-axis acceleration data for the stroke cycle which can be applied on the raw acceleration data [26]. The maximum occurs when the leading arm hits the water surface. The second is the zero crossing detection algorithm which needs calibration and filtering for the raw acceleration data in order to detect the one arm stroke duration. This must be converted to stroke rate by checking the data points crossing the zero gravity axis. In this application the maximum detection algorithm was implemented. Fig. 9 shows the 3 axis acceleration data, the position of

Free-style swimming is the most common and fastest swim style. Fig. 12 shows a side view for the six phases of the arm movements: Entry and stretch (1-2) when a swimmer enters his hand into the water and stretches his arm forward. The downsweep to catch (2-3) is when the swimmer’s hand moves downward. This happens by extending the shoulder joint and a slight flexion of the elbow joint. The result is a curvilinear downsweep motion. Catch (3) is when the elbow rise up above the hand. The Insweep (3-4) is generated by extending swimmer’s shoulder and flexing his elbow joint with body roll. The swimmer’s hand moves to the midline of his body. During this phase the palm gradually rotates from out and back to in and up. The further extension in the shoulder and elbow joint causes the upsweep (4-5). The swimmer must change his hand pitch angle properly in order to produce sufficient propulsive force. The release and exit (5-6) is when the swimmer removes his hand from water [26], [28]. The availability of the link depends on the wrist and head position and orientation. The beginning of the insweep motion causes the peaks in the y axis acceleration [26]. The y axis acceleration data peaks occur in the insweep phase. The stroke phase information can be determined from the x axis and z axis acceleration data [26].


3149

Fig. 13. Histogram of the two fast and the two slow stroke data set. The two population are distinct (P < 0.01).

TABLE II P ERCENTAGE OF R ECEIVED O PTICAL D ATA W ITH T URNING ON D IFFERENT LED S IN A IR [25]

Fig. 11. Transmitter flowchart showing the stroke rate determination embedded algorithm for real time swimmers feedback where W is the window size chosen for the number of y-axis acceleration data points in order to find the stroke rate.

Fig. 12. Side view for the six phases of the stoke in free style swimming [17].

The best colours to use for the feedback system and the average stroke rate categories were investigated experimentally [27]. A recreational swimmer was asked to swim free style in a 50 m pool by doing fast strokes and slow strokes.

LED Number

1

2

3

4

5

6

7

8

Stroke number

22

20

20

20

21

20

19

20

% Received data

4.5

25

25

55

33

95

0

20

The stroke length for this swimmer was between 2.38-2.5 m/cycle while SR was between 1.76-2.09 s/stroke cycles. The time difference between strokes was between 2.104-2.23 s in slow swimming and for the same swimmer during fast swimming the average time was in the range of 1.84-1.954 s. These values were used in the decision algorithm at the receiver [27]. The transmitter was activated by the time difference signal saved in the on board memory and sent optically to the receiver to activate the RGB light display in the goggles. A t-test was done for the two slow and the two fast stroke data sets in order to establish the probability of a relationship between the data sets. The probability value P was less than 0.01 for all the data showing that the two data sets are significantly different. Fig. 13 shows the histogram for the two fast and two slow strokes experiments. These values are different for each swimmer and depend on the swimming style and level of exhaustion. Therefore before using the system a swimmer should be calibrated by swimming slow and fast laps and then according to the collected data the calibration bounds applied. For this swimmer the bound was between 1.76-2.09 s/stroke cycles. Experiments were conducted in air in order to check the effect of each LED in the LED ring in freestyle swimming on the percentage of received data during a whole stroke cycle. The swimmer was asked to replicate the swimming action with each LED turned on separately. The purpose of these experiments was to investigate the contribution of each LED separately to see which LED or LEDs are best to use. Table II shows the results. Another experiment was done for different swimming style and when all the LEDs were on. Table III presents the results.

3150


TABLE III P ERCENTAGE OF R ECEIVED O PTICAL D ATA FOR D IFFERENT S WIMMING S TYLE IN A IR AND IN WATER % of Received Data

Breast Stroke

% of Received Data

Backstroke

% of Received Data

Butterfly

% of Received Data

1-2 Entry and stretch

100

1-2 Outsweep

0

1-2 First downsweep

100

1-2 Entry and stretch

66.14

2-3 Downsweep to catch

100

2-3 Catch

100

2-3 Catch and first upsweep

0

2-3 Outsweep

100

3 Catch

100

3 Insweep

0

3-4 second downsweep

0

3 Catch

0

3-4 Insweep

0

3-4 Release and recovery

100

4-5 Second upsweep

98.42

3-4 Insweep

0

4-5 Upsweep

0

5-6 Release and exit

39.3

4-5 Upsweep

66.67


100

Free Styles Stroke (Front Crawl Stroke)


100

Continuous in air

68.5

Continuous in air

49.6

Continuous in air

28.5

Continuous in air

42.85

Continuous in water

54.3

Continuous in water

60.3

Continuous in water

45.25

Continuous in water

50.5

VIII. R ESULTS AND D ISCUSSION Experiments conducted include finding the stroke rate variations for an individual swimmer. The results show that the time difference between strokes for slow swimming was between 2.104-2.23 s and for fast swimming the range was between 1.84-1.954 s and these values were used in the embedded algorithms. The feedback experiments showed that red, blue and green were the best colors to use in the feedback system. The link availability experiments by turning on individual LED each time showed that LED4 and LED6 have the major contribution to the received optical data for the free style experiment. Note that this swimmer used thumb first entry into the water during the insweep phase. The link availability in water was highest for the breast stroke followed by free style swimming, butterfly and back stroke. IX. C ONCLUSION The wearable data acquisition, processing and feedback system was designed, implemented and tested based on visible light communication in order to give a real time feedback to a swimmer during swimming. Acceleration data is used for stroke rate determination and optimum transmission time in the stroke cycle. The time differences between strokes were sent to the receiver in order to keep the swimmer to a consistent stroke rate. Different challenges related to this system were discussed including the bubble effect and strong back ground light. Experiments were conducted in order to find the optimum link based on the data transferred by checking the best LED or LEDs which contribute more in the optical link and for different swimming style. The results showed that LED4 (center of wrist), LED6 (outside top edge of wrist) contributed with 55% and 95% of link availability respectively. For the different stroke experiments, the results in water showed that the breast stroke swimming had generally a high link availability. In future work and for the implementation of the circuit, circuit in plastic (CiP) can

be used as environment and water proof technology. A time division multiplexer (TDM) will be applied which allows data to be sent from different sensors on the swimmer’s body such as wrist, sacrum, and foot which give a whole picture about the swimmers performance. This will provide important information to the coaches. The transmission of real time data to a coach on pool side will allow a real time interaction between swimmers and their coaches. R EFERENCES [1] X. Li, “Real-time swimming monitor,” U.S. Patent 0 030 482, Feb. 4, 2010. [2] D. Anguita, D. Brizzolara, and G. Parodi, “Building an underwater wireless sensor network based on optical: Communication: Research challenges and current results,” in Proc. 3rd Int. Conf. Sensor Technol. Appl., 2009, pp. 476–479. [3] L. Liu, S. Zhou, and J.-H. Cui, “Prospects and problems of wireless communication for underwater sensor networks,” Wireless Commun. Mobile Comput., vol. 8, no. 8, pp. 977–994, 2008. [4] X. Liu, X. Yu, and M. Sui, “Evaluation of underwater wireless optical communication link with pspice simulator,” in Proc. Wireless Commun., Netw. Mobile Comput., 2007, pp. 1004–1007. [5] F. Schill, U. R. Zimmer, and J. Trumpf, “Visible spectrum optical communication and distance sensing for underwater applications,” in Proc. Austral. Conf. Robot. Autom., 2004, pp. 1–8. [6] Y. Liu and X. Ge, “Underwater laser sensor network: A new approach for broadband communication in the underwater,” in Proc. 5th WSEAS Int. Conf. Telecommun. Informat., 2006, pp. 421–425. [7] S. Arnon, “Underwater optical wireless communication network,” Opt. Eng., vol. 49, no. 1, pp. 015001–015006, 2010. [8] S. Arnon and D. Kedar, “Non-line-of-sight underwater optical wireless communication network,” J. Opt. Soc. Amer. A, vol. 26, no. 3, pp. 530–539, 2009. [9] G. Baiden, Y. Bissiri, and A. Masoti, “Paving the way for a future underwater omni-directional wireless optical communication systems,” Ocean Eng., vol. 36, nos. 9-10, pp. 633–640, 2009. [10] A. Mahdy and J. S. Deogun, “Wireless optical communications: A survey,” in Proc. IEEE Wireless Commun. Netw. Conf., vol. 4. Mar. 2004, pp. 2399–2404. [11] H. Brundage, “Designing a wireless underwater optical communication system,” M.S. thesis, Dept. Mech. Eng., Massachusetts Institute of Technology, Cambridge, MA, USA, 2010. [12] R. C. Smith and K. S. Baker, “Optical properties of the clearest natural waters (200–800 nm),” Appl. Opt., vol. 20, no. 2, pp. 177–184, 1981.


[13] M. Sui, X. Yu, and F. Zhang, “The evaluation of modulation techniques for underwater wireless optical communications,” in Proc. Int. Conf. Commun. Softw. Netw., Feb. 2009, pp. 138–142. [14] R. M. Hagem, D. V. Thiel, S. G. O’Keefe, and T. Fickenscher, “The effect of air bubbles on an underwater optical communications system for wireless sensor network applications,” Microw. Opt. Technol. Lett., vol. 54, no. 3, pp. 729–732, 2012. [15] R. M. Hagem, D. V. Thiel, S. G. O’Keefe, and T. Fickenscher, “Optical wireless link budget calculations for real time swimmers feedback,” in Proc. Int. Conf. Comput. Commun. Eng., Jul. 2012, pp. 180–184. [16] J. Griffiths, Radio Wave Propagation and Antennas: An Introduction. Englewood Cliffs, NJ, USA: Prentice-Hall, 1987. [17] R. M. Hagem, D. V. Thiel, S. G. O’Keefe, A. Wixted, and T. Fickenscher, “Low-cost short-range wireless optical FSK modem for swimmers feedback,” in Proc. IEEE Sensors, Oct. 2011, pp. 258–261. [18] R. M. Hagem, D. V. Thiel, S. G. O’Keefe, and T. Fickenscher, “Optical wireless communication for real time swimmers feedback: A review,” in Proc. Int. Symp. Commun. Inf. Technol., Oct. 2012, pp. 1085–1090. [19] A. Stamm, D. A. James, R. M. Hagem, and D. V. Thiel, “Investigating arm symmetry in swimming using inertial sensors,” in Proc. IEEE Sensors, Oct. 2012, pp. 1–4. [20] N. Davey, D. James, A. Wixted, and Y. Ohgi, “A low cost self contained platform for human motion analysis,” in The Impact of Technology on Sport II, A. Subic, F. K. Fuss, and S. Ujihashi, Eds. London, U.K.: Taylor & Francis, 2008, pp. 101–111. [21] M. Bachlin, K. Forster, and G. Troster, “SwimMaster: A wearable assistant for swimmer,” in Proc. 11th Int. Conf. Ubiquitous Comput., 2009, pp. 215–224. [22] B. H. Khoo, B. K. J. Lee, S. M. N. A. Senanayake, and B. D. Wilson, “System for determining within-stroke variations of speed in swimming (SWiSS),” in Proc. IEEE/ASME Int. Conf. Adv. Intell. Mech., Jul. 2009, pp. 1927–1932. [23] J. Pansiot, B. Lo, and Y. Guang-Zhong, “Swimming stroke kinematic analysis with BSN,” in Proc. Int. Conf. Body Sensor Netw., 2010, pp. 153–158. [24] K. Forster, M. Bachlin, and G. Troster, “Non-interrupting user interfaces for electronic body-worn swim devices,” in Proc. 2nd Int. Conf. Pervas. Technol. Rel. Assist. Environ., 2009, pp. 1–4. [25] R. M. Hagem, D. V. Thiel, S. G. O’Keefe, N. Dahm, A. Stamm, and T. Fickenscher, “Smart optical wireless sensor for real time swimmers feedback,” in Proc. IEEE Sensors, Oct. 2012, pp. 1–4. [26] Y. Ohgi and M. Yasumura, “Analysis of stroke technique using acceleration sensor IC in freestyle swimming,” in Proc. Eng. Sport, 2002, pp. 503–511. [27] R. M. Hagem, D. V. Thiel, S. G. O’Keefe, and T. Fickenscher, “Real time swimmers’ feedback based on a smart infrared (SSIR) optical wireless sensor,” Electron. Lett., vol. 49, no. 5, pp. 340–341, Feb. 2013. [28] E. W. Maglischo, Swimming Fastest, 3rd ed. Champaign, IL, USA: Human Kinetics, 2003.

Rabee M. Hagem received the B.Sc. degree in electronics and communication engineering from Mosul University, Mosul, Iraq, in 1998, and the M.Sc. degree in electronic and communication in 2001. He is currently pursuing the Ph.D. degree in underwater optical wireless communication for real time swimmers feedback. He is a member of the Centre for Wireless Monitoring and Applications with Griffith University, Brisbane, Australia. He has been a Lecturer with the Computer Engineering Department, Mosul University, since 2003. He is supported by a scholarship from MHED in Iraq. He was a R&D Engineer for North Mosul Electricity Distribution Company in 2001 and 2002 and as a QA/QC Engineer for ABB company for extension of Mosul 400 Kv substation project in 2007. His current research interests include underwater optical wireless communication, design and implementation of smart sensors, embedded systems for athletes monitoring, and sport data analysis.

3151

Steven G. O’Keefe received the B.Sc. and Ph.D. degrees from Griffith University, Brisbane, Australia, in 1984 and 1996, and the B.Sc. (Hon.) and M.Sc. degrees from LaTrobe University, Melbourne, Australia, in 1986 and 1989, respectively. He is currently an Associate Professor, the Deputy Head with the Griffith School of Engineering, and the Deputy Director of the Centre for Wireless Monitoring and Applications. His current research interests include wireless sensor networks, compact and mobile antenna design, and dielectric resonator antenna structures.

Thomas Fickenscher received the Dr.-Ing. and Habilitation degrees from Helmut Schmidt University (HSU HH), Hamburg, Germany, in 1997 and 2005, respectively. In 1997, he joined Siemens AG, Munich, Germany, where he was involved in the design and development of receiver front-ends for cellular phones. From 1998 to 2000, he was with Lucent Technologies, Bell Labs, Murray Hill, NJ, USA, and the Department of Optics and High Speed Circuit Development, Nuremberg, Germany, where he was responsible for optical receiver front-end design. Since 2000, he has been the Head of the Laboratory for HighFrequency Engineering, HSU HH. In 2011, he was an Associate Professor with Griffith University, Brisbane, Australia.

David Victor Thiel received the degree in physics and applied mathematics from the University of Adelaide, Adelaide, Australia, and the Masters and Ph.D. degrees from James Cook University, Townsville, Australia. He is currently the Director of the Centre for Wireless Monitoring and Applications with Griffith University, Nathan, Australia. He has co-authored Switched Parasitic Antennas for Cellular Communications. His interests include mathematical optimization techniques for antenna design. He has published over 100 journal papers, more than 150 papers presented at international conferences, and he has coauthored more than nine patent applications. He was a co-inventor of the new RoHS and WEEE compliant electronics manufacturing technology called “circuits in plastic.” His current research interests include electromagnetic geophysics, sensor development, electronics systems design and manufacture, and antenna development for wireless sensor networks. He is a fellow of the Institution of Engineers, Australia. He is currently the Chair of the IEEE Wave Propagation Standards Committee and serves on the IEEE Antenna Standards Committee.