MOBILE OPTICAL WIRELESS SYSTEM FOR HEALTHCARE CONTINUOUS MONITORING USING IR TECHNOLOGY

31 st August 015. Vol.78. No.3 005-015 JATIT & LLS. All rights reserved. ISSN: 199-8645 www.jatit.org E-ISSN: 1817-3195 MOBILE OPTICAL WIRELESS SYSTEM FOR HEALTHCARE CONTINUOUS MONITORING USING IR TECHNOLOGY 1 E. A. ALYAN, S. A. ALJUNID, 3 M. S. ANUAR 1,,3 Centre of Excellence Advanced Counication Engineering School of Coputer and Counication Engineering (CoE ACE-SCCE) E-ail: 1 ead_alyan@yahoo.co, Universiti Malaysia Perlis (UniMAP) syedalwee@uniap.edu.y, 3 anuars@uniap.edu.y ABSTRACT In this paper, indoor optical wireless syste using Infrared (IR) technology is developed for healthcare onitoring. This to insure the reliability, the security and the obility of the patient during his/her stay in hospital. Various studies were perfored to develop a obile optical wireless syste but using celling bounce odel which considers just one reflection and the transitter has to be pointed to the ground which liits the patient s obility and oveents. In this research, we consider an optical source (LED) coupled with the edical sensor and can randoly change its position within the xy-diensions of the roo and within height varying fro 0 to 1.3 which is the average height of the edical sensor to be put on the body. While the receiver is located on the centre of the ceiling. The line of sight and diffuse propagation links are considered to investigate the capability of the developed syste to transit the health inforation of the patient to the receiver after undergoing ultiple reflections fro different walls and objects placed in the roo. Using MATLB siulation, the average received power, RMS delay spread, SNR and BER are calculated in order to evaluate the perforance and capability of the syste with LOS, first and second order reflections for all positions that patient can reach to. Keywords: Infrared; Optical wireless; Diffuse; LOS; RMS delay spread; Healthcare onitoring. 1. INTRODUCTION Wireless counication has been widely used in healthcare for patient continuous onitoring purpose in last couple of years. This to insure the reliability and the obility of the patient during his/her stay in hospital [1]. In addition, it increases the safety of the patient, provides fast response and rehabilitation, and enhances the efficiency of the hospital staff. The exiting technology is generally based on radio frequency syste (RF) which has several drawbacks when it is used in hospital environent. These drawbacks are ainly caused by the induced electroagnetic interference (EMI) which ay cause alfunction to edical equipent, resulting in isdiagnosis, istreatent and that ay lead to dangerous edication[,3]. Therefore, optical wireless based on infrared (IR) technology was introduced to overcoe this proble, since it has several advantages over RF technology. These advantages are iunity to EMI, cost effective, supporting high data rates, operating in free licensed spectru band, and security [4]. Optical wireless counication can be classified into two ain configurations when IR technology is considered in indoor environent. These two configurations are Line Of Sight (LOS) configuration and Diffuse configuration. LOS requires alignent between transitter and receiver. Therefore, it has better power efficiency and lower ultipath dispersion. Besides, it has drawback due to the oving object across the direct path which causes shadowing. Diffuse configuration does not require any alignent between the transitter and receiver, and instead, ake use of reflections fro walls, ceiling, and other reflectors. However, diffuse transission links are usually affected by ultipath dispersion (which causes pulse spread and significant ISI), poor power efficiency and higher collection aount of abient light noise at receiver part [5,6]. In this paper, wireless infrared technology (IR) is investigated in order to ensure the quality of service of the counication channel between the coupled transitter with the edical sensors which 353

31 st August 015. Vol.78. No.3 005-015 JATIT & LLS. All rights reserved. ISSN: 199-8645 www.jatit.org E-ISSN: 1817-3195 are placed on patient s body and the established base station on the ceiling of the hospital. Based on previous studies, first and second order reflections will largely contribute to the receiver optical power, unlike, the third and higher order reflections are ignored due to their sall contribution. In this study, we consider that the transitter can randoly change its position within the xydiensions of the roo and within height ranging between 0 to 1.3 which is the average height of the edical sensor to be put on the body. Fro the siulation analysis and before concluding the reliability of the developed obile onitoring syste. The average received power, RMS delay spread, SNR and BER are calculated to evaluate the perforance and capability of the syste with first and second order reflections for all positions that patient can reach. The data rate is set to be 10 Mb/s which is enough to transit the health inforation of a patient such as ECG, EMG, pulse inforation and etc. []. The paper is organized as follows: in part we describe the developed syste. We analyze the received power in part 3. In part 4 we calculate the root ean square delay spread. Part 5 presents the result and analysis of SNR and BER, before concluding in part 6.. SYSTEM DESCRIPTION In this research, LOS and diffuse propagation links are considered to investigate the capability of the syste to transit the health inforation of the patient to the receiver after undergoing ultiple reflections fro different walls and objects placed in the roo. As shown in Figure 1, we consider an epty roo with diensions of 3x3x3 3 and each wall surface was divided into a nuber of equalled size of reflection eleents with area of da and reflection coefficient ρ. The reflections fro windows, doors and objects are considered to be identical to the walls reflections []. We consider in our research that the patient will be freely oved within a volue of 3x3x1.3 3. The transitter part consists of an optical source (LED with eye safety) while the receiver part consists of a photo-detector which is placed on the centre of the ceiling with field of view (FOV) directed downward to the floor. Figure 1: Architecture Of Infrared Diffuse Channel For Patient Monitoring Syste 3. RECEIVED POWER OF A MOBILE EMITTER In this odel, we consider up to two reflections, which can lead to a reasonably accuracy for the ipulse response. Moreover, the third and higher order reflections are ignored due to their sall contribution to the overall ipulse response. The radiation intensity of the transitter with a generalized Labertain radiation can be odelled as: 1 R 0 ( ) = cos ( ), (1) Pt where is the Labertian order of eission which is related to the half power sei-angle and represented as [7]: = ln( ) ln(cos 1 ) ( ) The syste consists of LOS propagation link which represents the direct path between the transitter and receiver and its channel DC gain can be atheatically calculated as: 1 D d h(0) LOS x 0 A R T F ( )Tc( )... cos ( ) cos( ) 0 ψ ψc ψ ψc ( 3 ) 354

31 st August 015. Vol.78. No.3 005-015 JATIT & LLS. All rights reserved. ISSN: 199-8645 www.jatit.org E-ISSN: 1817-3195 where D d is the direct distance between the transitter and receiver. A R is the physical area of the photo-detector. T F ( ) is the gain of the optical filter. T c ( ) is the gain of the optical concentrator. is the angle of incidence and ψ is the reception angle. ψ c donates the width of the field of view at the receiver [4]. The optical concentrator T c ( ) can be given as n Tc( ) sin ( c) 0 n represents the refractive index. 0 ψ ψc ψ ψc Moreover, the channel DC gain for the first and second order reflections are given respectively as [7]: h(1) ref ( 1)( 1) eleent A ( ) ( ) cos ( )... RdA11T F Tc 4 D1 D (5) eleent x cos( 1) cos ( 1) cos( ), 0 c 0 c (4) P R = (h(0) LOS N eleent P t) (h(1) ref P t) n... n 1 N eleent (h() ref P t) n n 1 In which N eleent is the total nuber of reflecting eleent. P t is the average transitted power. TABLE 1: SIMULATION PARAMETERS Paraeters Values Roo diensions (x,y,z) 3x3x3 3 Reflection coefficient (ρ) Walls, ceiling, floor: 0.8, 0.8, 0.3 Transitted optical power 300 W Labert s order 1 Position of receiver, Rx (x, y, z) physical area of a photodetector (1.5,1.5,3.0) 1.0 c Half-angle FOV 70 [deg.] Sei-angle at half power 70 [deg.] Pixel size (Δx, Δy) 0.11x0.11 Gain of an optical concentrator 1 Refractive index of a lens at a 1.5 photodetector In these analyses, the listed paraeters in TABLE I were used. (7) h() ref ( 1)( eleent 1) A... 3 R da1da 1 8 D1 D3 D4 eleent x TF ( ) Tc ( ) cos ( ) cos( 1 ) cos ( )... eleent x cos( ) cos ( 3 ) cos( ), 0 c 0, c (6) where D 1, D, D 3, D 4, γ 1, γ, 1, and 3 are illustrated in Figure 1. ρ 1 and ρ are the reflection coefficients of the first and second surfaces respectively. eleent is considered as an ideal Labertian reflector and equal to 1. The obtained data using two reflections include three coponents (1) LOS, () first reflection off of floor surface, and (3) second reflection off of floor and all surfaces. The total received power resulting fro all coponents can be calculated as Figure : Ipulse Response Of All Channels As shown in Figure, the LOS path gives the highest contribution to the overall response, then the first and second order reflections coe respectively. Therefore, the ipulse responses will decline as the order of reflection increase. Previous studies perfored based on the ceiling bounce which consider the first reflection only which lose the contribution of the second reflection [11]. 355

31 st August 015. Vol.78. No.3 005-015 JATIT & LLS. All rights reserved. ISSN: 199-8645 www.jatit.org E-ISSN: 1817-3195 Figure 3: The Average Received Power With LOS Coponent And Without LOS Coponent At Different Transitter Heights. Figure 3 shows that the average received power increases gradually as the height of the transitter increased. The obtained data were fitted into a 4 th degree polynoial curve fitting. For instance, the average received power in the case of including LOS propagation is -6.4 to -3.4 db for all height points ranging fro 0 to 1.3. The generated polynoial equation for that particular case is: 4 3 yp,los 1.41x -.94x 1.87x 1.79x- 6.4 (8) In which y p is the predictable average received power. x is the transitter height. For the case of excluding LOS propagation, the average received power is -31.7 to - 7.9 db and the obtained polynoial equation is: yp,nlos 4 3 4.46x - 9.10x 5.57x 1.4x - 31.70 (9) Figure 4: Residuals For Eq.8 And Eq.9 4. ROOT MEAN SQUARE DELAY SPREAD RMS delay spread defines the delay of reflections. Taking into account the LOS path, first order reflections and second order reflections. The RMS delay spread is coputed using the ipulse response h(t) as follows [8,9]: 1 / ( t μ) h ( t) dt Drs (10) h ( t) dt where µ, the ean delay spread and calculated as: th ( t) dt μ (11) h ( t) dt The average received power with LOS coponent is about -5 db larger than the case of excluding LOS. By looking to the residuals for both curves as shown in Figure 4, it is illustrated that both equations have good accuracy, although eq.8 perfors better than eq.9. The highest residual values for eq.8 and eq.9 are at transitter height of 0.5 and 0.78 which are equal to 0.01 and 0.039 respectively. (a) 356

31 st August 015. Vol.78. No.3 005-015 JATIT & LLS. All rights reserved. ISSN: 199-8645 www.jatit.org E-ISSN: 1817-3195 For the OOK odulation, sybols are transitted over the Infrared channel. And the electrical SNR of equiprobable sybols such as x {0,Pt} at the reception is defined as R H SNR = Pt (14) R b N0 where R b is the data rate. N 0 is the shot noise since it is considered as the doinant noise source N 0 = I b q = 6.4x10-3 W/Hz, by considering I b = 00 µa and q = 1.6x10-19 C [11]. (b) Figure 5: Distribution Of RMS Delay Spread: (A) With LOS Coponent (B) Without LOS Coponent Figure 5(a) and (b) obtain the RMS delay spread versus width and length of the roo for easured ultipath channels. The channels without LOS path suffer of higher delay spread because the power received of LOS coponent is uch stronger than the received power of the first and second order reflections. Therefore, in the presence of LOS coponent, the RMS delay spread is ranging fro 0.16 ns to about 0. ns. Unlike, in the case of absence of LOS path where the RMS delay spread is slightly increasing to a range of 0.95 ns to about 1.1 ns. Since the axiu value of RMS delay is 1.1ns, thus axiu acceptable bit rate that can be transitted within the channel without requiring equalizer is about 91 Mb/s. the relationship between the axiu bit rate and the RMS delay is given as [4,10]: 1 R b (1) 10 D rs (a) 5. SYSTEM ANALYSIS AND PERFORMANCE OF SNR AND BER The intensity odulation and direct detection (IM/DD) schee is eployed to transit the patient data which eans that the signal received by the photo detector is ainly depended on the incident optical power and the detector sensitivity R. The IR channel with received signal y, channel gain state H and Additive White Gaussian Noise n can be odelled as [7]: where x is the input signal. y = R.H. x + n (13) (b) Figure 6: Distribution Of SNR For All Channels: (A) With LOS Coponent, (B) Without A LOS Coponent. The SNR can deterine the quality of the developed counication syste. As a result, Figure 6 (a) shows the perforance of all channels including LOS coponent. The iniu SNR value is 14 db and its axiu value is 4 db. On the other hand, Figure 6 (b) represents the 357

31 st August 015. Vol.78. No.3 005-015 JATIT & LLS. All rights reserved. ISSN: 199-8645 www.jatit.org E-ISSN: 1817-3195 distribution of SNR with no LOS coponent which gives SNR values ranging fro db to 14 db. Fro these two obtained cases, case (a) has better perforance than case (b) because case (a) includes LOS coponent which generates stronger signal due to the direct path between the transitter and receiver. Furtherore, the received power is inversely proportional to the direct and indirect distances between the transitter and receiver Moreover, the BER (bit error rate) for OOK odulation is given as BER Q SNR (15) 6. CONCLUSION In this paper, we have reported an optical wireless counication syste that tends to enhance the efficiency of the hospital staff. Through the siulation, we investigated the average received power at different transitter heights and showed the effect at each height point. Then we analysed the distribution of SNR and RMS delay spread within the roo. Moreover, we analysed the of the transitter height on the bit error rate. The overall perforance of the syste has iproved to the acceptable level, and the direction of the transitter has becoe ore obile by considering different directions. ACKNOWLEDGEMENT The authors would like to express their cordial thanks the Universiti Malaysia Perlis (UniMAP) for the constant support and encourageent. REFRENCES: Figure 7: SNR Vs. BER For Different Heights Figure 7 shows the relationship between the BER and SNR. The transitter height is taken into account and varied fro 0 to 1.3. by increasing the value of transitter height, better SNR and BER are achieved. In this case, LOS coponent was considered in obtaining the results. We have evaluated the perforance of the developed syste based on NZ-OOK odulation. The achieved BER for different positions of the transitter is reduced as the transitter height increases fro 0 to 1.3 throughout the roo. The best achieved BER is equal to 10-9 at height of 1.3 for data rate of 10 Mb/s which is enough to transit the edical inforation (EEG, ECG and etc.). Noting that, the acceptable BER level in optical counication is 10-9 and below. [1] M. Paksuniei, H. Sorvoja, E. Alasaarela, and R. Myllyla "Wireless sensor and data transission needs and technologies for patient onitoring in the operating roo and intensive care unit," Engineering in Medicine and Biology Society, 005. IEEE-EMBS 005. 7th Annual International Conference of the, vol., no., pp.518,5185, 17-18 Jan. 006 [] A.M, Khalid, G. Cossu, and E. Ciaraella, Diffuse IR-optical wireless syste deonstration for obile patient onitoring in hospitals, Transparent Optical Networks (ICTON), 013 15th International Conference on, vol., no., pp.1,4, 3-7 June 013 [3] J. B. Carruthers, and P. Kannan, Iterative site-based odeling for wireless infrared channels, IEEE Transactions on Antennas and Propagation, vol. 50, pp. 759 765, May 00 [4] H.Q. Nguyen, J. Choi, M. Kang, Z. Ghasselooy, D.H Ki, S. Li,T. Kang, and C.G. Lee, A MATLAB-based siulation progra for indoor visible light counication syste, Counication Systes Networks and Digital Signal Processing (CSNDSP), 010 7th International Syposiu on, vol., no., pp.537,541, 1-3 July 010 358

31 st August 015. Vol.78. No.3 005-015 JATIT & LLS. All rights reserved. ISSN: 199-8645 www.jatit.org E-ISSN: 1817-3195 [5] R. J. Green, H. Joshi, M. D. Higgins, and M. S. Leeson, Recent developents in indoor optical wireless systes, IET Counications, vol., pp. 3-10, 008. [6] S. Jivkova and M. Kavehrad, Indoor wireless infrared local access, ulti-spot diffusing with coputer generated holographic beasplitters, in Counications, 1999. ICC '99. 1999 IEEE International Conference on, vol. 1, pp. 604-608, 1999. [7] J.M. Kahn, and J.R. Barry, Wireless infrared counications, Proceedings of the IEEE, vol.85, no., pp.65,98, Feb 1997. [8] B. Carruthers, S. M. Caroll, and P. Kannan, Propagation odelling for indoor optical wireless counications using fast ultireceiver channel estiation, IEE Proceedings- Optoelectronics, vol. 150, pp. 473-481, 003. [9] Z. Ghasselooy, W.O. Popoola, and S. Rajbhandari, Optical Wireless Counications Syste and Channel Modelling with Matlab, CRC publisher, USA, August 01. [10] Y. Tanaka, T. Koine, S. Haruyaa and M. Nakagawa, Indoor Visible Light Data Transission Syste Utilizing White LED Lights, IEICE Trans. Coun. E86-B, 440-454 (003). [11] S.S. Torkestani, N. Barbot, S. Sahuguede, A. Julien-Vergonjanne, and J.P. Cances, Perforance and transission power bound analysis for optical wireless based obile healthcare applications, Personal Indoor and Mobile Radio Counications (PIMRC), 011 IEEE nd International Syposiu on, vol., no., pp.198,0, 11-14 Sept. 011 359