Uwb channel simulation using ray tracing algorithm

DUBLIN CITY UNIVERSITY SCHOOL OF ELECTRONIC ENGINEERING UWB channel simulation using Ray Tracing algorithm Tam N Huynh August 2009 MASTER OF ENGINEERING IN TELECOMMUNICATION Supervised by Dr C Brennan UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Acknowledgements I would like to thank my supervisor Dr Conor Brennan for his guidance, enthusiasm and commitment to this project Thanks are also due to Prof Charles McCorkell for his kindly supports from the first day I came to DCU Declaration I hereby declare that, except where otherwise indicated, this document is entirely my own work and has not been submitted in whole or in part to any other university Signed: ii Date: UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Abstract Ultra-wideband (UWB) communication systems have recently received great attenuation in both academia and industry for many applications in wireless communications Channel modelling for UWB has occupied an essential part in system design and implementation Among the variety of channel models, Ray Tracing (RT) has raised as an attractive deterministic channel modelling method However, there were some impacts on using Ray Tracing in simulating the UWB channel This dissertation will present RT and its application Beside that, we also propose some novel methods for efficiently computing Ray Tracing channel response A novel method of Parallel Ray Approximation (PRA) Ray Tracing for UWB indoor channel modelling is presented in this study with a great accuracy in phase calculation The performance of method was proved mathematically and from the simulation results Compared to the former PRA method, the proposed method has shown out a good improvement in error reducing with a slightly increasing in computation time Simulations have been performed for convention Ray Tracing, former PRA and the proposed PRA at some resolutions and grid size within the band of UWB Even when the resolution obtains 0.2m the relative error of proposed method is just 0.06% while the former PRA method shows out an unacceptable error of 24.3% The results show that our new way of approximating Ray Tracing algorithm is suitable for UWB in door channel modelling We also submit a communication letter for this novel method to IEEE Transaction on Antenna and Propagation Moreover, we combine the proposed PRA method with another method of efficient approximating Ray Tracing in frequency domain to save more computation time From the database obtained from RT channel simulation We propose a novel method for UWB localization application based on signal correlation The method has been performed in the simulation context and show that this is a potential way for UWB localization in both academia and industry application iii UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Table Of Contents ACKNOWLEDGEMENTS II DECLARATION II ABSTRACT III TABLE OF CONTENTS IV TABLE OF FIGURES VI CHAPTER - INTRODUCTION 1.1 UWB COMMUNICATION SYSTEMS AND APPLICATIONS 1.1.1 Regulatory bodies 1.1.2 Impulse Radio UWB (IR-UWB) 1.1.3 UWB Applications 1.2 RAY TRACING AND UWB CHANNEL MODELLING 1.2.1 Ray Tracing algorithm 1.2.2 UWB channel modelling CHAPTER – RAY TRACING FOR UWB CHANNEL MODEL 11 2.1 RAY TRACING FOR UWB CHANNEL MODELLING 11 2.1.1 Electric field of Dipole Hertz antenna 12 2.1.2 Calculation of reflection coefficient 14 2.1.3 Diffraction in channel modelling 22 2.2 SIMULATION RESULT 23 CHAPTER – EFFICIENT RAY TRACING APPROXIMATION FOR UWB CHANNEL MODELLING 29 3.1 FORMER PARALLEL RAY APPROXIMATION 29 3.2 PROPOSED PARALLEL RAY APPROXIMATION 36 3.3 REFLECTION COEFFICIENT APPROXIMATION IN FREQUENCY DOMAIN 42 CHAPTER – APPLYING RAY TRACING IN UWB LOCALIZATION 45 4.1 SIGNAL CORRELATION IN UWB RAY TRACING LOCALIZATION 45 4.2 PSEUDO RECEIVED SIGNAL USING SEMI-DETERMINISTIC MODEL 47 4.2.1 Saleh-Valenzuela (SV) model adopted in 802.15.4a 47 iv UWB channel simulation using Ray Tracing algorithm – Tam N Huynh 4.2.2 Semi-Deterministic channel model for UWB 48 4.3 RAY TRACING UWB LOCALIZATION SIMULATION 51 4.3.1 Characteristic of correlation along the searched curve 51 4.3.2 Efficient pre-computed RT for localization 53 4.3.3 Error evaluation for RT localization and tracking 54 CHAPTER - CONCLUSIONS AND FURTHER RESEARCH 58 5.1 CONCLUSION 58 5.2 FURTHER RESEARCH 59 REFERENCES 60 APPENDIX 63 v UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Table of Figures FIGURE 1.1 THE FCC EMISSION LIMITS FOR THE INDOOR UWB COMMUNICATIONS [17] FIGURE 1.2 GAUSSIAN AND MONOCYCLE PULSE IN TIME DOMAIN FIGURE 1.3 GAUSSIAN AND MONOCYCLE PULSE IN FREQUENCY DOMAIN FIGURE 1.4 SINUSOIDAL GAUSSIAN PULSE IN TIME DOMAIN FIGURE 1.5 SINUSOIDAL GAUSSIAN PULSE IN FREQUENCY DOMAIN FIGURE 1.6 UWB POTENTIAL APPLICATIONS [17] FIGURE 1.7 RAY TRACING ALGORITHM FIGURE 2.1 GEOMETRICAL ARRANGEMENT OF AN INFINITESIMAL DIPOLE (A) AND ITS ASSOCIATED ELECTRIC-FIELD COMPONENTS ON A SPHERICAL SURFACE (SOURCE [13]) 13 FIGURE 2.2 GEOMETRY FOR CALCULATING THE REFLECTION COEFFICIENT 15 FIGURE 2.3 GEOMETRY FOR CALCULATING THE ARBITRARY POLARIZATION REFLECTION COEFFICIENT 16 FIGURE 2.4 FREQUENCY DEPENDENCE OF REFLECTION COEFFICIENT AT ߠ݅ ൌ ߨ/3 AND ߛ ൌ ߨ/4 18 FIGURE 2.5 PARALLEL AND PERPENDICULAR POLARIZATION RC DEPEND ON INCIDENT ANGLE AT FC=7GHZ AND ߛ ൌ ߨ/4 19 FIGURE 2.6 TOTAL RC DEPEND ON INCIDENT ANGLE ߠ݅ AT FC=7GHZ AND ߛ ൌ ߨ/4 19 FIGURE 2.7 REFLECTION COEFFICIENT DEPEND ON ANGLE Γ AT ߠ݅ ൌ ߨ/3 AND FC=7GHZ 20 FIGURE 2.8 REFLECTION COEFFICIENT DEPEND ON ANGLE Γ AND ߠ݅ AT FC=7GHZ 21 FIGURE 2.9 GEOMETRY FOR COMPUTATION DIFFRACTION IN CASE OF SINGLE SCREEN 22 FIGURE 2.10 GEOMETRY FOR COMPUTATION DIFFRACTION IN CASE OF SINGLE WEDGE 23 FIGURE 2.11 RAY TRACING PATHS IN 10MX10MX5M ROOM (UP TO SECOND ORDER REFLECTION) 24 FIGURE 2.12 RECEIVED SIGNAL OF LOS PATH 25 FIGURE 2.13 DISTORTION RECEIVED SIGNAL FROM REFLECTION 25 FIGURE 2.14 RECEIVED SIGNAL FROM MULTI-PATH CHANNEL UP TO FIRST ORDER REFLECTION (A) AND SECOND ORDER REFLECTION (B) 26 FIGURE 2.15 RECEIVED SIGNAL FROM AWGN CHANNEL UP TO FIRST ORDER REFLECTION27 FIGURE 2.16 RECEIVED SIGNAL FROM AWGN CHANNEL UP TO SECOND ORDER REFLECTION 28 FIGURE 3.1 PRA FOR ADJACENT POINT M FROM CALCULATED POINT N 30 FIGURE 3.2 PRA FOR ADJACENT POINT M FROM CALCULATED POINT N 31 FIGURE 3.3 COMPARE THE RECEIVED SIGNAL AT POINT M AT RESOLUTION OF 0.025M (A), 0.1M (B), 0.2M (C) 33 FIGURE 3.4 ERROR OF PRA RT IN GRID SIZE 12X12 COMPUTATION POINTS AT RESOLUTION OF 0.025M (A), 0.1M (B), 0.2M (C) 35 FIGURE 3.5 COMPARE RECEIVED SIGNAL BETWEEN FULL RT, FORMER PRA RT, AND OUR PROPOSED PRA RT AT RESOLUTION OF 0.025M (A), 0.1M (B), 0.2M (C) 38 vi UWB channel simulation using Ray Tracing algorithm – Tam N Huynh FIGURE 3.6 PERCENTAGE ERROR OF RECEIVED SIGNALS IN OUR PROPOSED PRA RT AT RESOLUTION OF 0.025M (A), 0.1M (B), 0.2M (C) 40 FIGURE 3.7 POLYNOMIAL APPROXIMATION OF REFLECTION COEFFICIENT (A) AND TH PERCENTAGE ERROR OF METHOD WHEN ORDER OF APPROXIMATED POLYNOMIAL IS (A) TH AND 10 (B) 43 FIGURE 4.1 CURVE OF SIMULATED POINTS FOR LOCALIZATION 46 FIGURE 4.2 PRINCIPLE OF SV MODEL (SOURCE [1]) 47 FIGURE 4.3 RECEIVED SIGNAL FROM CLUSTER SEMI-DETERMINISTIC CHANNEL MODEL WITH (A) AND WITHOUT (B) AWGN NOISE 49 FIGURE 4.4 RECEIVED SIGNAL FROM RANDOM SEMI-DETERMINISTIC CHANNEL MODEL WITH AWGN NOISE 50 FIGURE 4.5 SAMPLE SEARCHED CURVE (A) AND THE CORRELATION ALONG THE SEARCHED CURVE (B) 52 FIGURE 4.6 APPROXIMATION ON THE PRE-COMPUTED RT GRID 53 FIGURE 4.7 ERROR OF RT LOCALIZATION WHEN THE AUXILIARY RAYS COME TO RECEIVER RANDOMLY (A) AND IN RAYLEIGH CLUSTER (7) 55 FIGURE 4.8 PROPOSED ALGORITHM TO IMPROVE THE PERFORMANCE OF RT TRACKING 56 FIGURE 4.9 ERROR OF PROPOSED ALGORITHM WHEN PSEUDO RECEIVED SIGNAL COME IN CLUSTER 56 FIGURE 4.10 TRACKING ERROR WHEN ADDING NOISE TO TOA 57 vii UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Chapter - Introduction Ultra wideband (UWB) communications is based on the transmission of very short pulses with relatively low energy [4] Recently, UWB systems have received great attenuation in both academia and industry for applications in wireless communications In particular, UWB is a promising area offering enormous advantages for short-range communications In order to evaluate and design any communication systems, the channel should be modelled as an essential part of the process Unlike the conventional narrow band systems, some unique features occur when transmitted signal in UWB has extremely large bandwidth 3.1 GHz to 10.6 GHz, follow the regulation of FCC issued in 2002 In general, there are two prevalent types of channel modelling: deterministic (or sitespecific) modelling and statistical modelling [1] In this dissertation, the Ray Tracing (RT) algorithm – one of the deterministic channel models will be investigated The extension of RT channel model to semi-deterministic model war also investigated The big impact when apply Ray Tracing in UWB is that the computation time increases prohibitively with amount of computation points 1.1 UWB communication systems and applications Wireless communication systems have evolved substantially over the last two decades The explosive growth of the wireless communication market is expected to continue in the future, as the demand for all types of wireless services is increasing Ultra Wideband (UWB) communication systems have an unprecedented opportunity to impact communication systems The enormous bandwidths available, the wide scope of the data rate/range trade off, and the potential for very-low-cost operation leading to pervasive usage UWB has a number of features which make it attractive for consumer communications applications In particular, UWB systems have some unique characteristics [1]: • Potentially low complexity and low cost • A noise-like signal spectrum • Resistant to severe multi-path and jamming • Very good time-domain resolution allowing for location and tracking applications UWB channel simulation using Ray Tracing algorithm – Tam N Huynh 1.1.1 Regulatory bodies The Defence Advanced Research Projects Agency (DARPA) provided the first definition for UWB signal based on the fractional bandwidth Bf of the signal The first definition provided that a signal can be classified as an UWB signal if Bf is greater than 0.25 The fractional bandwidth can be determined as the formula (1) ‫ܤ‬௙ ൌ ௙೓ ା௙೗ ௙ ି௙ ೓ ೗ (1) Where fl is the lower and fh is the higher −3 dB point in the spectrum, respectively In Feb 2002, the FCC issued the FCC UWB rulings that provided the first radiation limitations for UWB, and also permitted technology commercialization The FCC approved the development of UWB on an unlicensed basis in the 3.1 – 10.6 GHz band subject to a modified version of Part 15.209 rules The essence of the rulings is that the power spectral density (PSD) of modulated UWB signal must satisfy the spectral mask specified by spectrum-regulating agencies as in Table 1.1 and Figure 1.1 Table 1.1 FCC radiation limits for indoor and outdoor communication (source [1]) UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Figure 1.1 The FCC Emission limits for the indoor UWB communications [17] The IEEE also established the 802.15.3a Study Group to define a new physical layer concept for short-range, high-data-rate applications and 802.15.4a for low-data-rate with very low power, low complexity systems [1] 1.1.2 Impulse Radio UWB (IR-UWB) UWB systems have historically been based on impulse radio concept Impulse radio refers to the generation of a series of very short pulses, of the order of hundreds of picoseconds The basic model for an IR-UWB pulse trains was given in (2) ஶ ‫ݏ‬ሺ‫ ݐ‬ሻ ൌ ෍ ‫ܣ‬௜ ሺ‫ݐ‬ሻ‫݌‬൫‫ ݐ‬െ ݅ܶ௙ ൯ ሺ2ሻ ௜ୀିஶ Where Ai(t) is the amplitude of the pulse equal to േඥ‫ܧ‬௣ ( Ep is the energy per pulse ), p(t) is the received pulse shape with normalized energy, and Tf is the frame repetition time (A UWB frame is the time interval in which one pulse is transmitted) The variety of pulse shape p(t) was introduced in [1] By far the most popular pulse shapes in IR-UWB literature the Gaussian pulse and its derivatives The analytical description of Gaussian pulse for UWB is described in (3) and its first derivative (or known as Monocycle pulse) in (4) భ ೟షഋ మ ቁ ഑ ‫݌‬ሺ‫ݐ‬ሻ ൌ ‫ܣ‬଴ ݁ ିమቀ భ ೟షഋ మ ቁ ഑ ‫݌‬ሺ‫ݐ‬ሻ ൌ ‫ܣ‬ଵ ሺ‫ ݐ‬െ ߤሻ݁ ିమቀ (3) (4) UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Cluster of rays (a) (b) Figure 4.3 Received signal from Cluster Semi-Deterministic channel model with (a) and without (b) AWGN noise 49 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Another proposed approach to generate the pseudo received signal also used in this dissertation is that: instead of assuming received signals come in cluster, we assume the auxiliary rays not come in cluster Beside the “main” components which obtained in received signal from RT, the other auxiliary components will have the randomly attenuations and phase shifts respect to uniform distribution The equation (47) is expressed for the channel response ௅ ௄ ௟ୀ଴ ௞ୀ଴ ݄ ሺ‫ݐ‬ሻ ൌ ෍ ߙ௟ expሺ݆߶௟ ሻ ߜ ሺ‫ ݐ‬െ ߬௟ ሻ ൅ ෍ ߙ௞ expሺ݆߶௞ ሻ ߜ ሺ‫ ݐ‬െ ߬௞ ሻ ሺ47ሻ In (47), L first components are obtained from RT as discuss above The remain K components will have ߙ௞ , ߬௞ follow the uniform random process The distortion in phase ߶௞ also respect to uniform distribution in [0 2ߨ] Figure 4.4 will depict the received signal of this model when the number of auxiliary random rays up to 49 rays The AWGN noise was also accounted in channel simulation Figure 4.4 Received signal from Random Semi-Deterministic channel model with AWGN noise 50 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh 4.3 Ray Tracing UWB localization simulation 4.3.1 Characteristic of correlation along the searched curve The main idea of this localization method is based on correlation as introduced previously So the characteristic of the signal correlation along the searched curve should be investigated firstly Figure 4.5 will be useful in this issue (a) 51 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh (b) Figure 4.5 Sample searched curve (a) and the correlation along the searched curve (b) The searched curve in this case, Figure 4.5(a), is a complete circle with BS at the centre of the circle, the radius of searched circle is obtain from the TOA as discussed in the previous section The resolution of the searched curve is 0.01m (i.e distance between two points on the curve which used in calculating received signal by RT is 0.01m) According to the correlation along the searched curve in Figure 4.5 (b), the correlation value was distributed in some groups, and each group has its local maximum The position of real MS is the point which associated with the global maximum correlation However, when the resolution of searching process (i.e the resolution of the searched curve) is reduced to save computation time (will be investigated in next), the sampling distance between two adjacent points on the correlation line (in Figure 4.5(b)) will be increased That leads to a problem: the real global maximum correlation may be missed and another local maximum (not associated with the real position of MS) becomes the global maximum This problem leads to the error in RT localization that we must deal with Moreover, the computation time for the localization process above is quite large In this scenario, simulation run on Dell D600 laptop (Pentium M 1.7 GHz, 1GB RAM), and RT 52 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh computed up to first order reflection, the computation time for tracing occupy from tens to hundreds of second Because of RT must be implemented at every point on the searched curve to calculate the received signal The more searched points on the searched curve, the longer calculating time for localization process The number of searched point is decided by the distance from MS to BS (i.e the length of searched curve), and the resolution of searching process The sub section 4.3.2 will propose an efficient way to solve this problem 4.3.2 Efficient pre-computed RT for localization As discuss in the previous sub section, most of the time for localization process is paid for RT computation on each point of the searched curve In this subsection, we propose that the RT received signals should be pre-computed and load into RAM before we localization So the computation time now just paid for taking correlation However, because of RT is pre-computed on a grid, so we must some approximation for the position of searched points as explained in Figure 4.6 Figure 4.6 Approximation on the pre-computed RT grid According to the Figure 4.6, the pre-computed RT just can be computed in a coarse grid due to the large of database for it In this simulation, in an area of 10x10m room, with the resolution of 0.1m, the length for simulated signal up to 212 points in Time Domain, the size of database is above 600 MB So we can in crease the resolution (due to the limit of 53 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh hardware 1GB RAM) The problem occur now some points on the searched curve is fixed with the grid (as point E), but some not fix well and need an approximation (as point A) And this approximation will lead to errors in localization However, trading off with the errors occur, the computation time was significantly decreased Calculating time was reduced from tens, hundreds of second down to less than 3s (from 1s to 3s, depend on the number of searched point) And this shows an optimistic application for industrial RT localization The errors occur from approximated points and sampled correlation will be evaluated in the next subsection To improve the efficient of method, some simple algorithm will be proposed and investigated for localization and tracking moving object 4.3.3 Error evaluation for RT localization and tracking The error was given as the distance between real position of MS and predicted position of MS from RT localization For evaluating error of the methods, the pseudo received signals were generated as two methods (cluster and non-cluster for auxiliary rays) discussed in previous section The simulation created 100 random MS positions in the room and localization was implemented to specify the object It’s worth to note that the resolution for searching process is 0.1m (limited by the resolution of pre-computed RT grid) and assuming we have the exact TOA (i.e we have the exact distance d between MS and BS) Here we use the pre-computed RT method as proposed above to save the calculating time and evaluated the error of approximation as discussed in sub section 4.3.2 Figure 4.7 shows out the performance of RT localization (a) 54 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh (b) Figure 4.7 Error of RT localization when the auxiliary rays come to receiver randomly (a) and in Rayleigh cluster (7) According to the Figure 4.7(a), when the Semi-Deterministic with random auxiliary rays is used to simulate the pseudo received signal, the errors occur less than if we assume they come in cluster respected to Rayleigh distribution In some case, there was a very small error occurs here, that come from the error of approximation in the grid as we discuss above Moreover, both figures show that, sometimes, the error is very large (over 6m) That’s reasonable! Because the characteristic of sampled correlation along the searched curve as we investigated above In these cases, the global maximum of correlation was missed due to the sampling resolution and the wrong local maximum becomes the global maximum However, for tracking moving object using RT, a simple algorithm (proposed next) can be applied to reduce these errors In tracking moving object, the object moving with an arbitrary velocity, so the position of the object at a moment will be relative with the position of object at the previous moment That leads to our simple algorithm for tracking object using RT as below The threshold is quite important and it is specified from the velocity of moving object For example, the object moving with speed of 1.5m/s in the room, and our tracking simulator need 3s for calculating, so the threshold should be set at 4.5m Figure 4.9 depicted the errors of tracking when apply this simple but efficient algorithm The pseudo received signal used in Figure 4.9 was followed the Rayleigh cluster model 55 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Yes Figure 4.8 Proposed algorithm to improve the performance of RT tracking Figure 4.9 Error of proposed algorithm when pseudo received signal come in cluster According to the Figure 4.9, the error is improved significantly Most of the remained errors now caused by the approximation on the grid of pre-computed RT The result of this improvement also can be seen in the *.avi files for tracking simulation (Appendix): • “track_cluster1.avi” • “track_cluster2.avi” • “track_random1.avi” • “track_random1.avi” 56 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Where “track” for tracking simulation, “cluster” for pseudo received signal followed the cluster model, “random” for pseudo received signal with random model The number is presented for tracking without proposed algorithm and number is with the proposed algorithm Until now, we still the estimated TOA is exact Now let add some uniform distribution error on the TOA so that the distance d from MS to BS will be added some noise Figure 4.10 shows out that d is a very important coefficient in RT localization application The errors in that figure is quite large and more frequently than error without d-noised as in Figure 4.7 and Figure 4.9 Figure 4.10 Tracking error when adding noise to TOA The simulated of tracking process in case of d-noised with and without our proposed algorithm also present in “track_clus_dnoise1.avi” and “track_clus_dnoise2.avi” Where number and present for without and with our proposed algorithm, respectively The results also show that the proposed algorithm still works well Even though there were some errors occur, it still track the motion of the object 57 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Chapter - Conclusions and Further Research 5.1 Conclusion From the beginning up to this chapter, we have known that Ultra wideband (UWB) communication is based on the transmission of very short pulses with relatively low energy This technology may see increased use in the field of wireless communications and ranging in the near future UWB technique has a fine time resolution which makes it a technology appropriate for accurate ranging The UWB radio signal occupies a bandwidth of more than 500 MHz or a fractional bandwidth of larger than 20 % Channel modelling in UWB was raised up as an important part in UWB system design and implementation Among plenty methods for UWB channel modelling, Ray Tracing has shown its strengths in this manner And this project was laid out this method and some novel efficient ways to enhance the performance of UWB RT Firstly, the characteristic of the UWB channel was investigated in some sections From that, the full 3D RT channel simulation was implemented and evaluated We also investigate the Parallel Ray Approximation method for approximate the adjacent point in the computed gird as proposed in [5] Beside of the efficient saving in computation time, this method shows an unacceptable error when we decrease the resolution of computed grid That leads to our proposed PRA RT method for increase the accuracy with a very small trade off in simulation time The proposed method was proved mathematically and shown out by comparing the simulation results This method show a new way of reducing the simulation time in conventional Ray Tracing method with a great accuracy in phase Even when apply to an area with a very low resolution (0.2m), the method still shows its advantage with significant saving simulation time compared with conventional Ray Tracing In addition, the communication also shows that the modified PRA method is suitable way for simulating the UWB channel in indoor environment where the Ray Tracing simulator must due with a huge computation Moreover, the method of efficiently approximating the reflection coefficient in frequency domain by a low order polynomial in [6] was also investigated again This leads to novel 58 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh combined method that combines our proposed PRA RT for the adjacent points with this frequency approximation method for accurate calculated points Some simulations were taken and this improvement in saving computation time From database obtained by RT UWB channel simulation, we also propose its application in localization in chapter A new method based on RT and the signal correlation was proposed and performed in simulation context Some models for generating pseudo received signals were proposed for evaluating the novel localization method Moreover, noises in propagation channel and in estimation of TOA were added to perform this proposed method We also propose a simple algorithm to reduce the error when tracking a moving object Simulation results prove that this is a potential application for RT and UWB localization 5.2 Further Research Both UWB channel modelling and UWB localization are still the active research topics nowadays Through out this dissertation some methods were presented and investigated The further study of this project should be for researching on combining RT with other statistical channel modelling as presented in [1], [2], [3] The more accurate model we obtain, the more efficient in UWB communication systems design will be Among them, the Semi-Deterministic model in this dissertation should be considered with more improvement to obtain the real signal in time domain Beside that applying RT in UWB localization is a novel idea and need to be investigated more Even though this is an efficient way for tracking moving object, the error will be high if there were noises The simple algorithm for reduce error in dissertation should be improved for accurate by applying some kind of filter (Kalman filter for example) That should be a potential application in both academia and industrial fields 59 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh References [1] Andreas F Molisch et al., “UWB Communication Systems a comprehensive overview”, New York, Hindawi, 2006, pp1-17, pp40-94 [2] X Shen, M Guizani, R C Qiu, T Le-Ngoc, “Ultra-wideband wireless communications and networks”, West Sussex, Wiley, 2006, pp37-39, pp279-301 [3] Jeffrey H Reed et al., “An Introduction to Ultra Wideband communication systems”, Boston, Prentice Hall, 2005, pp73-151 [4] Andreas F Molisch, “Ultrawideband Propagation Channels-Theory, Measurement, and Modeling”, IEEE Transactions Vehicular Technology, Vol 54, no 5, pp 1528 – 1545, Sept 2005 [5] G Tiberi, S Bertini; W.Q Malik, A Monorchio, D.J Edwards ,and G Manara, “Analysis of Realistic Ultrawideband Indoor Communication Channels by Using an Efficient Ray-Tracing Based Method”, IEEE Trans Antennas Propag., vol 57, no 3, pp 777 – 785, Mar 2009 [6] J Diskin, C Brennan, “Accelerated Ray-Tracing for Indoor Ultra-wideband Propagation Modelling”, presented at IEEE 65th Vehicular Technology Conference, pp 418 - 422, Apr 2007 [7] R Yao, W Zhu, Z Chen, “An efficient time-domain ray model for UWB indoor multipath propagation channel”, presented at IEEE 58th Vehicular Technology Conference, vol 2, pp 1293 – 1297, Oct 2003 [8] G Tiberi, S Bertini, A Monorchio, F Giannetti and G Manara, “Computationally efficient ray-tracing technique for modelling ultrawideband indoor propagation channels”, IET Microwaves, Antennas & Propagation, vol 3, no 3, pp 395 – 401, Apr 2009 [9] X Li, Z Du, K Gong, Z Feng, “UWB Channel Modeling and Simulation”, presented at International Conference on Microwave and Millimeter Wave Technology, pp - 4, Apr 2007 [10] A Muqaibel, A Safaai-Jazi, A Bayram, A.M Attiya, S.M Riad, “Ultrawideband through-the-wall propagation”, IEE Proceedings on Microwaves, Antennas and Propagation, pp 581 – 588, Dec 2005 [11] G Tiberi, S Bertini, W.Q Malik, A Monorchio, D.J Edwards, G Manara, “An Efficient Ray Tracing Propagation Simulator for Analyzing Ultrawideband 60 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Channels”, IEEE International Conference on Ultra-Wideband, pp 794 – 799, Sept 2007 [12] Y Liu, W Dong, Y Zhang, C, Wei, “An efficient method of time-domain ray tracing based on the deterministic splitting method of ray tubes”, Asia-Pacific Microwave Conference APMC 2008, pp – 4, Dec 2008 [13] Constantine A Balanis, “Antenna theory analysis and design”, New Jersey, John Wiley & Sons, 2005, pp 184 – 204 [14] M Ghavami, L B Michael, R Kohno, “Ultra Wideband signals and systems in communication engineering”, West Sussex, John Wiley & Sons, 2004, pp.97-125 [15] Theodore S.Rappaport, “Wireless communications principle & practice ”, New Jersey, Prentice Hall, 1996, pp.78-100 [16] Andreas F.Molisch, “Wireless communications”, West Sussex, John Wiley & Sons, 2005, pp.43-61 [17] H Nikookar, R Prasad, “Introduction to Ultra Wideband for Wireless communications ”, Springer, 2009, pp.11-27, pp.43-64 [18] Andreas F.Molisch et al., “IEEE 802.15.4a channel model - final report”, IEEE document of the IEEE 802.15.4a channel modeling subgroup, 2007 [19] Xiaobing Sun, Yugang Ma, Jin Xu, Jian Zhang, Junjun Wang, “A High Accuracy Mono-Station UWB Positioning System”, IEEE International Conference on UltraWideband ICUWB 2008,pp 201 – 204, Sept 2008 [20] Dongwoo Kang, Young Namgoong, Suckchel Yang, Sungsoo Choi, Yoan Shin, “A Simple Asynchronous UWB Position Location Algorithm Based On Single RoundTrip Transmission ”, , 2006 The 8th International Conference Advanced Communication Technology ICACT 2006, pp.1461-1465, Feb 2006 [21] L Pierucci, P.J Roig, “UWB localization on indoor mimo channels ”, International Conference on Wireless Networks, Communications and Mobile Computing, pp 890 – 894, June 2005 [22] Yung-Hoon Jo, Joon-Yong Lee, Dong-Heon Ha, Shin-Hoo Kang, “Accuracy Enhancement for UWB Indoor Positioning Using Ray Tracing ”, IEEE/ION Position, Location, And Navigation Symposium, pp 565 – 568, April 2006 [23] J Kunisch and J Pamp, “Radio channel model for indoor UWB WPAN environments”, presentation P802.15 02/281SG3a, IMST GmbH, June 2002, IEEE P802.15 SG3a contribution 61 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh [24] J.Kunisch and J.Pamp, “Measurement results and modeling aspects for the UWB radio channel, ” in Proceedings of IEEE Conference on UltraWideband Systems and Technologies UWBST ’02, pp.19–23, May2002 [25] J Chuang, S Bashir, D.G Michelson, “Automated Identification of Clusters in UWB Channel Impulse Responses ”, Canadian Conference on El, ectrical and Computer Engineering CCECE 2007, pp.761 – 764, April 2007 [ 26] M.L Tobin, J.E Richie, “A 2-D Ray Tracing Model for the Characterization of Spatial and Time-Domain Properties of the Indoor Propagation Channel ”, 1995 AP-S Digest Antennas and Propagation Society International Symposium, pp 1948 – 1951, June 1995 [27] Xu Yong, Luying-Hua, Hepeng-Fei, Wangye-Qiu, “Electromagnetic Simulation of UWB Propagation in Indoor Environments ”, The 2006 4th Asia-Pacific Conference on Environmental Electromagnetics, pp 913 – 916, Aug 2006 [28] M Porebska,T Kayser, W Wiesbeck, “Verification of a Hybrid Ray-Tracing/FDTD Model for Indoor Ultra-Wideband Channels ”, 2007 European Conference on Wireless Technologies, pp 169 – 172, Oct 2007 62 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Appendix T.N Huynh, C Brennan, “An efficient Ray-Tracing method for indoor UWB channels”, will submitted to IEEE Trans on Antenna and Propagation List of *.avi files in tracking simulation: • • • • • • “track_cluster1.avi” “track_cluster2.avi” “track_random1.avi” “track_random2.avi” “track_clus_dnoise1.avi” “track_clus_dnoise2.avi” “3D_RayTracing_wave.avi” 63 ... channel model as in [18] 10 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh Chapter – Ray Tracing for UWB channel model The fundamentals of Ray Tracing and UWB channel modelling were... 0.025m, 0.1m, and 0.2m respectively 31 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh (a) (b) 32 UWB channel simulation using Ray Tracing algorithm – Tam N Huynh (c) Figure 3.3... MODELLING 1.2.1 Ray Tracing algorithm 1.2.2 UWB channel modelling CHAPTER – RAY TRACING FOR UWB CHANNEL MODEL 11 2.1 RAY TRACING FOR UWB CHANNEL MODELLING