Department of Electronic Engineering FINAL YEAR PROJECT REPORT

Department of Electronic Engineering FINAL YEAR PROJECT REPORT BEngIE-2006/07-QTZ-05 Multiple-Access Schemes for IEEE 802.15.4a Student Name: Wong Ching Tin Student ID: Supervisor: Prof. ZHANG, Keith Q T Assessor: Dr. Wong, W K Bachelor of Engineering (Honours) in Information Engineering 1

Student Final Year Project Declaration I have read the student handbook and I understand the meaning of academic dishonesty, in particular plagiarism and collusion. I declare that the work submitted for the final year project does not involve academic dishonesty. I give permission for my final year project work to be electronically scanned and if found to involve academic dishonesty, I am aware of the consequences as stated in the Student Handbook. Project Title: Multiple-Access Schemes for IEEE 802.15.4a Student Name: Wong Ching Tin Student ID: Signature Date: 26-4-2007 2

No part of this report may be reproduced, stored in a retrieval system, or transcribed in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of City University of Hong Kong. 3

Table of Contents Acknowledgements-----------------------------------------------------------------------------iii List of figures------------------------------------------------------------------------------------iv Abstract-------------------------------------------------------------------------------------------v Chapter 1: Introduction-------------------------------------------------------------------------1 1.1 Objectives-----------------------------------------------------------------------2 1.2 Report Outline------------------------------------------------------------------2 Chapter 2: Background--------------------------------------------------------------------------3 2.1 Description of IEEE 802.15.4a-----------------------------------------------3 2.1.1 Definition of IEEE 802.15.4a------------------------------------------3 2.2 Description of ultra wideband impulse radio (UWB-IR)---------------3 2.2.1 Definition of UWB-IR-------------------------------------------------3 2.2.2 Advantages of using UWB-IR----------------------------------------5 2.2.3 Applications of UWB-IR----------------------------------------------5 2.3 Summary------------------------------------------------------------------------6 Chapter 3: Time Hopping (TH)----------------------------------------------------------------7 3.1 Description of TH--------------------------------------------------------------7 3.1.1 What is TH?--------------------------------------------------------------7 3.1.2 Application of TH-------------------------------------------------------7 3.2 TH Modulation Schemes------------------------------------------------------8 3.2.1 Time hopping pulse position modulation (TH-PPM)------------8 3.2.2 Time hopping binary phase shift keying (TH-BPSK)----------10 3.3 Channel Model----------------------------------------------------------------12 3.4 Receiver Model---------------------------------------------------------------12 3.5 Theoretical Analysis----------------------------------------------------------12 i

3.5.1 Comparison of theoretical BER and simulation BER-------------13 3.6 Summary-----------------------------------------------------------------------14 Chapter 4: Consideration in the real environment-----------------------------------------15 4.1 Transmitter structure--------------------------------------------------------15 4.2 Receiver structure------------------------------------------------------------16 4.3 Selection of pulse------------------------------------------------------------17 4.4 Transmission medium/ channel--------------------------------------------17 4.5 Summary----------------------------------------------------------------------17 Chapter 5: Methodology-----------------------------------------------------------------------18 Chapter 6: Simulation results and analysis-------------------------------------------------20 6.1 Variation in number of users, Nu-------------------------------------------20 6.2 Variation in number of paths, Npath---------------------------------------23 6.3 Variation with processing gains, Nh and Ns------------------------------26 6.4 Summary---------------------------------------------------------------------- 28 Chapter 7: Conclusion-------------------------------------------------------------------------29 References---------------------------------------------------------------------------------------30 Appendix A: Program codes used for simulation------------------------------------------31 ii

Acknowledgments I would like to extend my sincere thanks to my supervisor, Prof. Zhang, my assessor, Dr. Wong, and Dr. Song. I need to thank my supervisor, Prof. Zhang, for his support and encouragement. When I felt frustrated of my project, he could always make me feel more confident after talking with him. Prof. Zhang also gave me useful suggestions. Dr. Wong gave me good advices on presentation strategies. I need to say thank you to Dr. Song in particular for spending long time to discuss the issues concerning my work frequently. When I had difficulties, he always discussed the possible problems of my work. Thanks for his patience guidance. Finally, thank you to my good classmates for their support and help, and also my family. iii

List of figures Figure 2.1 UWB spectrum compared to narrowband signal------------------------------4 Figure 2.2 Spatial capacity comparisons-----------------------------------------------------6 Figure 3.1 Illustration of Time Hopping-----------------------------------------------------7 Figure 3.2 Illustration of TH-PPM-----------------------------------------------------------10 Figure 3.3 Illustration of TH-PPM-----------------------------------------------------------11 Figure 3.4 Comparisons of Simulation and Theoretical Result--------------------------13 Figure 4.1 Transmitter structure--------------------------------------------------------------15 Figure 4.2 RAKE receiver structure---------------------------------------------------------16 Figure 6.1 TH-PPM Variation of Number of Users, Nh=20, Ns=8----------------------21 Figure 6.2 TH-BPSK Variation of Number of Users, Nh=20, Ns=8--------------------22 Figure 6.3 TH-PPM Variation of Number of Paths, Nh=20, Ns=8----------------------24 Figure 6.4 TH-BPSK Variation of Number of Paths, Nh=20, Ns=8---------------------25 Figure 6.5 TH-PPM Variation of Processing Gains, Nh*Ns=100------------------------27 iv

Abstract Ultra wideband (UWB) impulse radio (IR) is a prospective transmission technology for low-rate indoor communications, as described in the physical layer proposals for IEEE 802.15.4a wireless personal area networks (WPAN). Time hopping (TH) is considered as an access scheme for multi-user UWB-IR systems. Communication over the indoor wireless channel is a technical challenge, not only because it is varies from place to place, but also because the signal suffers from heavy multi-path propagation. The motivation to search for improved modulation schemes is to improve the performance of data transmission, reduce the Bit Error Rate (BER) and increase the multiple-access capacity. In this project, two modulation schemes, namely, time hopping-binary phase shift keying (TH-BPSK) and time hopping-pulse position modulation (TH-PPM) were studied. Simulation is done using MatLab to verify that the performance is closed to theoretical value and TH-BPSK outperforms TH-PPM. The influence of different combinations of TH processing gains, multipaths and multi-users on BER performance of the two schemes is analysed and numerical results are presented for illustration. v

Chapter 1: Introduction The IEEE 802.15 WPAN focuses on the development for short distance WPAN. In this standard, a new physical layer concept for low data rate applications using ultrawideband (UWB) technology at the air interface is defined compared with the high-rate communications in IEEE 802.15.3a. The study group addresses new applications that require only low to moderate data throughput, long battery life such as low data rate WPAN and small networks. The WPAN is aimed at short distance wireless networking of portable and mobile computing devices, such as palm and notebook. To implement this short distance network, UWB is used for transmission of data through channel, which interferences and noises may introduce inside the channel. UWB modulation was initially proposed for military communications due to its resistance to jamming, and since it allows hidden communications. UWB was considered also for civilian applications. Advantages of UWB include resilience to multi-path and the capacity of coexisting with other systems sharing the same band. There are two optional physical layers in this standard; one is Chirp Spread Spectrum (CSS) and the other is the main focus in this project, which is Impulse Radio (IR). IR is a form of spread spectrum (SS) signalling technique belonging to the UWB systems, typically on the order of nanosecond. IR is probably the best-known form of UWB for communications, having a simple implementation is additional advantage. IR can spread the energy of the radio signal very thinly from near dc to a few gigahertz. One method for spreading the spectrum of UWB low-duty-cycle pulse trains is time hopping (TH), with data modulation accomplished by additional pulse modulation at the rate of many pulses per data symbol. TH is a simple means for spreading spectrum of these UWB low-duty-cycle pulse trains. A TH sequence was used to eliminate catastrophic collisions in UWB multi-user communication systems. 1

In the real environment, there are several factors that affect the transmission and lower the performance also. For example, multi-path, environment location (indoor environment), other signals interferences etc. 1.1 Objectives In this project, the performance of two time hopping modulation schemes, namely time hopping binary phase shift keying (TH-BPSK) and time hopping pulse position modulation (TH-PPM) were compared. The processing gains (Nh, Ns) were adjusted. For different combination of processing gains, performance was different. The most suitable processing gains that had better performance were found in this project also. 1.2 Report Outline Chapter 2 is a background chapter on IEEE 802.15.4a and UWB-IR. In chapter 3, TH is discussed in detail, with the two modulation schemes, TH-PPM and TH- BPSK. Signal model, channel model and receiver model are discussed and comparison of these two schemes is done. Theoretical analysis of TH is discussed in detail and comparison with simulation result is done also. After analysis of theoretical case, other factors that need to be considered in the real environment are discussed in chapter 4. In chapter 5, methodology for this project is given. In chapter 6, performance (BER) of TH-PPM and TH-BPSK are investigated by simulation using MatLab as a tool. Parameters are varied to observe any effects. Discussion is done in this chapter, which is focused more on variation of processing gains since processing gains can be a factor for better performance. Chapter 7 is a conclusion chapter. 2

Chapter 2: Background This chapter provides background information on IEEE 802.15.4a and UWB- IR. Definitions of IEEE 802.15.4a and UWB-IR are described, and also the advantages of using UWB-IR are given. 2.1 Description of IEEE 802.15.4a 2.1.1 Definition of IEEE 802.15.4a According to [1], the IEEE established the IEEE 802.15.4a to define a new physical layer concept for low data rate applications utilizing UWB technology at the air interface. This standard addresses new applications that require only low to moderate data throughout, but long battery life such as low-rate WPAN, sensors and small networks, as discussed in chapter 1. 2.2 Description of UWB-IR 2.2.1 Definition of UWB-IR UWB-IR is a signal occupying more than 25% of the fractional bandwidth or more than 1.5 GHz [2, 3]. Fractional bandwidth can be defined as follows B f f h i = 2 --- (2.1) f h + f f i where fh is the highest frequency in the allotted band and fi is the lowest frequency in the allotted band. UWB-IR is a form of spread spectrum signalling technique and is very narrow in the time domain because of the inverse relationship between time and frequency. According to the current Federal Communications Commission (FCC) guidelines, Part 15(d) power limits are imposed on UWB-IR transmission between frequencies 3

3.1 to 10.6 GHz. For frequencies below 3.1 GHz, the signal needs to be attenuated by approximately 15-30 db over existing Part 15(d) limits. This is done to ensure that the UWB signal does not interfere with other frequencies. UWB-IR is different from other wireless communication technologies. Short pulses (impulses) are transmitted compared with modulating carrier signal in conventional techniques. So UWB-IR is a carrierless signal or impulse radio. [18,19] In Figure 2.1, it is observed that the power can be transmitted in narrowband system is much higher that UWB signals and the frequency range of UWB is wider than narrowband signal. Figure 2.1 UWB spectrum compared to narrowband signal 4

2.2.2 Advantages of using UWB-IR There are many advantages of using UWB-IR technology. The first one is higher bit rate. Nowadays, the demand for the Internet and multimedia services is increasing sharply. Existing technologies can t provide enough bit rates. More bandwidth is a must for achieving higher bit rate. Thus UWB-IR signals is the best choice which is about 100 times more than most conventional signals. The second advantage is multi-path resistance. Multi-path fading is due to destructive combining of several copies of the same signal at the receiver. In order for multi-path effects to occur in UWB-IR systems, interference from the same pulse or interference from the other pulses of the same user must be considered. The third advantage is the low cost. The transceiver and receiver structure are simple and can be manufactured at very low cost. 2.2.3 Applications of UWB-IR As UWB-IR has many advantages, it has many applications. Some of the applications are shown below. 1. WPAN: UWB-IR is considered as the physical layer for the IEEE 802.15 WPAN stand. Figure 2.2 by Intel Architecture Labs [4] compares the existing technologies and UWB in terms of spatial capacity. 2. Precise geo-positioning: the location accuracy is as high as 1 cm. This can be used to detect location of person. It can also be used in conjunction with GPS because of its high accuracy. [5] 3. Sensors: Electronic hand tools, liquid level sensors are commercially available [6]. 5

Figure 2.2 Spatial capacity comparisons 2.3 Summary In this chapter, definition of IEEE 802.15.4a and UWB-IR are described. UWB-IR is carrierless and has very wide in bandwidth as compared to narrowband systems. UWB-IR offers many advantages, especially in WPAN. 6

Chapter 3: Time Hopping (TH) 3.1 Description of TH 3.1.1 What is TH? TH is a kind of multiple access method. It can be illustrated by the following figure. Tc 2Tc Tf Figure 3.1 Illustration of Time Hopping 2Tf From the graph shown above, each packet consists of several bits, which further composed of several pulse. Each pulse lies within one frame time. That is, in Figure 3.1, let s say there are two pulses for one bit and the first pulse is in between 0 to Tf and the second pulse is in between Tf and 2Tf. Each frame is further divided in several chips. In TH, a pseudorandom code (i.e. TH code) is assigned for each pulse. The range of this code is limited to Tf and should be an integer. The corresponding pulse will shift according to the code. For example using the above figure, if the code is {4, 5}, the first pulse is located with the 4 th sub-slot in the first Tf, and 5 th sub-slot in the second frame. For more than one user, different TH code will be assigned for them. This really reduces the probability of collision and to improve the performance of the system. 7

3.2 TH modulation schemes Two modulation schemes are introduced namely TH-BPSK and TH-PPM. 3.2.1 Time Hopping - Pulse Position Modulation (TH-PPM) [7] TH-PPM is a very popular scheme in UWB systems because optimum use of transmitted power has to be there. In this the pulse sent early or late depending upon data to be transmitted. The signal model is given by s ( k ) ( t) = E N b s Ns 1 j= 0 w( t jt f c ( k ) j Tc δd ( k ) ) --- (3.1) where w(t) is a transmitted pulse, T f is the pulse repetition time or frame time, 8 (k ) c j is the time hopping (PN) sequence of the k th user, δ is modulation index, Tc is the time-shift unit incurred by a PN sequence, (k ) d is the binary (0 or 1) symbol of the k th transmitter. When there are multiple pulses per bit we can represent this as j / Ns for any arbitrary j th bit. Equation (3.1) consists of the following [8]: - Uniform pulse train: A pulse train consists of monocycle pulses T f seconds spaced in time (refer Fig. 4.1). As shown in the figure, a packet is comprised of many bits, which further consists of many pulses. A pseudorandom time shift is added to each pulse. This frame time or pulse repetition time could be a hundred to a thousand times the pulse width. Typically, the pulse has a low duty cycle and high repetition rate. One can imagine a frame in which there are thousands of slots and a user will transmit its pulse in one of these slots. Multiple-access can be provided if different users transmit in different slots assuming a synchronous system. Since the duty cycle is very low, the probability of collisions becomes large.

- Pseudorandom time-hopping: To eliminate collisions, each user is assigned a unique pseudorandom time hopping code. This code is periodic with period Np, i.e., (k ) c j +i Np = c. Each code element is an integer in the range 0< c (k ) j is the highest number for which hopping is allowed. The time hopping code, (k ) j < Nh, where Nh therefore, provides a shift to each pulse in the pulse train between 0 and Nh Tc (refer to Fig. 4.1). Take an example to illustrate this point. Let the time hopping code for a given user be {1,5,3,2, }. The pulse is sent in the first slot of the first frame. In the second frame for that user, the pulse is transmitted in fifth slot, in the third slot in the next time frame, and so on. For this reason it is called time hopping. - Data Modulation: The data sequence (k ) d j of the k th transmitter is a binary (0 or 1) symbol stream that conveys information. Usually, this is an oversampled modulation system with Ns monocycles transmitted per symbol. But, for gaining higher data rates, Ns can also be equal to one and the duty cycle can be higher. Multiple pulses per bit are needed particularly when the interference from other users or from the channel is significant. Using the PPM modulation method, when the data symbol is 0, no additional time shift is added, but a time shift of δ is added to the monocycle when the symbol is 1. Other forms of data modulation can be employed to benefit either the performance of the synchronization loops, the interference rejection, and the implementation complexity and so on. 9

Figure 3.2 Illustration of TH-PPM 3.2.2 Time Hopping Binary Phase Shift Keying (TH-BPSK) The signal model is given by s ( k ) Ns 1 Eb ( t) = d N s j= 0 ( k ) w( t jt f c T ) ( k ) j c --- (3.2) As compared to TH-PPM (refer to equation (3.1)) in which we shift the pulse by δ depending on the data symbol; in this scheme we send a positive or a negative pulse at the same position. As shown in the Fig. 3.2 a packet comprises many bits, which further consists of many pulse. A pseudorandom time shift is added to each pulse. The data modulation is done by inverting the pulse. The benefit of this scheme over TH-PPM is higher immunity to 10

noise. Also, TH-BPSK will be more robust to timing jitter as compared to TH-PPM, since in TH-PPM, the pulse if sent early or late depending on the data. Figure 3.3 Illustration of TH-PPM 11

3.3 Channel Model The channel model is given by [9] h( t) = L 1 j= 0 α ( t) δ ( t j jt w ) --- (3.3) where α (t) is the channel gain of the path j, Tw is the length of the pulse width (i.e. j resolution bin) and it consists of L paths. In this project, the transmitted pulse is same width as Tw and equal to one chip time. Rayleigh channel is selected in the simulation. 3.4 Receiver Model The receiver model is given by [9] Nu Ns 1 L 1 Eb u ( k) r( t) = hij w( t itf ci Tc δd N s k= 1 i= 0 j= 0 ( i) jt w ) + n( t) --- (3.4) where u h ij denotes the channel gain of path j for user u and n(t) is the Additive White Gaussian Noise (AWGN). 3.5 Theoretical Analysis The theoretical BER is given by [10] P e = 1 2 erfc 2 SNR + 1 2 K T T f p N s --- (3.5) where K is number of users, Tp is the pulse duration, Tf is the frame time and Ns is number of pulses used for one bit. 12

3.5.1 Comparison of theoretical BER and simulation BER Simulation is done to verify that the performance using equation (3.5) can be done by simulation. To be fair, the simulation is done in AWGN channel and the result is shown in the following figures. Figure 3.4 Comparisons of Simulation and Theoretical Result From the figure above, it shows that the simulation result (blue line) is closed to theoretical one (red line) under AWGN channel. 13

3.6 Summary In this chapter, TH is discussed in detail. Two modulation schemes have been discussed in detail, namely, TH-PPM in which a pulse is sent early or late, TH-BPSK in which a pulse is inverted at the same position. Signal model, channel model and receiver model are described. Comparison between the simulation value and theoretical value is also described in this chapter. 14

Chapter 4: Considerations in the real environment In this chapter, the transmitter and receiver structures are described. The main advantage of these structures is low cost. To obtain the best performance, RAKE receiver is considered in this project. Also, different pulses have different performances. The transmission medium/ channel is also discussed in which Rayleigh channel is used in this project. 4.1 Transmitter structure The transmitter structure can be shown in the following figure [11] Pulse Generator Code Generator Time Delay Modulation Clock Oscillator Data (bit) Figure 4.1 Transmitter structure The transmitter has no power amplifier since the power of UWB-IR is very low. Time delay is used for time hopping codes generation for the pulses. Modulation can be implemented by changing the position of the pulse or changing the amplitude of the pulse. The clock oscillator should be as precise as the pulse generation, which is of 1 ns. 4.2 Receiver structure 15

There are several receiver structures can be used. In this project, RAKE receiver is chosen and the structure of RAKE is shown below [12]. g ( t 1 ) T () dt 0 r 1 r(t) g ( t 2 ) T () dt 0 r 2 Decision Algorithm decision (t) g k T () dt 0 r k Figure 4.2 RAKE receiver structure In RAKE receiver structure, there are enough fingers for detection. For the received signal, it contains several copies of transmitted signal with time delay. By correlated with the conjugate of the channel gain for each signal path and with the time hopping code, the signal can be detected. Since more signals can be detected with more paths accurate, the performance will be better. However, if the RAKE receiver can t have enough fingers, the performance will be poorer because limited number of signals is detected. 16

4.3 Selection of pulse There are many methods for generating UWB signals. In the simulation of this project, a rectangular pulse with pulse width equal to chip length, Tw. For pulse width greater than Tw, it may overlap with other pulse for different users, which affect the performance. For pulse width smaller than Tw, in multi-path situation, two pulses or more may lie within Tw, which also lower the performance. As a result, the pulse selection must be very careful. Furthermore, the time hopping codes must also be considered carefully to try to prevent any collision or interference with other pulses. 4.4 Transmission medium/ channel The indoor channel mode is difficult to characterize. It has been observed that propagation within the building is dependent on the layout of the building and the construction materials used. Refraction, reflection and scattering of signals can be occurred inside the building and it suffers from heavy multi-path fading. The channel selected in this project simulation is Rayleigh channel. For the theoretical comparison, AWGN channel is also considered. This is the simplest channel model. The transmitted signal will experience some attenuation and delay in this channel. 4.5 Summary In this chapter, the transmitter and RAKE receiver structure are described. For the selection of pulse, it should be considered along with the time hopping codes and number of users to prevent collision problem. For transmission medium, since Rayleigh channel is the worst case of multi-path fading, it is considered in this project simulation. 17

Chapter 5: Methodology In this project, the methodology and the steps are discussed in detailed in the followings. Firstly, so many journals and information related to the topic of ultra-wideband were read. Through reading the abstracts of the journals, the main ideas of UWB were understood. Since there are many areas to discuss UWB in different methods such as modulation techniques and different forms of UWB such as Impulse Radio, focusing on one or two areas is the next step needed to do. After selecting the focused area, detailed understanding in this area is needed and this is also an important step. As impulse radio, TH-PPM and TH-BPSK are the three major areas focused in this project; I selected two journals for detailed reading, one about impulse radio and the other is about TH-PPM and TH-BPSK. The characteristics of these were studied, e.g. how the signal transmitted and received in different modulation techniques, what the transmitted waveform look likes etc. Furthermore, the performance is also studied since the performance comparison and evaluation for multiple access techniques are the main focus in this project. After studying the focused area, the simulation is the next step. Simulation will be done by MatLab. During studying the selected materials, only theoretical information and simulation result done by the writers were known. Although their simulation results were true, many unrealistic assumptions were made. Through simulation, I can prove whether the simulation result is correct or not and at the same time, we can find out the weaknesses of previous simulation and try to improve it. 18

Finally, after finishing all simulations on TH-PPM and TH-BPSK, analysis was the next step. Comparison and observation of the graphs are done. Any irregularities and similarities were recorded and the reasons of these were found. If these are man-made error, corrections can be made. By variation of different parameters, performances will be affected also. This may be a critical point for the design of transceiver structure. 19

Chapter 6: Simulation Result In this chapter, simulation results under three main situations will be shown. All the simulation is achieved by using MatLab. The three considerations are variation of number of users, number of paths transmitted (i.e. multi-path) and processing gains (Nh and Ns). For each consideration, TH-BPSK and TH-PPM will be considered separately. By comparison of the two graphs, performance of TH-BPSK and TH-PPM is discussed also. Performance comparison and reason for such a graph is shown after each consideration. 6.1 Variation of number of users (Nu) The key features in this simulation are shown below: 1. Processing gain, Nh = 20 2. Processing gain, Ns = 8 3. Rayleigh fading 4. Number of bits transmitted each time = 1000 bits 5. Number of paths = 2 6. Rake receiver with enough number of fingers for detection 7. Number of synchronous users = 2, 8 20

TH-PPM The simulation result is shown by the following graph. Figure 6.1 TH-PPM Variation of Number of Users, Nh=20, Ns=8 Blue line represents two users and red line represents eight users. It is clearly observed that the performance is better for two users. When number of users increase, the performance is getting worse. 21

TH-BPSK The simulation result is shown below. Figure 6.2 TH-BPSK Variation of Number of Users, Nh=20, Ns=8 Blue line represents two users and red line represents eight users. The trend is as same as TH-PPM. It is clearly observed that the performance is better for two users. When number of users increase, the performance is getting worse. Reason for more users lead to poor performance For more number of users, probability of collisions is increased. Also, time hopping codes for each user may overlap. If more users use the same time hopping code, errors are introduced more. This leads to poor performance. To overcome this problem, processing gains, especially Nh, should be increased to prevent any collisions between users. 22

Comparison of TH-PPM and TH-BPSK From the two graphs on variation of number of users, it is shown that TH-BPSK outperforms TH-PPM. Let s take an example, for SNR of 10 db and two users, the BER of TH-PPM is about 10-2, whereas for TH-BPSK, the BER is closed to 10-3. TH-BPSK performs better. This is because for TH-BPSK, the receiver detects both phases of the signal (1 and -1), whereas in TH-PPM, only one phase detection is needed. As more error may introduce for one phase detection and noise presents, TH-PPM is worse than TH-BPSK. In Mathematical explanation, the energy needed for TH-BPSK is but the energy for TH-PPM is 2E, which is much smaller. 2 E b b 6.2 Variation of number of paths (Npath) The key features in this simulation are shown below: 1. Processing gain, Nh = 20 2. Processing gain, Ns = 8 3. Rayleigh fading 4. Number of bits transmitted each time = 1000 bits 5. Number of users = 2 6. Rake receiver with enough number of fingers for detection 7. Number of paths = 2, 3 23

TH-PPM The simulation result is shown by the following graph. Figure 6.3 TH-PPM Variation of Number of Paths, Nh=20, Ns=8 Red line represents three paths and blue line represents two paths. From the graph, it is observed that when number of paths increased, the performance is better. 24

TH-BPSK The simulation result is shown by the following graph. Figure 6.4 TH-BPSK Variation of Number of Paths, Nh=20, Ns=8 Same as TH-PPM, red line represents three paths and blue line represents two paths. From the graph, it is observed that when number of paths increased, the performance is better. Reason for more paths leads to better performance As the receiver structure is RAKE receiver, which have enough fingers for detection for each signal transmitted in the corresponding paths. The receiver can combine all the signals received from each finger and determine the sending bit. For more users, there will be more copies of signal and the detection will be more accurate. Thus, more paths will lead to better performance. 25

Comparison of TH-PPM and TH-BPSK It is the same reason shown in the variation of number of user part. 6.3 Variation of processing gains (Nh, Ns) The key features in this simulation are shown below: 1. Rayleigh fading 2. Number of bits transmitted each time = 1000 bits 3. Number of users = 2 4. Rake receiver with enough number of fingers for detection 5. Number of paths = 2 6. Variation of Nh, Ns = Nh*Ns = 100 7. TH-PPM Case 1: Nh = 10, Ns = 10 Case 2: Nh = 5, Ns = 20 Case 3: Nh = 20, Ns =5 26

The simulation result is shown by the following graph. Figure 6.5 TH-PPM Variation of Processing Gains, Nh*Ns=100 From the graph, it is observed that based on the performance of Nh = 10, Ns = 10, when Nh is increased and Ns is decreased, the performance is better. On the other hand, performance is worse when Nh is decreased and Ns is increased. Reason for such results In this graph, the SNR is ranged for 0 to 16 db. Nh*Ns should be equal for fair comparison. For larger Nh, the performance is better. It is because the when chip size (Nh) is large, collision frequency is smaller. Thus improve the performance. For larger Nh (i.e. smaller Ns), collision occur more frequently. However, for the red and blue lines, it seems that these two lines are closer starting from 12 db, so these two lines should be overlapped after certain increase of SNR. As a result, Nh and Ns should have a trade off and thus the design of Nh and Ns value should be 27

assigned properly. It is not good when Nh is set too high and Ns is set too low. If Ns is small, the detection is just based on this symbol. If it is wrong, error introduced. So, Ns should not too small. 6.4 Summary In this chapter, we have shown several simulation results according to the key features set during the simulation. It is clearly shown that TH-BPSK outperforms TH-PPM in all consideration. Receiver structure is needed to consider since it may affect the performance. RAKE receiver is used in this simulation is based on MRC and have enough fingers. This can improve the performance of the received signal accuracy. Fewer fingers may lead to poor performance also. For different combination of processing gains, higher Nh leads to better performance. However, it must be considered that the SNR values also. For higher SNR, large Nh may not lead to so good performance. 28

Chapter 7: Conclusion In this project, performance comparison of two modulation schemes, namely, TH- BPSK and TH-PPM over Rayleigh channel. The BER performance was analyzed for the two modulation schemes in presence of multi-user interference and multi-path interference. It is observed clearly that for any number of users and number of paths, TH-BPSK outperforms TH-PPM. This can be explained Mathematically and non-mathematically as discussed in the Simulation Result chapter. TH-PPM needs to increase SNR or adjust the processing gains to achieve same performance as TH-BPSK. For variation of number of users, it is shown that more users lead to poor BER performance because of increasing the chance of collision. For variation of number of paths, it is shown that high number of multi-paths lead to better performance. This is valid only for the receiver structure is RAKE receiver, which has enough number of fingers for detection of different copies of signal. For variation of processing gains (Nh, Ns), it is shown that better performance for larger Nh. However, there should have a tradeoff between Nh and Ns. Although it is true as shown in the Simulation Result chapter, but for even larger SNR, it can easily predicted that the trend may not the same. 29

References [1] Ian Oppermann, Matti Hamalainen and Jari Iinatti (2004) UWB Theory and Applications, pp7 [2] C. Fowler, J Entzminger, J. Corum, Assessment of Ultrawideband(UWB) technology, IEEE AES magazine, Vol 5, Issue 11, Nov.1990, pp45-49 [3] http://www.uwb.org [4] G.L. Stuber, Principles of Mobile Communication, Kluwer Academic Publishers, 2 nd edition, 2001 [5] http://www.aetherwire.com [6] http://bwrc.eecs.berkeley.edu/ [7] N. Morinaga, M. Nakagawa, R. Kohono, New concepts and technologies for achieving highly reliable and high capacity multimedia wireless communications systems, IEEE communications magazine, Jan. 1997, pp28-33 [8] M.Z. Win, R.A. Scholtz, Impulse Radio: How its works, IEEE communication letters, Vol 2, Issue 2, Feb 1998, pp36-38 [9] S.H. Song, Q.T. Zhang, TH-CDMA-PPM with Non coherent Detection for Low Rate WPAN, IEEE communication letters [10] Khairi Ashour Hamdi, Xuanye Gu, Bit Error Rate Analysis for TH-CDMA/PPM Impulse Radio Networks, IEEE communication letters, 2003, pp167-172 [11] K. Siwiak, Ultra-wide band radio: introducing a new technology, IEEE 53 rd VTC Spring 2001, Vol 2, pp1088-1093 [12] X. Shen, M. Guizani, R. C. Qiu, T. Le-Hgoc (2006) Ultra-wideband wireless communications and networks 30

Appendix A: Program code used for simulation Code for generating time hopping codes for u = 1:Nu for s=1:ns TH(u,(Ns-1)+floor(Ns.*rand(1))+1)=1; end end Code for generating channel gain in Rayleigh channel Rayleigh(i)=(1/sqrt(2))*(randn(1)+j*randn(1)); Code for detection of received data d=sum(rx_data); if d>0 d=1; else d=-1; end if(d ~= user(1)) error = error+1; end 31

Code for calculating the BER p=error./sim; ber(snr)=p; Code for plotting BER versus SNR semilogy(snr,ber) xlabel(snr) ylabel(ber) 32