Complementary Unipolar Duty Cycle Division Multiplexing (CUDCDM): a Novel and Economical Multiplexing Technique

Transcription

1 International Review on Computers and Software (l.re. CO.S.), Vol 3, N. 2 March 28 Complementary Unipolar Duty Cycle Division Multiplexing (CUDCDM): a Novel and Economical Multiplexing Technique M. K. Abdullah, A. Malek Mohammadi, A. F. Abas, G A. Mahdiraji, M. Mokbtar Abstract ~ A new multiplexing technique based on duty cycle division is proposed, under the name: Complementary Unipolar Duty Cycle Divi.-iion Multiplexing (CUDCDM). CUDCDM can be applied in both electrical and optical domains. The new technique allows for more efficient use of time slots as well as the spectrum taking advantage of both the conventional time division multiplexing TDM and FDM. The basic properties based on theoretical analysis as well as simulation studies have been done to evaluate the performance of this technique ba.sed on the signal energy and symbol error rate (SER). In this paper the performance of complementary unipolar duty cycle division multiplexing compared with multi level M-aiy signaling as well as the conventional (TDM) technique. The simulation has been set for wireless transmi.ssion based on free space propagation model with adaptive white Gaussian noise (A WGN). PSK and QAM are used as modulation schemes to evaluate these techniques against data rate and numher of users. The study shows that the energy per bit in CUDCDM, unlike that of TDM technique, increases with the number of users. The simulation results correspond with the theoretical study in which the CUDCDM technique has better SER than that of TDM even in more distances. Copyright 28 Praise Worthy Prize S.rJ. - Ali rights reserved. Keywords: Multiplexing technique. Duty cycle, Wireless systems I. Introduction Multiplexing is one ofthe fundamental and essential parts in today's digital communications. The expected growth in demand for existing narrowband services and fiiture broadband-interactive, multimedia-entertainment, and educational services has led to a need for high capacity networks. The success and the increasing diffusion of wireless system have made band width a scarce resource. Therefore, efficient use of limited available band width is mandatory. Whenever the transmission capacity of a medium linking two devices is greater than the transmission needs of the devices, the link can be shared, much as a large water pipe can carry water to several separate houses at once. Multiplexing is the set of techniques that allows the simultaneous transmission of multiple signals across a single data link. There are several ways in which multiplexing can be achieved stich as Freqnency Division Multiplexing (FDM) [l]-[3]. Time Division Multiplexing (TDM and OTDM) [4]-[6], Code Division Multiplexing (CDM and OCDM) [7]-[8] and multilevel M-ary signaling. The goals of all multiplexing techniques are to support as many users at as high speed and at the lowest cost possible [9]-[]. Multilevel M-ary is normally used to transmit multiple amplitudes, each one representing two bits or more. The main purpose of M-ary is to increase the bandwidth efficiency whereby, many bits can be transmitted by the signal symbol. If each bit is assumed to represent a different user, then M-ary can also be used for multiplexing. The main issue in multilevel M- ary is the voltage levels which is increased as 2"(n is the number of users), that makes the system impractical for high number of users. DM is the most widely used multiplexing technique in today's communication. However, for multiplexing high number of users with high data rates, high speed multiplexer and de-multiplexer are required, resulting in very high cost for TDM systems. At high speed, clock recovery is another essential issue that may render the system highly complicated and costly for TDM systems []. Therefore, many investigations have been done to design and develop reliable and cost-effective clock recovery modules for TDM in both electrical and optical versions [II]. Realizing these problems, in this paper complementary unipolar duty cycle division mulliplexing is proposed as an alternative multiplexing technique. In this technique, different users can share conimtinication medium and transmit data simultaneously by using the same frequency band, but with a different duty cycle. At the receiver side, based on the signal amplitude and duty cycle, the original data can be easily distinguished from the received signal. The proposed technique also has an inherent property which allows for better clock recovery (however, this is not within the scope of this paper). Manuscript received and revised February 28, accepted March 28 Copyright 2S Praise Worthy Prize S.r.l. - All rights reserved 2

2 M K. Abdullah, A. Malek Mohammadi, A. F. Abas. G. A. Mahdiraji. M. Mokhtar The purpose of this paper is to introduce the new multiplexing techniques based on duty cycle division and compare this multiplexing technique with multilevel M-ary as well as with TDM. In section II ofthe article base on theoretical analysis the basic properties of CUDCDM are explained. The results from the simulation study are discussed in subsequent section which is followed by a conclusion. II. The Working Principle //. /. Basic Properties The Complementary Unipolar Duty Cycle Division Multiplexing (CUDCDM) is based on having each channel modulated with a unique RZ duty cycle. In this technique each multiplexing user transmits bit '' with zero volt and bits ' ' with +A volts. Based on the linear distribution of duty cycle the i"'' multiplexing user transmits bit T within T second which is calculated as: () where n represents number of multiplexing users, and Ts is the symbol duration. The first user has the shortest pulse width which is calculate using equation (), when i= and the n''' user has the longest pulse width, when i= '«'. For example, for multiplexing three users, the P', 2"'' and 3"^ user uses duration of (75/4), (2Ts/4) and (3 Ts/4) respectively to transmit bit I s. Therefore, different users share the communication medium to transmit in the same time and the same frequency but different duty cycles. Table I shows an example of eight possible combinations of bits for three users. Figs. l(a),(b).(c), represents the signals with (Ts/4), i2t.-,-/4) and {its/4) duty cycle respectively, represents as different users. Fig. l(d) shows the combination of three u.sers and Fig. l(e) shows the pattems when the 3 users are complementary multiplexed with the same period of Ts. The multiplexed signal can then be modulated by using a single modulator. At the receiver side, after demodulating the received signal, the original signal can be extracted from the demodulated signal based on the signal voltage amplitude and the duty cycle's time duration. Based on the 2" possible bits combination (Table I), each of these combinations produce a unique symbol for the complementary multiplexed signals. Having the knowledge about this uniqueness at the receiver side, the original data for each user can be easily distinguished and recovered by taking one sample per slot for '/?+/' slot per Ts seconds. This technique allows for automatic bit error detection and correction based on the sequence of sampled amplitudes per symbol duration. For the case of multiplexing 'n' users, if only one sample per slot is taken, then, the first sample (taken from the first slot), has '«+/' possible levels, the second sample {taken from second slot), has 'n' possible levels, the n"' sample has oniy two possible levels { or A volts), and the last sample has one possible level which is volt. Note that the maximiun amplitude ofthe multiplexed signal is: (2) The minimum amplitude that the multiplexed signal may take is volt. The minimum amplitude only happens, when all users transmit bit I. For example in Fig., case 8 has the minimum amplitude of volts in the first slot. There are '«+7' number of slots per symbol in the multiplexed signals. All of these slots have an equal duration that can be calculate using Eq. (3): Tslotk = TABLI-: EiGirr POSSIBLE COMBINATIONS OF BITS FOR 3 USERS Case Userl User2 UKer3! 2 3 Fig.. (a), (b), (c) 3 multiplexing users with duty cycle of 25%.5O% and 75%Ts for user, 2 and 3 respectively and (d) summation of 3 users signal and (e) is complementary multiplexed signal for 3 users 4 I //. 2. Signal Energy (3) Average energy/power per bit can be calculated from energy power spectrum density (E/PSD) or more simply, it can be calculated by taking the average energy/power per symbol for all the 2" possible combination of bits. The later method was utilized in this paper. The average energy per bit for this method is formuhzed as: Copyright 2S Praise Worlhy Prize S.r.l. - AH rights resen-ed International Review on Computers and Software. Vol. 3. M 2 2

3 M. K. Abdullah, A. Malek Mohammadi, A. F. Abas,. A. Mahdiraji, M. Mokhtar Eavg/bit = 6X(H+) where M' is the signal amplitude in volt, 'Ts'' is signal symbol duration in second, '?;' is number of users and 'jfc' is equal to(l, 2... w). Unlike TDM the energy per bit in CUDCDM will be increased linearly as the number of multiplexed user increased. This is important as it helps to improve the signal transmission quality at a larger number of users. Fig. 2 illustrates an example of calculating average energy per bit for the case of multiplexing 3 users, the energy per bit for symbols are: (4) Two cottiparisons are done which are based on the number of signal voltage levels and on the average energy per bit versus number of users. Fig. 3 shows the theoretical results for the signal voltage level versus number of users for the CUDCDM, TDM and M-ary techniques. The comparison shows that the number of voltage levels in the TDM remained fixed to two levels. For the CUDCDM technique it increased as (n+l), and for the M-ary it increased as 2". Thus in order to multiplexed 7 users, the TDM, CUDCDM and M-ary techniques require 2, 8 and 28 signal voltage levels respectively. This result shows the disadvantage of the M-ary technique which makes it impossible to use as a multiplexing technique for high number of users, there for the next comparison was continued between the TDM and CUDCDM only. 4/ I. 4 and the average energy per bit will be: Eavg = 3A 3*8 El ti:i34S67892 Number of users Fig. 3. Number of signal voltage levels versus number of users for the M-ary, TDM and CUDCDM techniques HI - F.2 Next comparison is done based on the theoretical results for the signal energy; the average energy per bit for the CUDCDM is calculated based on Eq. (4) as presented in Fig. 2. For the time division multiplexing also same method applied, which is detlned as: IJ.3. m J - mm Fig. 2. Complementary multiplexed symbols for the 8 cases Comparison of CUDCDM with Multilevel M-ary and TDM Techniques M-ary signaling is normally used to transmit mtiitiple amplitudes, each one representing two bits or more. The main purpose of M-ary is to increase the bandwidth efficiency whereby, many bits can be transmitted by the signal symbol. If each bit is assumed to represent a different user, then M-ary can also be used for multiplexing, thus worthy of comparison with the new scheme. Eavg / bit =\ 2 (5) Fig. 4 shows the average energy per bit of TDM and CUDCDM for multiplexing up to users. Ihe figure shows that the average energy per bit of CUDCDM increased linearly with the number of multiplexing users whereas the average energy per bit for DM reduced. So CUDCDM is able to suppori higher number of multiplexing users and also transmit in more distances compared with the TDM considering the energy per bit only. Other advantage of CUDCDM including simple transmission design is capability of better clock recovery. For all possible symbols, there are transitions in the first and middle of each symbol except for the case that all users transmit bit '' which probability is low (/2"). The transition is very useftil to recover the receiver clock (synchronization); especially the Copyright 2H Praise Worlhy Prize S.r.l. - All rights resetted International Review on Computers and Software. Vol. 3. N. 2 22

4 M K. Abdullah. A. Malek Mohammadi. A. F. Abas, G. A. Mahdiraji, M. Mokhtar transition that occurs in the first of each slot can be used for symbol recovery. The elaboration of this feature of CUDCDM is not within the scope of this paper. As far as transmitter design is concemed, the OTDM technique require one modulator for each user [2] and the use of multiple modulator is costly and lead to cross-talks [5] but, in CUDCDM only one modulator for all users is being applied. This is much cheaper and at the same time it avoids cross talk problems. antenna gain,'c' is the speed of light which is 3** m/s [5], W is the distance between transmitter and receiver in meters, '/ is the carrier frequencies. 'Pt* is assumed to be the same as baseband signal over which is calculated from the energy content of the multiplexed signals shown in Eqs. (4H5) for CUDCDM and TDM respectively. Transmitter Receiver Energy per bit versus number of users user User 3 3 Si : [ +f Vicdium, SHR Dt voll A'n v<iils k. c 4 c. c d / Rased or ^Ilplitu<li & duty cyck Sampler c tc (. c User T 2 User S 9 3 Fig. 4. Normalized average energy per bit versus number of users for CUDCDM and TDM III. Simulation Study A simulation study has been done in order to compare the performance of the CUDCDM with the TDM technique. The schematic diagram of the CUDCDM is shown in Fig. 5. In the transmitter side, the duty cycle of each user changed based on some algorithms. Users with different duty cycles are combined together and then., the complementary of that signal will transmit. It was assumed in this study that the medium is wireless and the entire configuration for wireless transmission with adaptive white Gaussian noise (AWGN). M-PSK and M-QAM was applied as modulation schemes, ///. /. Simulation Setup The parameters used in this simulation study are presented in Table II, applicable for both the CUDCDM and TDM techniques. The simulation was done based on two main important factors which are attenuation and noise. As the transmitter and receiver antenna were assumed to be in light of sight, therefore, the free space propagation model was used to calculate the attenuation in the communication media calculated based on the Friis free space equation [3], [4] which is: Pr [4x piyx^'xlx where "Pt\ 'Pr' are the transmit and receive power in watts, 'Gf' and '^Gr' are the transmitter and receiver (6) Fig. 5. Block diagram of CUDCDM TABLI-; II DF.SICIN AND PE:RfOKMANCE-:PARAMF':rhRS['{)K Till' SiMri AIION Design Parameter Number of pulses per symbol Numher of multiptaxing users Number of sample per symbol Modulation Scheme Propagation Model Carrier Frequency Transmitter Power Transmitter and receiver antenna No ise parameter Signal voltage amplitude per user Transmitter antenna gain(gl) Rfceivcr aiilcnihiliir) V;iliif 5 Pulses Up to 3 + number of users M-QAM, M-PSK Free Space Propagation Loss(FSPL) 4 GHz Baseband signal power Fixed, line-of-sight K=l.38e-'' j/"k. I-- OdB,Tant-3"K,TO=3"K. 5mV 6 db 2dB The amount of noise that is considered in communication media is calculated by [3], [6]; Fn = KxTnx Bn (7) where 'Pv' is thermal noise power in watts, 'A" is the Boltzmann's constant which is.38*io"~' J/"K, '7,,' is equivalent noise temperature and 'B,,' is the equivalent noise bandwidth. The noise temperature is calculated as illustrated in [3], [7] by: TO{F-l) (8) Copyright 28 Praise Worthy Prize S.r.t. - Ail rights reserwd International Review on Computers and Software. Vol. 3. N. 2 23

5 M. K. Abdullah, A. Malek Mohammadi, A. F. Abas, G. A. Mahdiraji, M. Mokhtar where Tun,, is noise temperature ofthe antenna and *F' is antenna noise figure in db and T,/ is temperature of source. Temperature of antenna and source are assumed to be as 27"^'[3]. The simulation is modeled in MATLAB environment in order to calculate the SER of CUDCDM and TDM with both M-PSK and M-QAM modulation schemes. In MATLAB the authors used direct method for calculating the SER, which means comparing the transmitted and received data (i.e., number of error/number of transmitted bits) III.2. Simulation Result Two characteristics of multiplexing systems are studied. First SER versus data rate and second SER versus number of users for both CUDCDM and TDM techniques. In the first case for CUDCDM, number of multiplexing users are fixed for the simulation setup and data rate varied from 5 to 3 Mbps. The distance between transmitter and receiver is 2 Km. Two modulation schemes of M-PSK and M-QAM are used to modulate the output signal of the CUDCDM multiplexer. Fig. 6 shows the SER versus data rate of the CUDCDM for PSK and QAM modulation schemes. Based on this result, QAM present better SER than PSK in difterent data rates. In the case of number of users, the number of multiplexing users varied from 5 to 3 in the simulation setup. All users are assumed to transmit in equal bit rate of 2Mbps. The distance between the transmitter and receiver is 3 Km. Fig. 7 shows comparison between two modulation schemes for CUDCDM. Based on this resull QAM. present better SER than PSK considering different luiniber of users. In the case of comparison between different multiplexing techniques, as QAM modulation scheme pertorms much better than PSK it's selected as the modulation scheme for comparison between CUDCDM and TDM. By comparing the two multiplexing techniques. Figs. 8-9 shows the result of SER versus data rate and. SER versus number of users for the CUDCDM and TDM technique respectively. Fig. 8, shows comparison between two multiplexing schemes when simulating for 7 users and using 8-QAM as modulation scheme. The distance between transmitter and receiver tor TDM is Km. and for CUDCDM is 4 Km, based on this result CUDCDM has better SER even in more distances compare to TDM. As illustrated in Fig. 9 in the case of number of users, CUDCDM performs better than TDM technique especially at the high number of users. This is because of the higher energy per bit of CUDCDM signals as discussed in section (II.B). Although the symbol error rate is high, the main objective of comparing the performance of the two techniques is nevertheless achieved. CUDCDM_QAM CUOCDU^PSK data rate Fig. 6. SER versus data rate for CUDCDM using 32-PSK and 32-QAM l.ooe+oo l.ooe- l.ooe-2 l.ooe-3 i.ooe-4 l.ooe-5 Number of Users Fig. 7. SER versus number of users for CUDCDM using M-PSK and M-QAM (M=8, 6 and 32) -OOE+OO l.ooe- l.ooe-2 l-ooe-3.e-o4 l.ooe 5 OOE-6 Data Rate per users (Mbps) B TDM_KM CUOCOM 4 KM Fig. 8. SER versus data rate using AM modulation for CUDCDM and TDM Copyright C* 2S Praise Worthy Prize S.r.l. - All rights reserved Internolional Review on Computers and Software. Vol. 3. N. 24

6 M K. Abdullah, A. Malek Mohammadi, A. F. Abas, G. A. Mahdiraji, M. Mokhtar g l.ooe+oo l.ooe- - l.ooe-2 l.ooe-3 l.ooe-4 l.ooe-5 Number of users TDLMOKM CUEJCDM 4 KM Fig. 9. SER versus number of users for CUDCDM and TDM IV. Conclusion In this study the principle of complementary unipolar duty cycle division multiplexing (CUDCDM) technique is discussed by comparing with multilevel M-ary and time division multiplexing (TDM) techniques. Theoretical result showed thai using the M-ary technique as multiplexing technique is impractical because the number of signal voltage level is increased by 2", 'H' is the number of users. Although TDM has the advantage of smaller amplitude levels, its energy content decreases with the number of users. On the other hand, CUDCDM energy content improves with the number of users. Simulation results clearly showed the better performance of CUDCDM than that of TDM for supporting higher numher of multiplexing user and bit rate even in more distances. For CUDCDM as showed in the simulation result, QAM proves to be a better modulation scheme. Although the simulation considers oniy noise and attenuation in the communication media the generality of the transmission performance of CUDCDM in comparison with TDM is maintained. The other inherent advantages of CUDCDM consider for future reports are simpler transmitter, better error detection, correction and better clock recovery. References [I] Wayne Toinasi. AIIVOIKCII EkxliDnic Coniwunkalions Systems 4th Edition, Prentice Hall i'lr. 998 [2] Edmond zahcdi, Di^iial Dtiia commiiiiicaiidii prentice-hall, inc 22 [3] Jayalath. A.D.S., Rajatheva. R.M.A.P,. Ahnieci, K.M. Coded oriliogonat frecjuency division mulliplexing for wireless" Vehicular Technology Conference. 999 IEEE 49thVolume 3, Issue, Jul 999 Page{s): vol.3 [4J R. S. Tucker, G. Eisenstein, S. K. Korotky, "Optical titne division multiplexing for very high bit-rate Mansmi.ssion", Journal. Light wave Technology, Volume 6, Issue, Nov. 988, pp [5] M. D. Spirit, A. D. Ellis, and P. E. Bamsley, "Optical time division multiplexing; Systems and networks," IEEE Commun. Mag., vol. 32, no. 2, pp , Dec [6] F. Babich, "Consideratiotis on adaptive techniques for timedivision multiplexing radio systems," IEEE Trans, on Vehicular Tech.. vol. 48,no. 6. pp , Nov [7] S. A. Aljunid, M. Ismail, A. R. Ramli, B. M. Ali, M. K. Abdullah, "A new family of optical code sequences for speclralamplitude-coding optical CDMA systems", Pholon. Teclmol. Let!.. Vol. 6, Issue, Oct. 24. pp [8] J. A. Salehi, "Code division multiple-access techniques in optical fiber networks Parts I: Fundamental principles,"!eee Trans. Commun.. vol.37, pp, , Aug [9] Joseph c.palais. y^it'r optic communicatk)ns,4"' edition. Prentice Hall,998 [] M. K. Abdullah, M. F. Abdalla, A. F. Abas" Novel and Economical Optical Mulliplexing and Electrical Demultiplexing technique for High Speed Fiber Optics Networks" Wireless and Optical Communications Networks, 27. [llj Mahdiraji, M.K.Abdullah.A.F.Abus," Duty Cycle Division Multiplexing Technique for l^irele.'is Communications "Wireless and Optical Communications Networks, 27. [2] Leon W.Couch II. Digital and analog communication systems Macmiilan publishing company.inc.usa. 993 [3] T. S. Rappaport. Wireless communications: principles and practice. 2"'' Edition, Prentice-Hall, Inc., USA, 22. [4] Harold kolimbiris. Digital communication.systcni-i with.satellite and filler optics applications. Prentice-Hal I, Inc., USA 2OO [5] Bahaa E.A.saleh, Maluin carl Teich. Fundamentals of photonics.]qhn wilay &sons.inc 994. [l(i] A. F, A. Ismail. D. Sandcl, A, Hidiiyat, B, Milivojevic, S. Bhandare, H. Zhang, R. Noc, "2.56 Tbit/s,.6 bit/s/hz, 4 Gbaud RZ-DQPSK polarization division multiplex transmission over 273 km", 9th Optoelectronics and Communications Conference/3rd International Conference on Optical Internet (OECC/COIN2il4i.. PDI'4. Yokohama,.lapan, July 2-6, 24. [7] D. Roddy. Satellite communications 3^ Edition, McGraw-Hill, Singapore,2 Authors' information Amin Malek Mohammadi, received his B.Sc. in Computer Engineering from Central Tehran Univei^sity. Tchian, Iran, in 22 and M.Sc. in Electronic Engineering from Indian Institute of Science (lisc). Bangalore. India in 26, He is I currently working toward the PhD degree in Computer System [-Jiginociing at University ' Putra Malaysia. He is a member of IHEE,..iiiiniinicaiions society and member of International Association of Engineering, Canada. His research interest includes optical fiber communications, photonics devices, multiplexing techniques and wireless communications. am in ma lek_in wieec.org I'tiif Mohd. Khii/ani Abdullah (I'hD) is i:iim;ntly a Professor and Head of the Photonics and Fiber Optic Systems Laboratory, LIniversiti Putra Malaysia (UPM). He obtained his BSc and MSc from University of Missouri at Rollii, USA in 99 and 993 respectively, and PhD from Universiti Malaya in 999. His research interest includes fiber optics devices, non-linear optics, DWDM, OSCDMA systems and digital communications technitjiics. He was awarded the Malaysian national young scientist award for the year 2 for his various contributions in the field of optical fiber communi cat ions. He has won several Lnedals at various international expositions and has successfully filed 9 patents for his inventions and scientific works. Copyright 28 Praise Worthy Prize S.r.l. - All righl.s resened International Review on Computers and Software. Vol. 3. M 2 25

7 M. K. Abdullah, A. Malek Mohammadi, A. F. Abas, G. A. Mahdiraji, M. Mokhtar Ahmad Fauzf Abas, Dr.-lng., MIEEE. received his B.Eng. Degree in Telecommunications Engineering from Ij'niversiti Malaya, Kuala Lumpur (2). He was awarded his M.Sc. degree by Utiiversiti I'utra Malaysia specializes in Communieations and Networks Engineering (22). In 26, at the age of 29, he received his Doctor of Engineering (Dr.-lng.) degree from University of Paderbom, Germany, in the field of Optical Fiber Communications. Currently he is working at the Depanment of Computer atid Communication Systems Engineering, Universiti Putra Malaysia as a lecturer. At the depanment level he is appointed as the Research Coordinator. His research interests are optical fiber communications, photonics devices and optical sensors. Currently, he is a leader and collaborator of several research projects and holds several research grants. He also acts as a consultant to a few projects with the private and govemment agencies. He has authored and coauthored 38 technical papers which include journals articles and conference proceedings. He also has one patent granted and two patents filed. M. Mokhtar (PHD) received the BEng. (Hons.), degree from the University Kcbangsaan Malaysia, Malaysia, in 2 and the PhD. Degree in Electronic Systems Engineering from University of Essex, United Kingdom in 27. She is currently a lecturer in the Faculty of Engineering, University Putra Malaysia. Her research interests lie mainly in the channel eoding and optical comniunieation. Copyright 2QQ8 Praise Worthy Prize S.r.i - All rights reserved International Review on Computers andsofiware, Vol. 3. N. 2 26

8