DC-FREE TURBO CODING SCHEME FOR GPRS SYSTEM

Similar documents
A study of turbo codes for multilevel modulations in Gaussian and mobile channels

Space Time Equalization-space time codes System Model for STCM

Performance Analysis of Multi User MIMO System with Block-Diagonalization Precoding Scheme

Parameter Free Iterative Decoding Metrics for Non-Coherent Orthogonal Modulation

RESOURCE CONTROL FOR HYBRID CODE AND TIME DIVISION SCHEDULING

Impact of Interference Model on Capacity in CDMA Cellular Networks. Robert Akl, D.Sc. Asad Parvez University of North Texas

Resource Allocation Optimization for Device-to- Device Communication Underlaying Cellular Networks

The Performance Improvement of BASK System for Giga-Bit MODEM Using the Fuzzy System

Bit-interleaved Rectangular Parity-Check Coded Modulation with Iterative Demodulation In a Two-Node Distributed Array

Rejection of PSK Interference in DS-SS/PSK System Using Adaptive Transversal Filter with Conditional Response Recalculation

Comparative Analysis of Reuse 1 and 3 in Cellular Network Based On SIR Distribution and Rate

A thesis presented to. the faculty of. the Russ College of Engineering and Technology of Ohio University. In partial fulfillment

Calculation of the received voltage due to the radiation from multiple co-frequency sources

PRACTICAL, COMPUTATION EFFICIENT HIGH-ORDER NEURAL NETWORK FOR ROTATION AND SHIFT INVARIANT PATTERN RECOGNITION. Evgeny Artyomov and Orly Yadid-Pecht

Resource Control for Elastic Traffic in CDMA Networks

Review: Our Approach 2. CSC310 Information Theory

THE USE OF CONVOLUTIONAL CODE FOR NARROWBAND INTERFERENCE SUPPRESSION IN OFDM-DVBT SYSTEM

Markov Chain Monte Carlo Detection for Underwater Acoustic Channels

A Spreading Sequence Allocation Procedure for MC-CDMA Transmission Systems

MIMO-OFDM Systems. Team Telecommunication and Computer Networks, FSSM, University Cadi Ayyad, P.O. Box 2390, Marrakech, Morocco.

Adaptive Modulation for Multiple Antenna Channels

A Comparison of Two Equivalent Real Formulations for Complex-Valued Linear Systems Part 2: Results

Digital Transmission

The Application of Interpolation Algorithms in OFDM Channel Estimation

Phasor Representation of Sinusoidal Signals

Multicarrier Modulation

Approximating User Distributions in WCDMA Networks Using 2-D Gaussian

Power Minimization Under Constant Throughput Constraint in Wireless Networks with Beamforming

arxiv: v1 [cs.it] 30 Sep 2008

Efficient Large Integers Arithmetic by Adopting Squaring and Complement Recoding Techniques

NOVEL ITERATIVE TECHNIQUES FOR RADAR TARGET DISCRIMINATION

A NSGA-II algorithm to solve a bi-objective optimization of the redundancy allocation problem for series-parallel systems

Throughput Maximization by Adaptive Threshold Adjustment for AMC Systems

A NEW TRANSMISSION STRATEGY FOR SCALABLE MULTIMEDIA DATA ON OFDM SYSTEMS

Research of Dispatching Method in Elevator Group Control System Based on Fuzzy Neural Network. Yufeng Dai a, Yun Du b

An Efficient Energy Adaptive Hybrid Error Correction Technique for Underwater Wireless Sensor Networks

Dynamic Optimization. Assignment 1. Sasanka Nagavalli January 29, 2013 Robotics Institute Carnegie Mellon University

Keywords LTE, Uplink, Power Control, Fractional Power Control.

A MODIFIED DIRECTIONAL FREQUENCY REUSE PLAN BASED ON CHANNEL ALTERNATION AND ROTATION

Topology Control for C-RAN Architecture Based on Complex Network

LOCAL DECODING OF WALSH CODES TO REDUCE CDMA DESPREADING COMPUTATION

Performance Analysis of Power Line Communication Using DS-CDMA Technique with Adaptive Laguerre Filters

A new family of linear dispersion code for fast sphere decoding. Creative Commons: Attribution 3.0 Hong Kong License

DVB-T/H Digital Television Transmission and its Simulation over Ricean and Rayleigh Fading Channels

High Speed, Low Power And Area Efficient Carry-Select Adder

Index Terms Adaptive modulation, Adaptive FEC, Packet Error Rate, Performance.

Uncertainty in measurements of power and energy on power networks

The Spectrum Sharing in Cognitive Radio Networks Based on Competitive Price Game

AN IMPROVED BIT LOADING TECHNIQUE FOR ENHANCED ENERGY EFFICIENCY IN NEXT GENERATION VOICE/VIDEO APPLICATIONS

Spatial Multiplexing applied to Turbo Coded Multi-Carrier CDMA

Performance of Modified Iterative Decoding Algorithm for Multilevel Codes in Adaptive OFDM System

The Detection Algorithms Performance in BLAST Enhanced IEEE a WLAN Standard on Measured Channels. University of Bristol

Iterative Detection and Decoding (IDD) MIMO-OFDM HARQ Algorithm with Antenna Scheduling

FFT Spectrum Analyzer

Ergodic Capacity of Block-Fading Gaussian Broadcast and Multi-access Channels for Single-User-Selection and Constant-Power

Exponential Effective SIR Metric for LTE Downlink

New Wavelet Based Performance Analysis and Optimization of Scalable Joint Source/Channel Coder (SJSCC & SJSCCN) for Time-Varying Channels.

Distributed Channel Allocation Algorithm with Power Control

Robust Image Transmission Performed by SPIHT and Turbo-Codes

熊本大学学術リポジトリ. Kumamoto University Repositor

Joint Adaptive Modulation and Power Allocation in Cognitive Radio Networks

High Speed ADC Sampling Transients

Performance Study of OFDMA vs. OFDM/SDMA

Define Y = # of mobiles from M total mobiles that have an adequate link. Measure of average portion of mobiles allocated a link of adequate quality.

Channel Alternation and Rotation in Narrow Beam Trisector Cellular Systems

Energy Efficiency Analysis of a Multichannel Wireless Access Protocol

BER Performances of a Differential OFDM System in Fading Channels

HUAWEI TECHNOLOGIES CO., LTD. Huawei Proprietary Page 1

AN EFFICIENT ITERATIVE DFT-BASED CHANNEL ESTIMATION FOR MIMO-OFDM SYSTEMS ON MULTIPATH CHANNELS

Bit Error Probability of Cooperative Diversity for M-ary QAM OFDM-based system with Best Relay Selection

Performance of Reverse Link CDMA in a Multi-cell Environment with Moving Cells*

ETSI TS V8.4.0 ( )

SC-FDMA and OFDMA: An Efficient Wireless Image Transmission Schemes

On the Feasibility of Receive Collaboration in Wireless Sensor Networks

Joint Power Control and Scheduling for Two-Cell Energy Efficient Broadcasting with Network Coding

EFFICIENT FIELD PROGRAMMABLE GATE ARRAY IMPLEMENTATION OF A CONVOLUTIONAL TURBO CODE FOR LONG TERM EVOLUTION SYSTEMS

A MODIFIED DIFFERENTIAL EVOLUTION ALGORITHM IN SPARSE LINEAR ANTENNA ARRAY SYNTHESIS

Information-Theoretic Comparison of Channel Capacity for FDMA and DS-CDMA in a Rayleigh Fading Environment

A High-Sensitivity Oversampling Digital Signal Detection Technique for CMOS Image Sensors Using Non-destructive Intermediate High-Speed Readout Mode

Walsh Function Based Synthesis Method of PWM Pattern for Full-Bridge Inverter

Revision of Lecture Twenty-One

On Timing Offset and Frequency Offset Estimation in LTE Uplink *

MASTER TIMING AND TOF MODULE-

Near-Optimum Blind Decision Feedback Equalization for ATSC Digital Television Receivers

A NOVEL PREAMBLE DESIGN FOR CHANNEL ESTIMATION IN MIMO- OFDM SYSTEMS RESULTING IN ENHANCED THROUGHPUT

To: Professor Avitabile Date: February 4, 2003 From: Mechanical Student Subject: Experiment #1 Numerical Methods Using Excel

Side-Match Vector Quantizers Using Neural Network Based Variance Predictor for Image Coding

Capacity improvement of the single mode air interface WCDMA FDD with relaying

1.0 INTRODUCTION 2.0 CELLULAR POSITIONING WITH DATABASE CORRELATION

The Throughput of Hybrid-ARQ in Block Fading under Modulation Constraints

Generalized Incomplete Trojan-Type Designs with Unequal Cell Sizes

Error Probability of RS Code Over Wireless Channel

Figure.1. Basic model of an impedance source converter JCHPS Special Issue 12: August Page 13

King s Research Portal

Delay Constrained Fuzzy Rate Control for Video Streaming over DVB-H

An Efficient Secured Turbo Codes for Long Term Evolution System Enhancement

Inverse Halftoning Method Using Pattern Substitution Based Data Hiding Scheme

Low-Delay 16 kb/s Wideband Speech Coder with Fast Search Methods

California, 4 University of California, Berkeley

Comparison of Coding Schemes over FRBS Aided AOFDM Systems

Transcription:

DC-FREE TURBO CODING SCHEME FOR GPRS SYSTEM Prof. Dr. M. Amr Mokhtar & Eng. A. Refaey-Ahmed Electrcal Engneerng Department, Alexandra Unversty, Egypt. ABSTRACT A useful tool n the desgn of relable dgtal communcaton systems s channel codng. Turbo codes have been shown to yeld an outstandng codng gan close to theoretcal lmts n the Addtve Whte Gaussan Nose (AWGN) channel. Recently, they have been studed over Raylegh fadng channel as well and showed a remarkable performance. In ths paper, we propose a DC-Free Turbo codng scheme and Turbo codng scheme to replace the Convolutonal codng already n use n the General Packet Rado Servce (GPRS) system. Evaluaton of the proposed encoder by means of computer smulaton has shown a performance mprovements. I. INTRODUCTION Codng ams at mprovng transmsson qualty when the sgnal encounters any type of dsturbances. Turbo codes wth MAP decodng are one of the most wdely used forward error correctng technques. Ths s due to relatvely large codng gans that can be acheved. Thus t can be expected that the DC-free codes constructed from turbo codes have good error control performance. Turbo codes combne the concepts of parallel concatenated convolutonal codes, nterleavng and teratve decodng. The outstandng performance acheved was unexpected and rased an explosve amount of lterature [1,2]. Turbo code encoder consst as shown n Fgure 1, of two (or more) encoders grouped together n a parallel connecton, and separated by a random nterleaver. Although the component encoder can be of any type (block or convolutonal); typcally, Turbo codes are bult wth systematc recursve convolutonal encoders [3]. The decodng of Turbo codes s typcally accomplshed usng an teratve decodng algorthm. As shown n smplfed form n Fgure 2, there are two component decoders operatng together and passng soft nput nformaton to each other n a feedback loop. The component decoders must be capable to generate soft outputs correspondng to each nformaton symbol. Component decoders for carryng out ths task are based on the maxmum aposteror (MAP) symbol-by-symbol algorthm [4] or the soft output Vterb algorthm (SOVA). For DC-free Turbo codes as shown n Fgure 3, ts encoder and decoder structures are the same as those of the conventonal turbo codes wth slght modfcatons. The proposed scheme can be dvded nto two parts; the runnng dgtal sum (RDS) control encoder/ decoder and the Turbo encoder/ MAP decoder. The RDS control encoder generates only several code words of a wndow code for selectng a control vector [5]. Ths paper s organzed as follows. In secton II, the GPRS s revewed and detals wll be gven about ts already exstng channel codng technques. The proposed DC-free Turbo encoder/ decoder wll be presented n secton III, comparatve smulaton results are presented n secton IV, fnally the concluson n secton V. II GPRS SYSTEM AND ITS CODING SCHEMES The General Packet Rado Servce (GPRS) allows an end user to send and receve data n packet transfer mode wthn a publc land moble network (PLMN) wthout usng a permanent connecton between the moble staton (MS) and the external network durng data transfer. Ths way, GPRS optmzes the use of network and rado resources (RRs) snce, unlke crcut-swtched mode, no connecton between the MS and the external network s establshed when there s no data flow n progress. Thus, ths RR optmzaton makes t possble for the operator to offer more attractve fees.the prncples defned for the Global System for Moble Communcatons (GSM) rado nterface were kept for GPRS, snce the notons of tme slot, frame, multframe, and hyperframe have not changed for GPRS as compared wth GSM. The GPRS standard proposes multslot allocatons for data transmsson; the network may allocate up to eght tme slots per tme dvson multple access (TDMA) frame for a gven moble on uplnk and down-lnk. The GPRS standard proposes four channel codng types allowng throughput per slot rangng from 9.05 Kbps to 21.4 Kbps. Ths allows a theoretcal throughput gong up to 171.2 Kbps for data transmsson when eght tme slots are allocated to the MS. Four codng schemes (CSs), CS-1 to CS-4, have been defned for GPRS, offerng a decreasng level of protecton. The codng rate s the lowest wth CS-1 (maxmum redundancy) and s the hghest for CS-4 (no redundancy). The CS to be used s chosen by the network accordng to the rado envronment. Ths mechansm s called lnk adaptaton. The codng s based on a cyclc redundancy code (CRC), followed by a convolutonal encodng, for CS-1 to CS-3. There s only a CRC for CS-4. Puncturng s appled to adapt the convolutonal encoder output to the rado block length. Fnally, block nterleavng over the rado block makes t possble to mprove the decodng performance at the recever. The prncple for the codng of one rado block for CS-1 to CS-3 s shown n Fgure 4. The moble always transmts wth a CS ordered by the network, whereas n Rx the moble performs a blnd detecton of the used CS. Ths detecton s done by analyzng the stealng flags (8 bts per rado block, at the extremtes of the tranng sequences), one dfferent stealng flag pattern beng defned for each of the CSs. A summary of the four CS characterstcs s gven n Table 1. Ths table specfes the total codng rate for each CS and for a rado block.

III. Proposed DC-Free Error correctng Codes A. Introducton. DC-free codng s wdely employed n dgtal communcaton and storage areas. DC-free means that the coded sequence has no DC spectral component [7]. It s usually requred n dgtal transmsson and recordng systems to reduce the effect of baselne wander and match spectra of the transmtted sgnals to frequency characterstcs of the transmsson meda [8]. In ths paper, a novel DC-free turbo codng scheme s presented. Fg.3. represents the archtecture of the proposed scheme. Wth ths archtecture, t s avalable to explot soft decson decodng wth the MAP algorthm. Moreover, the dashed box part n fg.3. s exactly dentcal to a conventonal turbo codng system wth slght modfcaton. The RDS control encoder and decoder can be regarded as a front-end and a back-end of the conventonal codng system. Frstly, the user message sequence (u 0 u 1 u 2 )s encoded to ntermedate sequence (x 0 x 1 x 2 ) by a RDS control encoder. Then turbo encoder converts an ntermedate sequence to the coded sequence (y 0 y 1 y 2 ). After the bnary-bpolar converson, the coded sequence s transmtted over a nosy channel such as the addtve whte Gaussan nose (AWGN) channel. The term "RDS" means the runnng dgtal sum of a coded sequence. The DC-free property s acheved f and only f the absolute value of the RDS s bounded by a constant value for any tme nstant. B. Encodng Assume that a turbo code C together wth the parameter α, β, γ, and a decomposton matrx M are gven. The code C s called base turbo code. The message sequence (u 0 u 1 u 2 ) s dvded nto blocks of length β. The -th (=0, 1, 2 ) message block s denoted by The message sequences are encoded to the ntermedate sequences by the RDS control encoder. The ntermedate sequence (x 0 x 1 x 2 ) s dvded nto the ntermedate block of length L. The -th ntermedate block s defned by u = ( uβ, uβ + 1, uβ + 2, K, u( + 1) β 1) where O s the frst pm-tuple of x and n s the last γ+ β- tuple of x such that x = (x(γ+ β),x(γ+ β)+ 1, K x(γ+ β)+ L 1 ) x = (0 \ n ) We obtan a coded sequence (y o y 1 ) by encodng the ntermedate sequence wth the turbo encoder. The coded sequence s dvded nto the coded blocks of length r. The - th ( = 0, 1, 2,K ) coded block has the form ( y, y K y y = r+ qm r + 1 + qm, r ( + 1) 1+ qm Note that y 0 = ( y qm, K, y r 1+ qm ) The relaton between the message, ntermedate and coded sequences s shown n fgure 5 Notce that the ntermedate blocks x and x +1 are overlappng. The overlappng part corresponds to o. By applyng the addtve encoder to a wndow code, the overlappng s taken nto account. Wthn the ntermedate block x, only the vector n can be assgned freely wthout any nfluence of the prevous block. The overlappng part o s determned by the prevous ntermedate block x -1. As shown n fgure 6, the RDS control encoder adds redundancy (a control vector) to the message sequence and thus the codng rate defned between the message and ntermedate sequence becomes β / (γ + β ). The turbo encoder appends redundancy to the ntermedate sequence. Consequently, the overall rate becomes R ( pβ )/(q( γ + β)). The rate loss can be consdered as a prce for obtanng a RDS constrant. C. Decodng The decodng ssue for the proposed scheme s dscussed n ths secton. The receved sequence s frst decoded by the MAP decoder for the base turbo code C. let the set of all allowable sequences generated by the proposed scheme be C RDS. The mnmum free Hammng dstance defned on C RDS s denoted by d free proposed scheme, evdently, C. From the cascaded structure of the RDS s contaned n C and the nequalty d free dfree holds. The symbol d free denotes the mnmum free hammng dstance of C. IV. Smulaton Results In ths secton, smulaton results of BER versus E b /N o were plotted to show the nfluence of varous parameters. The channel model used s a frequency non selectve slow Raylegh fadng channel, assumed a fully estmated channel fadng values. The DC-Free Turbo decoder s used n an teratve fashon untl we acheved the 8 th teraton. The component decoder based on the MAP algorthm. The performance of convolutonal, Turbo and DC-Free turbo code for CS-1, CS-2 and CS-3 n terms of Bt Error Rate (BER) verses E b /N o are shown n fgures 6,7 and 8, as a channel nterleaver s used (an nterleaver for the dfferent consecutve symbols after codng) results n uncorrelated consecutve symbols. Also we assumed a fully estmated channel fadng values and the excepted mean value s used at the detector nstead of the true channel values. We can notce that at the same E b /N o DC-Free Turbo code gves the most effcent BER than convolutonal code and the Turbo code V. Concluson A new constructon of DC-free codes based on turbo codes whch can smultaneously meet the dc constrant and errorcorrectng requrement s proposed. The presented scheme dvded nto two parts: the RDS control encoder/decoder and the turbo encoder/decoder. The RDS control encoder generates several codewords of a wndow code for selectng a control vector. The decodng requres smpler tasks than the encoder. In ths paper, We have proposed the applcaton of a DC- Free Turbo code and a Turbo code n GPRS nstead of convolutonal code already used. We can notce that applyng DC-Free Turbo code on GPRS system gve the best performance than convoluton code and the Turbo code.

REFERENCES [1] S. Benedetto and G. Montors, 'Unvelng Turbo Codes: Some Results on Parallel Concatenated Codng Schemes," IEEE Trans. Inform. Theory, vol. IT-42, no. 2, pp. 409-428, March 1996 [2] G. Battal, "A Conceptual Framework for Understandng Turbo Codes," IEEE J. Selected Areas Commun., vol. SAC-16, no. 2,pp.245-254, Feb. 1998 [3] Yasmne Fahmy, Hala G. Abdel Kader, Magd El- Soudan, "Turbo Codng Scheme for GPRS System," Nneteenth Natonal Rado Scence Conference, Alexandra, March, 19-21, 2002 [4] L. R. Bahl, J. Cocke, F. Jelnek, and J. Ravv, "Optmal Decodng of Lnear Codes for Mnmzng Symbol Error Rate," IEEE Trans. Inform. Theory, vol. IT-20, pp. 248-287, March 1974 [5] Marwa, M. T., "Performance of DC-Free Error Correctng Codes,", M. Sc. Thess, pp 7-15& 85-90, Egypt, 2004 [6] Proaks, J. G., "Dgtal Communcatons," McGraw Hll, Inc., Forth edton, 2001 [7] M.C. Chu, DC-Free Error-Correctng Codes Based on Convolutonal Codes, IEEE Trans. Commun, vol. 49, no. 4, pp. 609-619, Aprl 2001. [8] T. Wadayama and A.J. Han Vnck, DC-Free Convolutonal Codng, IEEE Trans. Inform. Theory, vol. 48, no. 1, pp. 162-173, Jan. 2002. [9] Wcker, S. B., "Error Control Systems for Dgtal Communcaton and storage," New Jersey, Prentce- Hall, 1995 [10] Emmanuel S., Patrck S., Perre-Jean P., " EDGE for Moble Internet," Boston, London, Artech House, 2003 [11] M. W. Olphant, "The Moble Phone Meets the Internet," IEEE Spectrum, vol. 36, no. 8, pp. 20-28, Aug. 1999 [12] M. W. Olphant, "Rado Interfaces Make the Dfference n 3G Cellular Systems," IEEE Spectrum, vol. 37, no. 10, pp. 53-58, Oct. 2000 Table 1. Codng Parameters for the GPRS Codng schemes Table 2. Comparsons between usng convolutonal, Turbo and DC-Free Turbo code for the GPRS Codng Schemes at E b /N o = 4 db Fgure 6. CS-1 (E b /N o = 4) Fgure 7. CS-2 (E b /N o = 4) Fgure 8. CS-3 (E b /N o = 4) Usng Convolutonal code BER = ١ ٠٠ E-02 Usng Convolutonal code BER = ٦ ٠٠ E-02 Usng Convolutonal code BER = ٢ ٨٠ E-01 Usng Turbo code BER = ١ ٣٠ E-03 Usng Turbo code BER = ٣ ٨٠ E-02 Usng Turbo code BER = ٧ ٧٢ E-02 Usng DC-Free Turbo code BER = ٤ ٣٠ E- 05 Usng DC-Free Turbo code BER = ٩ ٤١٠E-03 Usng DC-Free Turbo code BER = ٥ ٣٢ E- 02

X X Encoder 1 Y1 Interleaver Encoder 2 Y2 Fgure 1. A typcal Turbo encoder Interleaver X Y1 Decoder 1 Interleaver Y2 Decoder 2 X Denterleaver Fgure 2. Iteratve Turbo decodng Process u o u 1 u 2.. Message Rate =β / (β+γ) RDS Control Encoder x o x 1 x 2.. Intermedate Rate = p/q Turbo Encoder Conventonal system y o y 1 y 2.. Coded Nosy Channel Estmated Message RDS Control Decoder SOVA Decoder Receved Fgure 3. Archtecture of DC-Free Turbo codes scheme

One rado block USF BCS CS-1: 224 bts CS-2: 287 bts CS-3: 331 bts Rate 1/2 Convolutonal codng USF Puncturng (CS-1: 0 bt, CS-2: 132 bts, CS-3: 220 bts) USF Block nterleavng 448 bts TS TS TS TS Stealng flag One normal burst 26 bts tranng sequence 2*1 bts stealng flags 2*57 bts data Fgure 4. Rado block encodng for CS-1 to CS-3. User message sequence β u -1 Control vector sequence (Redundancy) * RDS control ( b \ u )M Encoder Intermedate sequence x -1 x x +1 Coded sequence b * 1 u u +1 O 1 n 1 O n O +1 n +1 pm γ+β X G y -1 y y +1 γ * b Turbo encoder r Fgure 5. Relatons among message, ntermedate and coded sequences

Fgure 6. BER for CS-1 Fgure 7. BER for CS-2

Fgure 8. BER for CS-3