Quantifying Tradeoffs in the IEEE 802.15.4 protocol through simulation Submitted To Pierre Collinet Dr. Brian L. Evans Prepared By Chinmoy Gavini EE464 Senior Design Project Electrical and Computer Engineering Department University of Texas at Austin Summer 2007ii CONTENTS TABLES ............................................................................................................................... iv FIGURES ............................................................................................................................. vi EXECUTIVE SUMMARY.................................................................................................. v 1.0 INTRODUCTION....................................................................................................... 1 2.0 DESIGN PROBLEM STATEMENT ........................................................................ 2 3.0 DESIGN PROBLEM SOLUTION ............................................................................ 2 3.1 BACKGROUND INFORMATION.................................................................. 3 3.1.1 DBPSK Modulation………………………………………………………3 3.1.2 OQPSK Modulation……………………………………………………...4 3.1.3 Direct Sequence Spread Spectrum (DSSS) ............................................. 5 3.1.4 Channel Models…………………………………………………………..5 3.2 DESIGN DECISIONS ....................................................................................... 6 3.2.1 Rayleigh and Rician Models..................................................................... 6 3.2.2 IEEE 802.15.4a Channel Model ............................................................... 7 3.2.3 Timing recovery design decisions ............................................................ 8 4.0 DESIGN IMPLEMENTATION ................................................................................ 8 5.0 TEST AND EVALUATION....................................................................................... 9 5.1 Bit Error Rate Analysis ...................................................................................... 10 5.2 Data Rate and Implementation Complexity ..................................................... 14 6.0 TIME AND COST CONSIDERATIONS ............................................................... 15 7.0 SAFETY AND ETHICAL ASPECTS OF DESIGN .............................................. 15 8.0 CONCLUSIONS AND RECOMMENDATIONS .................................................. 16 REFERENCES ................................................................................................................... 17 APPENDIX A – GLOSSARY ......................................................................................... A-1 APPENDIX B – BLOCK DIAGRAM ........................................................................... B-1 APPENDIX C – SIMULATION SOURCE CODE....................................................... C-1iiiiv TABLES 1 Expected Characteristics of the IEEE 802.15.4 protocol ................................................. 2 2 BER Comparison of the OQPSK and BPSK system ..................................................... 14v FIGURES 1 Schematic representation of DBPSK modulation ............................................................ 6 2 BER plot of BPSK system with AWGN........................................................................ 10 3 BER plot of BPSK system with AWGN and channel model......................................... 11 4 BER plot of OQPSK system with AWGN..................................................................... 12 5 BER plot of OQPSK system with AWGN and channel model...................................... 13 6 IEEE 802.15.4 simulation block diagram .................................................................... B-2vi EXECUTIVE SUMMARY The goal of this project is to quantify the tradeoffs in the modulation methods of the IEEE 802.15.4 wireless sensor network protocol through baseband simulation in Matlab. The protocol specifies the Differential Binary Phase Shift Keying (DBPSK) modulation for the 868 MHz and the 915 MHz carrier frequencies and the Offset Quadrature Phase Shift Keying (OQPSK) modulation for the 2.4 GHz frequency. The three Physical Layers use Direct Sequence Spread Spectrum (DSSS). The simulation consists of the transmitter, receiver, channel, and Bit Error Rate (BER) analysis modules. Symbol generation, DSSS spreading, modulation, pulse shaping, and transmission occur in the transmitter, while the receiver performs matched filtering, demodulation, and DSSS despreading. The channel module corrupts the signal with Additive White Gaussian Noise (AWGN) and attenuates the signal with the channel impulse response. Binary Phase Shift Keying (BPSK) is a modulation technique in which the phase of the carrier wave assumes one of two values depending on the bit to be sent. In DBPSK, the data stream is encoded into an intermediate bit stream based on the XOR result of the current unencoded bit and the previous encoded bit. The phase of the carrier wave changes if the thn encoded bit is different from the previously encoded bit. In Quadrature Phase Shift Keying (QPSK), the phase of the carrier wave assumes one of four values. In OQPSK, the in-phase and quadrature-phase channels are offset by ½ of a symbol period. DSSS is an encoding technique for counteracting noise and widening the frequency band of the transmitted signal. The transmitter maps or “spreads” the symbols into predefined PseudoNoise (PN) sequences before modulation. The receiver recovers the symbols after demodulation. An appropriate channel model must account for multipath fading and noise because multipath propagation and noise are present in any wireless system. The three channel models I considered were the Rayleigh model, the Rician model, and the IEEE 802.15.4a channel modeling subgroup’s model. An accurate channel model must include a channel response in terms of a “time-varying Finite Impulse Response (FIR) filter, a time-varying gain for fading, and AWGN” [2]. The IEEE 802.15.4a subgroup’s model was optimal because it met all three constraints and generalized the Rician and Rayleigh models. I chose header correlation and filter delay compensation for timing recovery in the BPSK system and the