Zigbee Transceiver Implementation in MATLAB

This document describes a Zigbee transceiver implemented in MATLAB. It provides details on the transmitter part, with notes on how to implement the receiver as the reverse process.

Zigbee Transceiver Overview

The figure below illustrates a simple Zigbee transceiver’s internal modules. Data from the MAC layer is fed into the Zigbee PHY layer. The PHY baseband interfaces with the RF transceiver via DAC/ADC.

There are two Zigbee physical layer specifications: one for the 868/915MHz bands and another for the 2450MHz band. This example focuses on the 2450 MHz frequency band. For further information about the physical layers, refer to dedicated documentation on Zigbee Physical Layers.

As shown in the figure, the Zigbee PHY consists of the following modules:

• CRC Module: Adds redundancy to the Zigbee MAC payload.
• Bits to Symbol Conversion.
• Symbol to Chip Conversion.
• OQPSK Modulator.

Note: Simulation of half-sine wave pulse shaping is not included in this MATLAB code.

Zigbee MATLAB Code - Transmitter Part

The following MATLAB code implements the Zigbee transmitter:

% Data_gen = randint(1,10,255)
% D=dec2bin(Data_gen);
% D= reshape(D,1,80);
%% 8 bit CRC calculation
% D1=[D zeros(1,8)];
% crc_val=HCS(D);
% crc_bin=dec2bin(crc_val,8);
% D2=[D1(1:80) crc_bin];
% B1=reshape(D2,11,8);

% Binary number generator for 88 bits
bits_in=[];
for i=1:88
    num=round(rand(1));
    bits_in=[bits_in num];
end

%%Arranging 1 x 88  into 11 x 8 matrix form for bits to symbol conversion
bits_in1=[];
for i = 1:length(bits_in)/8
    b1=bits_in( (((8*i)-8)+1):(8*i)    );
    bits_in1(i,:)=b1;
end

%% bits to symbol conversion nibble wise
bits_in1=num2str(bits_in1);
bits_in1=bin2dec(bits_in1);

sym_out=[];
for i=1:11
    n1=bitand(bits_in1(i),15,'uint8');
    n2=bitand(bitshift(bits_in1(i),-4),15,'uint8');
    sym_out=[sym_out n1 n2];
end

%%Chip Table for mapping symbol(decimal from 0 to 15) to chips(32bit)
tab=['11011001110000110101001000101110'; ...
     '11101101100111000011010100100010'; ...
     '00101110110110011100001101010010'; ...
     '00100010111011011001110000110101'; ...
     '01010010001011101101100111000011'; ...
     '00110101001000101110110110011100'; ...
     '11000011010100100010111011011001'; ...
     '10011100001101010010001011101101'; ...
     '10001100100101100000011101111011'; ...
     '10111000110010010110000001110111'; ...
     '01111011100011001001011000000111'; ...
     '01110111101110001100100101100000'; ...
     '00000111011110111000110010010110'; ...
     '01100000011101111011100011001001';...
     '10010110000001110111101110001100'; ...
     '11001001011000000111011110111000'];

%Symbol to Chip conversion
chip_out=[];
for i=1:22
    sym_in=sym_out(i);
    chip_out=[chip_out tab(sym_in+1,1:32)];
end
chip_in = chip_out;
mapped_in=[str2num(chip_in(:))]';

%%Could not able to introduce half bit delay in matlab so introducing 1 bit
%%delay
% mapped_in1=mapped_in;
% for i =2:2:704
%     mapped_in1(i+2) = mapped_in(i);
% end
% mapped_in1(2)=0;
mapped_in1=mapped_in;

% %%QPSK IMPLEMENTATION
% mapper_out=mapping(mapped_in',2,0.707);
% plot(real(mapper_out),imag(mapper_out),'*r');

%%OQPSK IMPLEMENTATION
mapper_out=mapping(mapped_in1',2,0.707);
plot(real(mapper_out),imag(mapper_out),'*r');

data_td = ifft(mapper_out);  %converting frequency domain data to time domain before up conversion
plot(abs(data_td));

fc=1e3;  %1KHz RF Carrier Say
Fs=10e3; % 10 KHz sampling frequency
t=0:1/Fs:352*1e-4;   %time scale..
t=t(2:end);         % making no. of samples equal to size of data
data_td=data_td.';
data_td_uc=data_td'.*exp(j*2*pi*fc*t); % upconverting baseband to 1KHz
data_fd = fft(data_td_uc); % converting back to frequency domain for plotting purpose
%data_fd=fliplr(data_fd);
pwelch(real(data_fd),[],[],[],Fs,'twosided');

Important Consideration: The above code introduces a 1-bit delay instead of the required half-bit delay due to limitations within MATLAB. This half-bit delay is crucial for proper OQPSK modulation. For hardware implementations using FPGAs (with VHDL or Verilog), this half-bit delay can be accurately implemented.

Zigbee MATLAB Code - Receiver Part

The Zigbee receiver implementation mirrors the transmitter process in reverse. It will include the following steps:

• OQPSK Demodulation.
• Chip to Symbol Conversion.
• Symbol to Bit Conversion.