function [ tx_burst , I , Q ] = gsm_mod(Tb,OSR,BT,tx_data,TRAINING)% GSM_MOD:  This MatLab code generates a GSM normal burst by%           combining tail, ctrl, and training sequence bits with%           two bloks of random data bits.%           The data bits are convolutional encoded according%           to the GSM recommendations%           The burst sequence is differential encoded and then%           subsequently GMSK modulated to provide oversampled%           I and Q baseband representations.%% SYNTAX:   [ tx_burst , I , Q ] = gsm_mod(Tb,OSR,BT,tx_data,TRAINING)%% INPUT:    Tb: Bit time, set by gsm_set.m%          OSR: Oversampling ratio (fs/rb), set by gsm_set.m%           BT: Bandwidth Bittime product, set by gsm_set.m%      tx_data: The contents of the datafields in the burst to be%               transmitted. Datafield one comes first.%      TRAINING: The Training sequence which is to be inserted in the%               burst.%           % OUTPUT:%     tx_burst: The entire transmitted burst before differential%               precoding.% 	     I: Inphase part of modulated burst.%            Q: Quadrature part of modulated burst.%% WARNINGS: No interleaving or channel coding is done, and thus the %           GSM recommadations are violated. Data simulations done using%           this format can only be used for predicting Class II performance.%% AUTHOR:   Jan H. Mikkelsen / Arne Norre Ekstr鴐% EMAIL:    hmi@kom.auc.dk / aneks@kom.auc.dk%% $Id: gsm_mod.m,v 1.10 1997/12/17 15:29:27 aneks Exp $% GENERATE BURST SEQUENCE (THIS IS ACTUALLY THE MUX).%tx_burst = burst_g(tx_data,TRAINING);% DIFFERENTIAL ENCODING.%burst = diff_enc(tx_burst);% GMSK MODULATION OF BURST SEQUENCE%[I,Q] = gmsk_mod(burst,Tb,OSR,BT);