The correct detection of the start of burst is very important in wideband networking radio operation as it directly affects the Time Division Multiple Access (TDMA) adaptive time slot algorithm. In this paper, we propose a robust Data Aided (DA) algorithm for burst detection in a hybrid CDMA/Adaptive TDMA based wideband networking waveform of a software defined radio. The proposed algorithm is based on a novel differentially modulated training sequence designed by using precoding sequence. The training sequence structure and precoding sequence are exploited in the calculation of proposed timing metric which is normalized by the signal energy. The precoding sequence is adequately designed for the timing metric to have a sharp peak. The algorithm shows excellent performance for multiuser scenario. It is shown through computer simulations that by increasing the active users from 1 to 8, the performance degradation is only about 1∼2dB. The proposed algorithm is compared with other algorithms and found to outperform them even in the presence of multipath fading effects. The proposed algorithm has been implemented on Field Programmable Gate Array (FPGA) platform for high data rate applications and it is shown that the results from hardware are identical to the simulation results.