Data aided algorithm for burst detection in wideband networking waveform with FPGA implementation on SDR platform

Abstract

This paper presents a new technique for detection of bursts in a wideband networking waveform of a software defined radio. For burst mode of transmission, the timing synchronization includes the recovery of time varying sampling clock offset and the detection of start of burst. Both these operations are very important in wideband networking radio operation as they directly affect the Medium Access adaptive time slot algorithm. The proposed burst detection scheme computes the decision metric on the basis of a specially designed training sequence. The threshold value for the proposed burst detection metric is very less dependent on channel conditions. It is shown that both the proposed algorithms achieve better performance as compared to existing algorithms. Computationally efficient prototype hardware implementation on Field Programmable Gate Array (FPGA) is also presented. Real-time hardware results have been presented to demonstrate the effectiveness of the proposed algorithms and architecture for systems requiring high data throughput.