According to the spectrum sensing demand under complex electromagnetic environment in cognitive radio (CR) network,a truncated sequential probability ratio test (SPRT) algorithm by utilizing signal proceeding in segment is proposed based on theory and application of energy detection (ED) and SPRT in this paper.By utilizing energy calculation of received signal in segment and introduction of the central limit theorem,the log-likelihood ratio of energy detection approximates Normal distribution,which greatly simplifies the following calculative process and theoretical deduction.Simultaneously,in order to make decision and reach the best test results in a limited sensing time,a truncated SPRT algorithm is introduced and its optimal truncated threshold is derived.Performance analysis and simulation results show that under same performance conditions,the proposed detection reduce average sample number (ASN) and sensing time required around 40%~60%,comparing with the fixed-sample-size (FSS) detection.It guarantees rapid and effective perception of unknown target signal under the complex electromagnetic environment.