Modified chi-square-AD joint detection on fast fading channels

Based on a transformation of goodness-of-fit test,this paper proposes methods of spectrum sensing.Specifically,Anderson-Darling(AD) test,a kind of goodness-of-fit test,is applied into a new sequence which is a transformation of received samples,conditioning on known or unknown noise variance.To enhance performance of sensing,a modified chi-square-AD joint test is also proposed under a condition of unknown noise variance.Simulations show that,for a fast fading channel,the modified chi-square-AD joint test greatly improves the performance of spectrum detection,and makes a breakthrough when the noise variance is unknown.