A GLRT-Based Two-Stage Coexistence Detection Scheme for OFDM-Based 60-GHz WLAN/WPAN

GLRT Approach for Robust Burst Packet Acquisition in Wireless Communications

“…The former treats the denominator and numerator of the decision variable in Eq. (12) as uncorrelated random variables, while the later assumes L is extremely large so that 1 L X H k X k can be approximated as a constant [24], [35]. For the Assumption of independence approach FAP and DP are P Q ncx2 (yΓ 6 , 1, 2, Lγ)p ncx2 (y, 1/L, 2L, γ) dy, where Q ncx2 (x, σ 2 , ν, s 2 ) is the right-tail probability of the noncentral chi-square variable with freedom ν, common variance σ 2 and noncentrality parameter s 2 [37], p ncx2 y, σ 2 , ν, s 2 = 1 2σ 2 y s 2 (ν−2)/4 exp − s 2 +y…”

“…Prior publications have demonstrate the feasibility of using GLRT in wireless packet acquisition [31]- [35]. Zhi Tian et al proposed the GLRT approach for detection of burst impulse radio (IR) UWB signals in [31] and [32], where the noise is real instead of complex, and the noise power is assumed known so the GLRT detector reduces to a cross-correlation noncoherent (CCN) detector.…”

“…In fact, the perfect autocorrelation feature can hardly be satisfied and thus the derivations cannot be used in a general sequence-aided burst acquisition scenario. Another frequently used method of performance analysis for GLRT is the assumption of large amount of observations [24], which is also used in [35], where the GLRT detector boils down to a CCN detector again. In short, GLRT for general sequence-aided burst packet acquisition is still unavailable; the derivations of precise expressions of FAP and DP for GLRT are challenging and the problem is still open in the literature for the time being.…”