“…(4) P e0 = l--erfcl where F, = A,(T k ) 2 (i=l, 2), T k is the sampling time for the k-th bit, d the decision threshold, whose optimum value is (5) The error probability is then given by P e = ni (Pr{Q k <0/m,b k =l}) m + 7i 0 (Pr{Q k >0/m,b k =0}} m (6) where Q is the decision variable ( Fig. la) and Pr{B/F} is the probability of the event "B" conditioned by the event "F", b k indicates the k-th bit, m which type of message has been transmitted and π!…”