New lower bounds on the error probability for signals over an AWGN channel

New lower bounds on the error probability when communicating one of M equally likely signals over an additive white Gaussian noise (AWGN) channel are proposed. The bounds are derived by improving on a recent lower bound on the probability of a union of events by de Caen, and applying it to this problem. The resulting bounds, specialized for linear codes, improve on the latest lower bound appearing in the current literature by Seguin.