Randomized on-off signaling for asynchronous interference channels

This paper addresses a Gaussian interference channel consisting of two active users. The channel from each transmitter to each receiver is modeled with quasi-static and non-frequency selective Rayleigh fading. Users are asynchronous, meaning there exists a mutual delay between their transmitted codes. Due to the randomness of delay, no user is aware of the location of the interference burst along its code. As such, no interference cancellation is performed, i.e., users treat each other as noise. By the same token, interference has a mixed Probability Density Function (PDF) as a result of ambiguity on the start of the interference bursts. A stationary model for interference is considered by assuming the starting point of an interferer´s codeword is uniformly distributed along the codeword of any user. Due to non-ergodicity of the model, outage analysis is an appropriate tool to study the network. All users follow a locally Randomized Masking (RM) signaling scheme where each transmitter quits transmitting its Gaussian signals independently from transmission to transmission with a certain probability. An upper bound on the probability of outage per user is developed using entropy power inequality and a key upper bound on the differential entropy of a mixed Gaussian random variable. It is shown that by adopting the RM scheme, the probability of outage is strictly lower than that of a scenario where both users keep transmitting their Gaussian signals known as continuous transmission.