Energy detection in ISI channels using large-scale receiver arrays

Abstract

We investigate the performance of an energy detection (ED) system with a large number of receiver antennas in multi-path propagation channels. Contrast to existing works focusing on inter-symbol-interference (ISI) free scenarios, in multi-path environments with resolvable path components, ISI jeopardizes the performance of the ED receiver. Asymptotically, when the number of receiver antennas is sufficiently large, we show that a zero-forcing equalizer, requiring only the energy of channel taps to compute the filter coefficients, can be employed to effectively remove the ISI. This significantly reduces the burden on acquiring channel state information (CSI), since only the energy of each channel taps is demanded instead of complete CSI. As a result of asymptotic properties, to decode symbols at the output of the equalizer, decision threshold values take rather simple forms, reducing the computational complexity. Monte Carlo simulation results show that compared with using instantaneous channel energies for channel equalization, employing average channel energy leads to resembling symbol error rate, but demanding a lower filter coefficients updating frequency.