DS-SS quasi-optimal detection under interference generated by computer platforms

Recently, it has been shown that the electromagnetic interference generated by computer platforms like laptops, smart phones, tablets, or PDAs fits a double-sided K-distribution statistic. This kind of impulsive noise can severely degrade the performance of linear receivers intended for operation under additive gaussian noise. Therefore, we propose the use of a quasi-optimal detection scheme with almost the same implementation complexity than the traditional linear receiver. Indeed, it only requires an additional memoryless saturating nonlinearity. This receiver is known to be optimal under a Laplacian noise distribution. We show that it is nearly optimal for the rest of the K-distribution family. Simulations results for the case of Wi-Fi 802.11b are presented quantifying the performance of this receiver compared with other interesting detection schemes. The results are very promising since improvements of 1 to 2 dB are shown to be achievable with small modifications in the linear receiver structure.