Optimal MAP Detection in Presence of Burst Interference for Low Power Wide Area Networks.

Low Power Wide Area (LPWA) networks are known to be very vulnerable to external in-band interference in terms of packet collisions, which may substantially degrade the system performance. In order to improve the performance under collisions, the so-called telegram splitting (TS) method has been proposed recently. This approach is based on the assumption of a bursty behavior of the interference, which can be efficiently accounted for by forward error correction. In this paper, we assume that all interferers have the same structure in terms of packet duration and signal variance, which are assumed to be known to the receiver. However, the number of simultaneously active interfering transmissions and their scheduling are unknown. Hence, the amount of interference observed in each symbol interval can vary which needs to be accounted for by the receiver in order to increase the reliability of transmission. For this scenario, we develop the optimal signal detection strategy. We also show that the proposed method substantially outperforms a baseline scheme in terms of packet error rate.