Simple Peak Interference Cancellation (SPIC): Interference Cancellation Prior to Packet Decoding in LoRa Networks

LoRa has been proposed for providing long range connection capabilities in Internet of Things (IoT) networks. LoR
$a$
is capable of supporting simultaneous transmissions of nodes when nodes use different Spreading Factors (SF) because signals modulated with different SFs are quasi orthogonal. In this paper, we propose a new interference cancellation technique, namely, Simple Peak Interference Cancellation (SPIC), which enhances the multiple access capability of LoR
$a$
networks by removing the interference of packets with other SFs when decoding packets with a particular SF. The unique feature of SPIC is to reconstruct the time-domain signal of a packet without decoding the packet, which is possible because the transmitted signal in LoR
$a$
is converted to a peak that can easily be identified and extracted. We demonstrate SPIC with traces collected in real-world experiments. Our results show that SPIC significantly outperforms other interference cancellation techniques despite its simplicity.