Résumé : Smart homes are one of the most promising applications of the emerging Internet of Things (IoT) technology. With the growing number of IoT related devices such as smart thermostats, smart fridges, smart speaker, smart light bulbs and smart locks, smart homes promise to make our lives easier and more comfortable. However, the increased deployment of such smart devices brings an increase in potential security risks and home privacy breaches. In order to overcome such risks, Intrusion Detection Systems are presented as pertinent tools that can provide network-level protection for smart devices deployed in home environments. These systems monitor the network activities of the smart home-connected de-vices and focus on alerting suspicious or malicious activity. They also can deal with detected abnormal activities by hindering the impostors in accessing the victim devices. However, the employment of such systems in the context of a smart home can be challenging due to the devices hardware limitations, which may restrict their ability to counter the existing and emerging attack vectors. Therefore, this paper proposes an experimental comparison between the widely used open-source NIDSs namely Snort, Suricata and Bro IDS to find the most appropriate one for smart homes in term of detection accuracy and resources consumption including CP and memory utilization. Experimental Results show that Suricata is the best performing NIDS for smart homes