One of the most important challenges is security (privacy, integrity, availability and encryption). Things can be deployed at large scale in IoT environments, and existing security architecture is designed from the perspective of people communication, not being suitable to IoT system. Data integrity, privacy, data ownership, energy-efficient cryptography algorithms, legal and liability issues have to be addressed accordingly.

Technological and semantic interoperability is significantly more challenging for the IoT environments than traditional internet environment due to (i) they have different technological capabilities, and (ii) they can connect people with things and things with things. This seamless interaction between people with heterogeneous devices generates a large amount of shared heterogeneous information. Thus, improvements have to be made regarding the information model to the devices interpret the shared information correctly and act accordingly. Moreover, for full interoperability, standards for IoT must be created and broadly used.

IoT reference architectures should be defined, and then, IoT concrete architectures must be specified based on such IoT reference architectures. By default, IoT is multi-service, providing different applications or services. Thus, multiple traffic types (e.g., throughput and delay tolerant elastic traffic classes) will be transmitted within the network, and many applications/services will need quality of service (QoS) compromise. Moreover, IoT involves shared wireless media, data, and tools available on clouds, which is already an environment needs QoS requirements. Therefore, providing quality of service in IoT environments can be a hard challenge.