Typically, “things” of the Internet of Things (IoT) refer to sensors, actuators, motors, and this type of small, unsophisticated gadgets. But what about connecting cameras, microphones, or other multimedia devices to the Internet in large numbers? Evidently, the quality of service requirements of such an Internet of Multimedia Things (IoMT)--for example, in terms of bandwidth, latency, jitter, error rate, processing, and memory capabilities--vastly differ from typical IoT use cases.

This paper is a first attempt at sketching the vision of IoMT, assessing the problem domain, identifying possible solutions based on existing technologies, and honestly addressing currently unresolved open issues for all. The proposed IoMT architecture, however, is too superficial (bordering on the trivial for IoT specialists, particularly when it comes to cloud-based “multimedia-aware middleware”) and too fragmented to be of (immediate) practical use.

Nevertheless, I recommend this paper to anyone working in or doing research regarding the IoT because of its wealth of technical information and its thorough and elaborate discussion of many multimedia issues in the IoT. For instance, the authors compare characteristics (for example, data rates and energy requirements) of several protocols (802.11, 6LoWPAN, ZigBee) and elaborate on traffic routing issues, device naming services, and processing requirements for video encoding (which dramatically exceed typical IoT thresholds in terms of central processing unit (CPU) capabilities and power consumption).

The easily accessible paper concludes by presenting six IoMT uses cases (for example, behavioral interpretation systems, telemedicine, and automated public security) and briefly glances at the possibilities of multiagent systems.