**2. Background and Related Work**

This section introduces how ENF can be generated from multimedia streams and how ENF can be used for the environmental fingerprint. Following that, we describe typical consensus protocols in blockchain and provide related work on IoT-blockchain integration.

## *2.1. ENF as a Region-of-Recording Fingerprint*

ENF is the supply frequency in power distribution grids, which has a nominal frequency of 50 Hz or 60 Hz depending on the location of the power grids. Due to environmental effects in the grid such as load variations and control mechanisms, the instantaneous ENF usually fluctuates around its nominal value. At a given time, variation trends of ENF fluctuations from all locations of the same grid are almost identical due to the interconnected nature of the grid [13]. ENF fluctuations are embedded in audio/video recordings either due to electromagnetic induction or background hum from devices connected to the power grid [14]. Thanks to the consistency and reliability of ENF at a time instant, ENF has been adopted as a forensic tool for identifying forgeries in multimedia recordings. All ENF signals estimated from simultaneous multimedia recordings at different locations have similar fluctuations throughout the power grid. Thus, there are multiple forensic applications based on ENF, such as validating the time-of-recording of an ENF-containing multimedia signal [14] and estimating its location-of-recording [15].

In IoVT systems, ENF signals extracted from video recordings are in the form of illumination frequency (120 Hz). The video recordings made under indoor artificial light include ENF fluctuations. The estimation of ENF signals depends on the type of imaging sensor used in a camera. The most commonly used imaging sensors are complementary metal oxide semiconductors (CMOSs) and charge-coupled device (CCD) sensors, which have different shutter mechanisms. In this work, we assume that ENF signals are extracted from video recordings generated by cameras with CMOS imaging sensors in an indoor setting with artificial light [11,16]. The estimation of ENF involves various signal processing techniques such as power spectral analysis and spectrogram-based techniques, which are beyond the scope of this paper.
