Characterization and Modeling of M2M Video Surveillance Traffic
Ivo Petiz, Paulo Salvador, António Nogueira · 2012
Abstract—The relevance of machine-to-machine (M2M) communications is growing significantly, largely driven by wireless networks. Internet of Things (IoT) applications will create new demands and challenges that will require high bandwidth, real-time communications or reliability in remote locations. Since M2M applications will generate traffic and transactions that compete for bandwidth and priority, network operators must be concerned of how to handle the enormous increases in their signaling traffic and, in particular, in finding new ways to meet each M2M application’s requirements and service level agreements, while protecting the network and making a more efficient use of the available resources. An efficient design and control of the future Internet needs to take into account the main characteristics of the supported traffic and, therefore, accurate and detailed measurements of M2M traffic have to be carried out in order to perform a complete characterization and develop stochastic models that are able to capture the most important statistical properties of the network traffic. This paper considers a particular case of a M2M service, a video surveillance application, and proposes a general methodology that performs a detailed statistical characterization of its traffic, while addressing the main challenges that are involved in the development of an appropriate stochastic modeling approach. The M2M traffic analysis relies on the use of wavelet scalograms, which describe the signal energy on a timescale/time domain and are constructed based on the wavelet coefficients obtained from the multi-scale decomposition of the traffic process, to identify all the different traffic features. An innovative M2M modeling framework based on a multiple state machine is also proposed: its parameters should be inferred from the salient features of the traffic and it should be able to characterize both the download and upload traffic and quantify the overhead that is necessary to guarantee a secure communication.