Edge-distributed stream processing for video analytics in smart city applications

Abstract

Emerging Internet of Things (IoT) applications based on distributed sensors and machine intelligence, especially in smart cities, present many challenges for network and processing infrastructure. For example, a single system with a few dozen monitoring cameras is sufficient to saturate the city’s backbone. Such a system generates massive data streams for event-based applications that require rapid processing for immediate actions. Finding a missing person using facial recognition technology is one of those applications that require immediate action at the location where that person is since this location is perishable information. An encouraging plan to support the computational demand for widely geographically distributed systems is to integrate edge computing with machine intelligence to interpret massive data near the sensor and reduce end-to-end latency in event processing. However, due to the limited capacity and heterogeneity of the edge devices, distributed processing is not trivial, especially when applications have different Quality of Service (QoS) requirements. This work presents an edge-distributed system framework that supports stream processing for video analytics. We investigate recent researches regarding massive IoT data stream processing, primarily focusing on the division of this processing in multiple types of tasks. Then, we propose an architecture to organize edge and cloud nodes for running various functions in a collaboration schema to process multimedia data streams. A method of distributing workload on edge nodes for event-based processing is also proposed, along with a scheme for reusing nodes that perform tasks of interest to various applications, such as a facial recognition task, for example. We have also developed an algorithm to allocate nodes with sufficient processing capacity to process the flows demand while meeting the applications’ QoS requirements. Finally, the simulations showed that the distribution of processing across multiple edge nodes reduces latency and energy consumption and further improves availability compared to centralized processing in the cloud.