A Cross-Layer QoE-Based Approach for Event-Based Multi-Tier Wireless Multimedia Sensor Networks

Introduction

The proliferation of multimedia content and the demand for new audio/video services in the ¨Internet of Things¨ (IoT) applications (Atzori et al., 2010) have ushered in a new era based on multimedia information. These applications have led to the evolution of Wireless Multimedia Sensor Networks (WMSNs) (Tavli et al., 2012), which have attracted interest of researchers because of the availability of low-cost and mature technologies in camera and scalar sensors. WMSNs are composed of wirelessly interconnected nodes equipped with multimedia sensor devices, such as cameras and microphones, to enable the nodes to retrieve video and audio streams, and still images, as well as scalar sensor data.

WMSNs enable a large class of applications ranging across diverse areas in both civilian and military areas, such as multimedia surveillance, environmental monitoring, traffic control, smart cities and other IoT applications. The multimedia content in these applications gives support to end-users (or systems) to visually determine the real impact of an event, be aware of what is happening in the environment, enhance the level of information collected, and provide multi-view support (Alaei & Barcelo-Ordinas, 2012). Thus, the users can make use of the visual information to take action for object/intruder detection. In this study, we focus on intrusion detection scenario with real-time video support.

Different network architectures have been proposed to provide different application requirements. However, the multi-tier network architecture with heterogeneous nodes is the most suitable solution for routing protocols in WMSNs to improve scalability, resilience and energy-efficiency, as shown by (Almalkawi et al., 2010). Moreover, this must be carried out through efficient routing decision to improve the video quality, energy-efficiency, and scalability in event-based WMSNs (Ehsan & Hamdaoui, 2012). Multimedia dissemination should support Quality of Service (QoS) and Quality of Experience (QoE) requirements so that video content can be delivered with, at least, a minimum video quality level from the user’s perspective, together with energy-efficiency and scalability. These issues impose more constraints on the designing of a reliable multi-tier routing protocol for WMSNs (Wijnants et al., 2010).

Solutions involving multimedia transmission must take into account the video characteristics from the user’s perspective to increase the user’s satisfaction, while keeping the content with a good quality level. In this context, frames with different priorities compose a compressed video, and according to the user’s perspective, the loss of high priority frames causes severe video distortion. Thus, a key principle for a QoE-aware multimedia dissemination in multi-tier WMSN applications is the protection of high priority frames in congestion/link error periods (Serral-Gracià et al., 2010). Furthermore, the constraints of sensor nodes increase the effects of wireless channel errors, and the application-level Forward Error Correction (FEC) can be employed as an error control scheme for handling losses in real-time WMSN communications (Yousof Naderi et al., 2012). FEC schemes achieve robust video transmission by sending redundant data, and when the original information is lost, the data can be recovered from the redundant information. Application-level FEC adds h redundant packets to a set of n original source packets. The redundant packets can be used to reconstruct a lost frame, and thus, enhance the video quality level to meet the requirements of human experience.

Multi-tier event detection applications with multimedia support require a cross-layer approach to provide video dissemination with a high quality level from the user’s perspective, together with scalability, energy-efficiency and low network overhead. In this context, the multi-tier routing protocol must provide a route selection scheme that employs cross-layer information to find reliable routes. Moreover, priority frames must be protected in congestion/link error periods by taking into account an application-level FEC mechanism, which considers the importance of the frame for creating the redundant packets. However, the existing cross-layer approaches do not include all of these key characteristics to support QoE-aware multimedia dissemination.