Design and implementation of enhanced IEEE 802.15.4 for supporting multimedia service in Wireless Sensor Networks

Abstract

Recently, there has been a growing demand to incorporate multimedia content delivery over the Wireless Sensor Networks (WSNs). This feature could not only enhance several existing applications in the commercial, industrial, and medical domains, but could also spur an array of new applications. However, the efficient gathering of still images, audio, and video information in WSNs imposes stringent requirements on the throughput and energy consumption. Most wireless communication standards with high or moderate data throughputs do not focus primarily on energy efficiency. The IEEE 802.15.4 WPAN standard provides a widely accepted solution for low-cost and low-power wireless communication, with a potential to cater to many types of application scenarios. IEEE 802.15.4 MAC includes features such as its dual operational modes (Non-Beacon-enabled mode/Beacon-enabled mode), which make this standard more attractive for providing multimedia services over the networked sensors. Its Beacon-enabled mode can conserve energy by using the RF sleep mechanism, but it is limited by the lower data throughput. On the other hand, the Non-Beacon-enabled mode can offer higher data throughput but at the expense of significant energy consumption, mainly due to the idle listening problem. In order to overcome these issues, we propose an enhancement to IEEE 802.15.4, named TEA-15.4, which adaptively adjusts the active period based on traffic information. To detect data traffic in the network, the proposed scheme utilizes two techniques: Arbitrary Traffic Signal (ATS) and Traffic Time-Out (TTO). By utilizing these two techniques, the proposed TEA-15.4 can not only support enough data throughput to carry out multimedia communications, but also offer lower energy consumption for the sensing device in WSNs. For performance evaluations, we implement our proposed scheme and the IEEE 802.15.4 full-standard on the TinyOS. Based on the results gathered from testbed experiments and the TOSSIM simulator, TEA-15.4 is shown to be a suitable mechanism for Wireless Multimedia Sensor Networks (WMSNs).

Introduction

Many believe that Wireless Sensor Networks (WSNs) [1] are an indispensable part of the latest wave that will revolutionize the way we do computing today. WSNs will enable our transition from the notion of “personal computing” to a technology infrastructure that allows us to integrate computing into the environment, a concept coined as “pervasive and embedded computing” [2], [3]. In this scenario, multimedia data with multiple modalities, such as still image, audio, and video, may be more influential than a large amount of conventional data observed and gathered from the physical environment. Consequently, the scope of its applicability and functionality goes beyond what traditional applications, such as habitat and healthcare monitoring, target tracking and military surveillance, and home automation and control, have to offer. With multimedia communication support, a new array of applications would be possible, where technology is envisioned to become more personal, for example the Healthcare Personal Area Network (PAN) [4]. Following the hallmarks of personal computing (i.e., inexpensive, small, and simple), a distinctive feature of such ad hoc wireless networks is the availability of limited power supply.

Applications that support multimedia communications would further exacerbate the situation, demanding lower latency and higher communication efficiency as well. Recently, there has been a growing need for research efforts to provide multimedia services over WSNs [4], [5], [6]. These survey papers have categorically argued for new techniques at each layer of the network protocol stack to cope with the foreseen challenges associated with the Wireless Multimedia Sensor Network (WMSN). In the context of contention-based channel access schemes, more emphasis was put on the needs for energy-aware and adaptive duty cycle calculations that could meet the foremost concerns of energy conservation without sacrificing data throughput and delay.

Most wireless communication standards with high or moderate data throughputs do not focus primarily on energy efficiency. However, the IEEE 802.15.4 Wireless Personal Area Network (WPAN) standard [7], [8] provides support for low-cost and low-power wireless connectivity among resource limited devices. The IEEE 802.15.4 Physical (PHY) and Medium Access Control (MAC) specifications promise to cater to diverse performance requirements from a wide spectrum of applications [9], [10], [18], [19]. The IEEE 802.15.4 MAC includes the features like low-duty cycle operation and self-organization for WPANs. These features make this standard more attractive for providing multimedia services over the networked sensors.

Akyildiz et al. [5] advocated for the suitability of a multi-tier, scalable network architecture consisting of heterogeneous elements for WMSNs. These heterogeneous elements provide an efficient cost-performance mix by employing less expensive and resource-constrained scalar sensors and high power superior elements such as image sensors. These elements send multimodal sensing data to a cluster head designated for performing resource intensive processing. This architecture is illustrated in Fig. 1. For scalar sensors, conserving energy is critical because they generate data traffic at lower rates. The IEEE 802.15.4 MAC achieves low-duty cycle operation, termed the “Beacon-enabled mode,” based on the RF sleep mechanism. On the other hand, image sensors demand higher data throughput in order to transmit the bulk of multimedia traffic. In this case, the Non-Beacon-enabled mode would be the preferred choice, since it operates without any sleep mechanism. Unfortunately, the operational mode is decided by the coordinator during the network initialization, forcing the IEEE 802.15.4 to support only one mode at a time. Consequently, this makes it difficult to meet both goals, i.e., the higher energy efficiency and data throughput in the original IEEE 802.15.4.

In this paper, we propose a novel scheme, named “Traffic and Energy Aware IEEE 802.15.4 (TEA-15.4)” to reduce energy consumption and improve data throughputs in the current IEEE 802.15.4 standard by utilizing data traffic information. In the proposed TEA-15.4, a coordinator can adaptively adjust the active period according to the data traffic information of the associated devices in the Beacon-enabled mode. When the data traffic load is low, the active period is decreased to reduce energy consumption caused by the idle listening [11], [12]. On the other hand, for higher data traffic, the active period is extended close to that of the Non-Beacon-enabled mode to improve the data throughput. In order to detect the traffic information, TEA-15.4 employs two techniques. The first mechanism is based on Arbitrary Traffic Signal (ATS) and the second utilizes Traffic Time-Out (TTO). The ATS scheme is designed to detect an arbitrary traffic frame or its collision signal that indicates the existence of data traffic, whereas the TTO scheme utilizes a time-out mechanism to detect data traffic information of the associated devices. These two mechanisms are periodically performed during the sentinel duration, i.e., a special epoch for detecting the traffic information as decided by the coordinator.

The main contributions of our work are as follows: first, we have proposed a novel solution for improving IEEE 802.15.4 performance with the adaptive active duration via two data traffic indication schemes (that is, ATS and TTO). Second, we have designed and implemented a real sensor platform and its camera module for testbed experiments of the Wireless Multimedia Sensor Networks. Finally, for performance evaluations, we have implemented our proposed scheme, as well as the original IEEE 802.15.4 full-standard, into the TinyOS [28]. Moreover, we have designed and implemented a new wireless propagation model and RF physical stack for simulations in the TOSSIM [29].

The rest of the paper is organized as follows: in Section 2, we give a brief overview of the IEEE 802.15.4 standard and a summary of related works. In Section 3, our motivation is described with some experimental results on the IEEE 802.15.4 protocol stack, and our WMSN test-bed implementation details are introduced. Section 4 explains our enhanced TEA-15.4 protocol and its sub-schemes, which include the proposed traffic indication mechanisms in detail. Section 5 presents extensive performance evaluations of the proposed scheme and its comparison with the original IEEE 802.15.4 standard. Finally, we conclude the paper in Section 6.