A geographic routing approach for IPv6 in large-scale low-power and lossy networks

Highlights

• We combine RPL and GOAFR protocols for IPv6-enabled large-scale wireless networks.

• Main drawbacks of the RPL protocol for P2P communication are discussed.

• A routing protocol (GeoRank) is proposed and its scalability is compared to RPL.

• Simulations were performed on networks extracted from real street maps.

• GeoRank is adaptive to variable link densities found in large-scale networks.

Abstract

In this paper, we propose GeoRank, a geographic routing approach for the IPv6-enabled large-scale low-power and lossy networks. We discuss the main drawbacks of the RPL (IPv6 Routing Protocol for Low-Power and Lossy Networks) for P2P (point-to-point) communication in large-scale 6LoWPAN networks. Then, we address such drawbacks by proposing a routing protocol, named GeoRank, which integrates RPL protocol with the position-based routing protocol GOAFR (Greedy Other Adaptive Face Routing). The results obtained with simulations show that GeoRank finds shorter routes than RPL in high link density conditions and than GOAFR in low link density conditions. Thus, GeoRank shows to be adaptive to variable link densities found in large-scale networks. Further, GeoRank avoids the use of DAO (Destination Advertisement Object) control messages required in RPL, while being more scalable in terms of memory usage than storing-mode RPL.

Introduction

With the increase in world population and the growing urbanization worldwide, cities will need to evolve towards an intelligent infrastructure in order to provide to its citizens a better living environment. This will be achieved through the use of information and communication technology (ICT), with the goal to provide public services that are more accessible and more efficient. This vision is embodied in the “smart city” concept [1], [2].

According to several researchers, to enable the smart city it will be required an urban “Internet of Things (IoT)” infrastructure, where sensors and actuators distributed across the city will be interconnected by a pervasive network [3], [4]. This urban “IoT” will enable a myriad of services and applications for the smart city, such as structural health monitoring, waste management, air quality monitoring, noise monitoring, traffic congestion avoidance, city energy consumption tracking, smart parking, smart lighting, and charging infrastructure for electrical vehicles [1], [2].

One of such applications is smart street lighting. The street lighting system (SLS) is one of the most important assets for a city, due to its role as a public service for the citizens and also due its pervasiveness. However, SLS is very costly to operate, with a share of about 40% of the total amount of electricity spent in a city [5]. Therefore, making it more energy efficient and less costly for maintaining is very important for the cities.

The use of LED (light emitting diode) lamps in street lighting started to be reported in the last few years, and it is being considered a very good solution to improve the energy efficiency of the system and also to reduce the maintenance cost, because of their long lifetime, high luminous efficiency and high color rendering index (CRI) [6]. Also, the addition of communication capability to the SLS decreases the energy consumption and the maintenance cost, because it allows for monitoring the status information of each lamp (state, current, voltage and power measurements, etc.) and to have better control of the system (e.g.: lamp dimming). Then, a smart SLS, based on LED lamps and a communication network, is an interesting application in a smart city context, not only by its clearly advantages but also because it can be one of the foundations for a urban “IoT” due to its pervasiveness across the city [5]. Therefore, it can also be an enabler for the aforementioned applications in the smart city.

However, a street lighting system features a large number of independent devices, with geographic distribution depending on the city streets. Then, adding communication capabilities to these devices requires a complex network topology, as well as interoperability, scalability, security, robustness, low cost, ease to use and maintenance. A few contributions in the literature suggest some technologies that can be used to control a street lighting system. Some use the power lines for data transmission (PLC) [7], while others use wireless communication, such as cellular networks [8], wireless sensor networks (WSN) [9] and both combined [10].

The main drawbacks of the PLC technology are noisy medium, high signal attenuation, susceptibility to interference from nearby devices, time-varying network topology, high complexity and poor scalability, despite its existing infrastructure and extensive coverage. The most suitable area for PLC seems to be the medium voltage distribution network, usually combined with wireless technologies in the low voltage distribution network through a hybrid architecture [11]. Similarly, the scalability and reliability of cellular networks are not questionable, especially under low load traffic. However, the literature has already demonstrated that the limitations of the cellular networks for energy-constrained and/or processing-constrained devices are the tight synchronization requirement and high signalling overhead [12]. Furthermore, the dependency on the cellular carriers are usually not desirable in the smart cities/internet of things context. On the other hand, wireless sensor networks scalability is highly dependent on the routing algorithm performance. Therefore, designing a routing algorithm with high scalability and low overhead is a challenging requirement. It is important to note that we do not completely discard the PLC and cellular technologies for smart cities, but just reinforced that such technologies may be implemented as hybrid solutions, in order to develop redundant or complementary systems.

Recently, the 6LoWPAN protocol has been standardized by the IETF (Internet Engineering Task Force) as an adaptation layer to enable IEEE 802.15.4-based low power and lossy networks (LLNs), such as wireless sensor networks, to reuse the existing IPv6 protocol for the networking layer [13]. Also, the IETF has standardized RPL as the routing protocol for IP-based LLNs [14]. According to the literature, the main reasons to use IPv6 protocol for the LLNs include the connectivity through the internet to enable the IoT, the 128-bit addressing space that enables uniquely addressable devices (“things”), the network layered architecture and the interoperability [15], [4].

However, RPL has some drawbacks, as higher amount of control messages than data messages, mainly for constructing routes from the sink to the nodes (downward routes in RPL vocabulary) [16], [17], [18]. In a network where nodes are static, a routing protocol could benefit of this condition to find routes with less control overhead than RPL. The routing algorithms that use position to find routes in a communication network are known as geographic routing algorithms and a large body of literature has been devoted to the research on those algorithms in last few decades [19], [20], [21].

Our main contribution in this paper is to propose a modification into RPL protocol, in order to better support P2P communication. The proposed technique intend to integrate a geographic routing algorithm in the RPL protocol to reduce the amount of control message required for P2P communication. Such a routing protocol can be used, for example, in a smart street lighting system wireless network for supporting diverse applications in an urban IoT.¹ This paper is organized as follows: Section 2 addresses the requirements of a wireless sensor network to implement an smart street lighting network. Also, we argue about the benefits of creating a network with such coverage for other public services. Section 3 presents an overview of the 6LoWPAN protocol, describing succinctly its operating modes, routing strategies and main drawbacks. In Section 4 we address the geographic routing technique, highlighting its pros and cons. In Section 5, the GeoRank algorithm is presented, depicting its working principle. We evaluate, by simulation, the performance of the proposed algorithm in Section 6. Finally, we conclude this paper in Section 7.