Energy efficient and QoS-aware routing protocol for wireless sensor network-based smart grid applications in the context of industry 4.0

Highlights

• We propose a novel BMO-based dynamic clustering algorithm to balance the data traffic and energy consumption load evenly among clusters in the smart grid.

• We propose an innovative BMO-based routing algorithm to solve energy consumption and QoS-aware reliable data transmission in the smart grid.

• The performance evaluations show that EQRP has successfully minimized the end-to-end delay and has improved the other routing QoS performance metrics, such as packet delivery ratio, efficient memory utilization, residual energy, and throughput.

Abstract

Recently, there have been great advances in internet of things (IoT) and wireless sensor networks (WSNs) leading to the fourth industrial revolution in power grid, namely, Smart Grid Industry 4.0 (SGI 4.0). In the Smart Grid Industry 4.0 framework, the WSNs have the potential to improve power grid efficiency by cable replacement, deployment flexibility, and cost reduction. However, the smart grid (SG) environment that the WSNs operate in is very challenging because of equipment noise, dust, heat, electromagnetic interference, multipath effects and fading, which make it difficult for current WSNs to provide reliable communication. For SGI 4.0 to come true, a WSN-based highly reliable communication infrastructure is essential for successful operation of the next-generation electricity power grids. To address this need, in this paper a novel dynamic clustering based energy efficient and quality-of-service (QoS)-aware routing protocol (called EQRP), which is inspired by the real behavior of the bird mating optimization (BMO), has been proposed. The proposed distributed scheme improves network reliability significantly and reduces excessive packets retransmissions for WSN-based SG applications. Performance results show that the proposed protocol has successfully reduced the end-to-end delay and has improved packet delivery ratio, memory utilization, residual energy, and throughput.

Introduction

The rise of new digital industrial technology known as Industry 4.0, has lately gained a lot of interest from researchers, manufacturers, and application developers [1], [2]. The objective of Industry 4.0 is to connect and integrate all objects of the traditional factory world for enabling faster, more flexible, and more efficient processes to produce higher-quality goods at low costs. Industry 4.0 will bring great opportunities for promoting industrial upgrades to increase manufacturing productivity, shift economics, faster industrial growth, and integration of supply and demand processes between the factories. This will modify the profile of the workforce-ultimately changing the competitiveness of companies and regions [3]. In Industry 4.0, the wireless or wired connected systems located in different remote places can interact with one another using standard internet-based protocols and analyze data to predict failure, configure themselves, and adapt to changes [4]. Currently, most devices within factories are connected based on wired infrastructure working over industrial protocols; however the wireless solutions are increasingly playing a complimentary role to wired solutions [5]. To this end, industrial wireless networks (IWNs) are the key technology enabling the deployment of Industry 4.0, since it can offer a reduction in energy consumption, increase economic benefits with least maintenance and breakdowns, and enable smart production. This does not only affect machine-to-machine communication, but will also have far-reaching consequences for the interplay of engineers and embedded system and wireless technologies [6].

In the context of Smart Grid Industry 4.0 (SGI 4.0), the cooperative communication requirement is focused on multiple factors, such as reliability, latency, scalability especially for a very large area of coverage and longevity of communicating devices [7]. In this respect, WSNs can significantly improve product quality, streamline operations, speed up production, make installation easier, increase the flexibility and reduce expenditure for the infrastructure in the SGI 4.0. The current and envisioned applications of WSNs in the SGI 4.0 span a wide range, including substation automation, overhead transmission line monitoring, energy management, advanced metering infrastructure, outage management, distribution automation, demand response and dynamic pricing, load control and energy [8]. Importantly, all these applications would lead to new products, processes and services, improving industrial efficiency and use of sustainable energy resources while providing a competitive edge for in the fourth generation global market place. At the same time, it would ensure the reliability of the electric power infrastructure, helping to improve the daily lives of ordinary citizens. However, the realization of all these currently designed and envisioned smart grid applications directly depends on reliable and efficient communication capabilities of the deployed network.

Recent field tests show that wireless channel in power grid has many unique challenges, which do not generally exist in other communication systems. These challenges, include high packet loss rates, electromagnetic interference, equipment noise, multipath effects and fading [7]. This leads to time and location dependent delay and link quality variations of wireless links in harsh nature smart grid environments. Thus, the key design issue in smart grid is to provide reliable link quality aware data delivery under adverse wireless communications conditions [9].

Prior to the SGI 4.0, many other advanced manufacturing schemes have been presented in research studies [10], [11], [12] to empower electricity industry. Some of them focus on improving reliability and end-to-end delay performance for efficient data collection in smart grid. In [13], the authors present a WSN-based reliable routing protocol (called ETL-AODV) for smart metering systems that is capable of overcoming a false indication caused by temporary loss of data, signal interference, or invalid data. In [14], the authors design a self healing routing mechanism (called HRL-AODV) that provides energy efficient and reliable two way communications of smart meters in the power grid. In [15], the authors study the issue of routing criterion and provide airtime cost metric that considers the data size and transmission rate for reliable communication in smart grid. The authors in [16] try to solve the issue of end-to-end transmission delay for providing data communication in smart grid. To guarantee the reliability of the system, an optimal sensor computational overhead mechanism for estimating the tree delay is presented, where a sensor with the highest capacity is scheduled to transmit its sensed information to the sink. Recent filed tests and measurements have shown that organizing nodes in several clusters and specifying a particular node in each cluster to perform cluster head (CH) task, not only allows aggregation of data, but also limits data transmission primarily within and among the clusters [17]. Moreover, the coordination provided by the CH allows sensor nodes to sleep for extended period and aid to save more energy in each node. Thus, clustering improves network scalability and longevity by reducing both the traffic and the contention for the channel [18]. However, highly dynamic nature of the WSN topology requires frequent re-clustering in smart grid due to the variations in the environment conditions. In recent, some other cluster based routing schemes [19], [20], [21], [22] have tried to improve the energy efficiency and packet delivery ratio for smart grid applications to empower electricity industry. In these schemes, the network is partitioned into clusters with a node elected as a cluster-head (CH) within each cluster. All the sensor nodes sense data and send it to their corresponding CH, which finally send it to the sink for further processing for smart grid applications. Although the existing studies focusing on routing issues provide valuable insights and guide design decisions for WSN-based smart grid applications, they generally ignore the impact of external interference and noise on transmission reliability during cluster formation in harsh smart grid environments. Moreover, in these schemes due to inappropriate inter-cluster and intra-cluster architecture between sensor nodes and CHs, a CH consumes available resources more rapidly than others nodes in the network. This unbalanced distribution of heavy data traffic load among the CHs consumes more energy and may lead to early death of nodes, which partitions the network and thereby degrade the overall performance of the WSNs in smart grid. In addition, the proposed schemes due to their excessive maintenance, high costs, energy wastage, and periodic service interruptions, they cannot distinguish the damage state of the network structure and the scope of clusters dynamics change fail to fulfill diverse QoS requirements for smart grid applications.

The existing studies provide valuable insights and guide design decisions for WSN-based smart grid applications. However, these traditional proactive methods are based on previously known global information, which is hard to obtain or update in the SGI 4.0. Moreover, none of the abovementioned studies consider the effective use of limited node memory and careful resource management in highly dynamic network topological smart grid environments. In addition, the existing routing solutions do not take into account the region affected by sparse or densely network deployment and are unaware of the changing data path opportunity due to lack of localization information. Therefore, these approaches are not fully able to optimize data path performance for reliable data transmission in an effective manner. Also, an effective neighbor discovery mechanism, which is important for robust routing is absent in most of these schemes. The omission leads to an excessive delay at hops in the network if a relay node dies in the routing path. Additionally, they generally ignore the impact of external interference and noise on transmission reliability in harsh SG environments. Therefore, these limitations of the previous studies motivate us to develop a new QoS-aware communication framework that improves memory utilization, throughput and provides reliable packet delivery with low network delay and network energy consumption for SGI 4.0 applications.

The contributions of this study can be summarized as follows: First, a novel BMO-based dynamic clustering algorithm is proposed to balance the data traffic and energy consumption among cluster members while constructing even size clusters in a fully distributed manner. Considering highly reliable links, the proposed algorithm performs cluster formation and cluster head selection stages concurrently which leads to low number of control packets. Second, an innovative BMO-based clustering routing algorithm is proposed to solve the imbalanced energy consumption among cluster heads caused by the non-uniform node distribution. The designed algorithm balances the energy consumption among CHs by adjusting the intra-cluster and inter-cluster energy consumption of CHs. Therefore, it can achieve the balance of energy consumption among nodes and prolong the network lifetime of WSN-based SG applications. Residual energy, proximity degree, number of hops through the path to the sink and actual distance each data traverses to reach the sink are such parameters involved in the selection of relay nodes. The designed scheme is fully capable to find robust data paths, load balancing and avoids data path loops in the network. Moreover, it significantly decreases the probability of packet loss and preserves high link quality among CHs in the smart grid in both sparse and densely deployment. Third, extensive simulations are carried out to evaluate the performance of the proposed protocol by comparing its performance with the existing approaches under realistic smart grid channel conditions. Performance results show that the proposed protocol has successfully reduced the end-to-end delay and has improved packet delivery ratio, memory utilization, residual energy, and throughput.

The organization of this paper is as follows: The following section briefly introduces the architecture of the bird mating optimization. Section 3 explains the proposed routing approach. Section 4 reports the network and path loss model, and simulation parameters. Section 5 provides experimental results where the effectiveness of the proposed routing scheme is compared to the existing routing schemes. Finally, section 6 concludes this paper by summarizing our results, significance, limitation and open issues for the potential future work.