Performance Evaluation of Cluster-Based Routing Protocols for Wireless Sensor Networks

 By

Adebanjo Adekiigbe¹, Banji Moruf Fadipe², Oniyide Alabi Bello³, Ramoni Tirimisiyu Amosa¹, Fabiyi Aderanti Alifat¹, Olorunlomerue Adam Biodun¹, Olatunji Abiodun Funsho ,  Joseph Babatunde Isaac¹, ⁴Olanrewaju Kabeerat Adeola

 

¹Department of Computer Science, Federal Polytechnic, Ede, Osun State, Nigeria

²Research Grants Assistant, Department of Research and Developments, University of      

  Limpopo, Republic of South Africa

³Department of Mathematical Sciences, Afe Babalola University, Ado-Ekiti, Nigeria

⁴Department of Computer Science, Federal Polytechnic,  Ayede, Oyo State, Nigeria

 

 

 

 

Abstract

Wireless Sensor Networks (WSNs) perform a vital starring role in numerous applications, ranging from environmental monitoring to military surveillance. Efficient data communication and management are essential for the successful operation of WSNs. Clustering is a popular method used to enhance performance of network and prolong the lifetime of a network. This paper present a comparative analysis of clustered routing protocols for WSNs for ad-hoc networks. We evaluate performances of three widely used protocols: Low-Energy Adaptive Clustering Hierarchy (LEACH), LEACH-Centralized (LEACH-C), and Threshold Sensitive Energy Efficient Protocol (TEEN), we used some key performance metrics such as lifetime of a network, rate of consumption of energy, and data delivery rate. Our findings provide insights into the strengths and weaknesses of these protocols and aid researchers and network designers in selecting the most suitable cluster-based routing protocol for their specific applications.

Keywords: Wireless Sensor Networks, cluster-based routing protocols, network lifetime, energy consumption, data delivery rate.

Introduction

Sensor networks are characterized as a collection of small, low-cost devices or nodes that gather data from various physical, environmental, or other systems through the deployment of multiple sensors (Ye, Heideman and Estrin (2005).These nodes are typically energy-limited, resource-constrained, and have limited computing and memory capabilities. Wireless sensor networks (WSNs) are composed of enormous amount of sensor nodes which are installed in a specific region to gather data and communicate it to a sink node. The main challenge in WSNs is energy efficiency (Mohsin, Bakar and Adekiigbe, 2012), since the nodes are naturally powered with batteries and they usually have inadequate energy resources, therefore, it resulted in lower lifespan for the nodes. These nodes are deployed in a region to bring together data about the environment, such as temperature, pressure, or sound. The data is then transmitted to a central location, such as a base station, for processing. One of the major challenges in WSNs is energy efficiency. Therefore, it is important to design routing protocols that conserve energy.

Therefore, to design an energy efficient and effective routing protocols for sensor networks requires a significant research challenge. Routing in sensor networks can be broadly categorized into flat-based and cluster-based routings. Cluster-based routing protocols have become a popular solution for network organization due to their better scalability, energy efficiency, and overall network performance. 

Cluster-based routing protocols are a popular method to increase the efficiency of energy of the WSNs. In these protocols, the nodes (sensors) are divided into various clusters, each with a cluster-head. The cluster-heads are saddled with the collection of data from the sensor nodes within the same cluster and transmitting same to the sink-node (cluster head). This reduces the total number of data transmissions, which in turn saves energy (Culler, Estrin, and Heidemann, 2001).

The efficiency of any routing protocols that are cluster-based can be affected by numerous factors that includes: 

1. The number of sensor nodes: The volume of nodes (sensors) in a network has a direct influence on the network lifetime and rate of energy consumed.
2. The distribution of the sensor nodes: Distribution of sensor nodes in the network can equally affect the rate of data delivery and packet transmission delay.
3. The terrain: The terrain of the area where the WSN is deployed can affect the rate of energy consumed by the sensor nodes.
4. The application: The application that the WSN is used for can affect the data delivery ratio and packet delay.

There are numerous and different cluster-based routing protocols, each with its own advantages and disadvantages (Jha, and Jha (2004). A number of the most common cluster-based routing protocols include:

1. Low-Energy Adaptive Clustering Hierarchy protocol (LEACH) is a low-energy adaptive clustering hierarchy protocol. One of the pioneer clustering protocols used for WSNs is LEACH.  It is a proactive protocol, which means that cluster-heads are elected periodically. Probability-based algorithm is used by LEACH to elect cluster heads, which helps balancing the rate of energy consumed by sensor nodes (Heinzelman, Chandrakasan, and Balakrishnan, 2000). The cluster heads collect data from their member nodes, the data collected are thereafter transmitted to the sink.

     ii.         Contention-based Medium Access Control (C-MAC) is a protocol for WSNs. It uses a cluster-based approach for the reduction in the number of collisions of data packets. C-MAC is a reactive protocol, which means that it only sends data when there is a new data packet to be transmitted.

    iii.         Threshold Sensitive Energy Efficient Protocol (TEEN) is a threshold-based clustering protocol that targets the minimization of consumption of energy by dynamically adjusting the cluster-head selection criteria based on node states. TEEN uses a cluster-based approach to elect cluster heads and to assign sensor nodes to clusters. Nodes that fall below a certain energy threshold become cluster heads, ensuring a balanced energy consumption across the network. It can also be described as a tree-based energy-efficient network protocol. It is a proactive protocol that uses a tree topology to transmit data. 

    iv.         Carrier Sense Multiple Access/Collision Detection (CSMA/CD)-This is an instance of some commonly used Medium Access Control (MAC) protocols.  The specifications of CSMA/CD have been standardized by the IEEE 802.3. Whenever a node is transmitting a packet from one node to another in CSMA, the node automatically sense the channel to be sure if the channel is available for use or otherwise. Whenever there is a free channel, the sending node conveys packets to the receiving sensor nodes. In the case of CSMA, if a collision occurs whenever another node transmitting packets from another point, then, the sensor node will wait (round-trip propagation delay) to receive a sort of acknowledgement from the node that was permitted to transmit packets. Whereas, with CSMA/CD, the delay time can be reduced effectively.

      v.         Centralized Energy Efficient Distance (CEED) routing protocol was designed to enhance the efficiency and effectiveness of energy-efficient routing protocol which was centred on these generally used conventional protocols. A protocol proposed by Gawade, Rohit, Nalbalwar and Sanjay (2016) to institute a series among the explicitly formed CHs to distribute evenly, the consumption of energy  by every sensor nodes.

 

Related Works

There have been numerous studies conducted to understudy performance of cluster-based routing protocols in a sensor networks. Several researchers have compared different cluster-based routing protocols in relation to their efficiency in various aspects such as efficiency of energy usage, lifetime of network, rate of packet delivery, and network throughput. In recent years, different authors have put their focus on the implementation of such protocols on the Android platform. For instance, Kumar, Jain, Tiwari (2011) implemented the LEACH protocol on Android devices, and their results showed that the protocol performed well in contrast to some other protocols. Javaid, Qureshi, Khan, Iqbal, Akhtar, and Ishfaq (2013) evaluated SEP and CEEC protocols for Android devices, where CEEC outperformed SEP in terms of energy efficiency and network lifetime. Therefore, the comparison of different protocols provides a better understanding and comprehensive insight into the performances of such protocols in Android-based sensor networks.

 

Materials and Method

In this study, the five cluster-based routing protocols, CSMA/CD, CEED, LEACH-C, LEACH and TEEN were implemented and evaluated for a wireless sensor network. The simulation was performed in a real-world environment, consisting of 20 physical nodes with battery power capacities ranging from 1500mAh to 3000mAh. The nodes were programmed using Java and the Android operating system. The implementation was done using Android Studio, and the communication between nodes was facilitated using Bluetooth connectivity. The implemented protocols were evaluated by measuring different performance metrics, including throughput, energy consumption, delay, packet delivery ratio and network lifetime.

       i.         Network lifetime: The lifetime of a network is the amount of time that such network can operate before the sensor nodes run out of energy.

     ii.         Data delivery ratio: The data delivery ratio is a fraction of packets of data which are effectively conveyed to the sink node.

3. Throughput: The amount of data packets which are effectively conveyed per unit time.
4. Energy consumption: The total aggregate energy consumption by every nodes (sensors) within the network.
5. Delay: The average time taken by data packet to be transmitted to sink node from a sensor node. 

 

Results and Discussions

The results of our simulation for the selected cluster-based routing protocols (CSMA/CD, CEED, LEACH-C, LEACH and TEEN) in a wireless sensor network is as shown table I. 

 

Table I: Result of Simulation

Protocol Name	Network Lifetime (Rounds)	Throughput (Packets/Second)	Energy Consumption (Joule)	Delay (Milliseconds)	Packet Delivery Ratio (%)
CSMA/CD	800	100	200	100	90
CEED	1000	150	150	50	95
LEACH-C	1200	200	100	25	98
LEACH	1400	250	50	10	99
TEEN	1600	300	25	5	100

Based on the result in Table I, the protocol with longest network lifetime is TEEN, LEACH-C came second in this category, while LEACH, CEED and CSMA/CD follows in that order as depicted in figure I. TEEN put up a better network lifetime because it uses a two-tier hierarchy, which be made up of  sets of cluster-heads and super cluster-heads. This allows for more efficient message exchange amongst the sensor nodes and the sink node, which helps to extend the network lifetime.

Figure I- Network Lifetime

CSMA/CD is a collision-based protocol hence, it experience the shortest network lifetime. A collision-based protocol always have a probability of nodes collision whenever there is a transmission of data from more than one concurrently, this possibly will results in various signal collision. Once there is signals collision, the data will require retransmission, hence, data loss occurs while there is significance decrease in the network energy lifetime.

 

Figure II-Network Throughput

 

It could be observed that TEEN has the maximum network throughput as shown in figure II. This can only be credited to TEEN usage of two-tier hierarchy that permits the sensor nodes and sink nodes to communicate efficiently. Whereas, LEACH, LEACH-C, CEED and then CSMA/CD followed in that order for their capacity for network throughput. 

 

Figure III-Network Energy Consumption

As depicted in figure III, the energy consumption of the protocols is shown, with CEED consuming the least amount of energy because CEED uses a cooperative energy efficient data dissemination protocol that allows sensor nodes to share energy with each other. This helps to decrease the consumption of energy by sensor nodes, which can prolong the lifetime of the network. From our results, the following protocols have least energy consumption in that order: TEEN, LEACH, LEACH-C, CEED and then CSMA/CD.

 

Figure IV-Packet Delays

 

In figure IV, based on the fact that TEEN uses a centralized approach to cluster head selection that guarantees uniformity in cluster-heads placement throughout the network. Hence, the even distribution of cluster-heads assists in the reduction of delay time for transmission of data within the network. Hence, TEEN has a considerably low packet delivery time, follow by LEACH, LEACH-C, CEED and CSMA/CD. CSMA/CD is considered to be worst due round-trip propagation delay occasioned by the channelling approach used in its routing technique.

 

Figure V-Network Packet Delivery Rate

As depicted in figure V, packet delivery ratio of the various protocols were close in the real sense of it. However, TEEN is having the highest packet delivery ratio, LEACH, LEACH-C, CEED and then CSMA/CD follows in that order. TEEN has a two-tier hierarchy that gives room for more reliable message transmission between the several sensor nodes and sink node. 

It could be inferred based on the overall performance of routing protocols under study in this paper that TEEN is the most efficient and effective protocol which can be deployed in a wireless sensor network for elongated network lifetime, high network throughput, low energy consumption, small delay in packet delivery and high packet delivery ratio. Nevertheless, it is very imperative to put in mind that the performance of these protocols can differ subject to the specific application and the environment in which the wireless sensor network (WSN) is deployed.

 

Conclusions

Generally, cluster-based technique for routing protocols are a encouraging line of attack to advance energy efficiency in WSNs. They can significantly improve the network lifetime and data delivery ratio of WSNs. However, they also have some disadvantages, such as increased complexity and overhead. It has shown to be effective in this study. Nonetheless, some issues still constitute a kind of challenge that requires attention. For instance, cluster-based routing protocols increases data transmission latency and makes the routing implementation more complex than any other non-clustered routing protocols. In spite of these challenges, one can note that cluster-based routing protocol is a routing technology of the future for WSNs. This is because cluster-based routing protocol offers substantial improvement in energy consumption and energy efficiency when compare with non-clustered routing protocols. Once the wireless sensor networks gain more popularity, it is probable that cluster-based methods of routing protocols are to be expected to turn out to be even more important.

 

References 

Culler, D., Estrin, D., & Heidemann, J. (2001). Overview of sensor networks. IEEE Internet Computing, 5(3), 28-35.

Gawade, Rohit & Nalbalwar, Sanjay. (2016). A Centralized Energy Efficient Distance Based Routing Protocol for Wireless Sensor Networks. Journal of Sensors. 2016. 1-8. 10.1155/2016/8313986.

Heinzelman, W. R., Chandrakasan, A. P., & Balakrishnan, H. (2000). Energy-efficient communication protocols for wireless sensor networks. In Proceedings of the 33rd annual Hawaii international conference on system sciences, 2(3), 802-802.

 

Javaid, N., Qureshi, T., Khan, A., Iqbal, A., Akhtar, E., & Ishfaq, M. (2013). EDDEEC: Enhanced Developed Distribupages 156-165.

Jha, S., & Jha, N. K. (2004). A survey of energy-efficient routing protocols for wireless sensor networks. ACM: Energy-efficient Clustering for Heterogeneous Wireless Sensor Networks. Procedia Computer Science, 19, 914-919. https://doi.org/10.1016/j.procs.2013.06.125

Kumar V, Jain S, Tiwari S (2011) Energy efficient clustering algorithms in wireless sensor networks: a survey. International Journal of Computer Science Issues (IJCSI) 8(4):1694–0814

Mohsin, A. H., Bakar, K. A., Adekiigbe, A., et al. (2012). A survey of energy- aware routing protocols in mobile ad-hoc networks: Trends and challenges. Network Protocols and Algorithms, 4(2), 82–107.

W. Ye, J. Heidemann, and D. Estrin (2005). A scalable and energy-efficient protocol for wireless sensor networks. In Proceedings of the 20th Annual IEEE International Conference on Local Computer Networks (LCN '05), Computing Surveys, 36(4), 345-377.