Elsevier

Computer Communications

Volume 29, Issue 12, 4 August 2006, Pages 2230-2237
Computer Communications

Design of a distributed energy-efficient clustering algorithm for heterogeneous wireless sensor networks

https://doi.org/10.1016/j.comcom.2006.02.017Get rights and content

Abstract

The clustering Algorithm is a kind of key technique used to reduce energy consumption. It can increase the scalability and lifetime of the network. Energy-efficient clustering protocols should be designed for the characteristic of heterogeneous wireless sensor networks. We propose and evaluate a new distributed energy-efficient clustering scheme for heterogeneous wireless sensor networks, which is called DEEC. In DEEC, the cluster-heads are elected by a probability based on the ratio between residual energy of each node and the average energy of the network. The epochs of being cluster-heads for nodes are different according to their initial and residual energy. The nodes with high initial and residual energy will have more chances to be the cluster-heads than the nodes with low energy. Finally, the simulation results show that DEEC achieves longer lifetime and more effective messages than current important clustering protocols in heterogeneous environments.

Introduction

Recent technological advances in hardware have enabled the deployment of tiny, low-power sensors with limited on-board signal processing and wireless communication capacities. Wireless sensor networks (WSN) become increasingly useful in variety critical applications, such as environmental monitoring, smart offices, battlefield surveillance, and transportation traffic monitoring. In order to achieve high quality and fault-tolerant capability, a sensor network can be composed of hundreds or thousands of unattended sensor nodes, which are often randomly deployed inside the interested area or very close to it [1].

Since WSN is usually exposed to atrocious and dynamic environments, it is possible for the loss of connectivity of individual nodes. Conventional centralized algorithms need to operate with global knowledge of the whole network, and an error in transmission or a failure of a critical node will potentially cause a serious protocol failure [2]. On the contrary, distributed algorithms are only executed locally within partial nodes, thus can prevent the failure caused by a single node. It is realized that localized algorithms are more scalable and robust than centralized algorithms. As each sensor node is tightly power-constrained and one-off, the lifetime of WSN is limited. In order to prolong the network lifetime, energy-efficient protocols should be designed for the characteristic of WSN. Efficiently organizing sensor nodes into clusters is useful in reducing energy consumption. Many energy-efficient routing protocols are designed based on the clustering structure [3], [4]. The clustering technique can also used to perform data aggregation [5], [6], which combines the data from source nodes into a small set of meaningful information. Under the condition of achieving sufficient data rate specified by applications, the fewer messages are transmitted, the more energy is saved. Localized algorithms can efficiently operate within clusters and need not to wait for control messages propagating across the whole network. Therefore localized algorithms bring better scalability to large networks than centralized algorithms, which are executed in global structure. Clustering technique can be extremely effective in broadcast and data query [7], [8]. Cluster-heads will help to broadcast messages and collect interested data within their own clusters.

In this paper, we study the performance of the clustering algorithms in saving energy for heterogeneous wireless sensor networks. In the sensor network considered here, each node transmits sensing data to the base station through a cluster-head. The cluster-heads, which are elected periodically by certain clustering algorithms, aggregate the data of their cluster members and send it to the base station, from where the end-users can access the data. We assume that all the nodes of the sensor network are equipped with different amount of energy, which is a source of heterogeneity. It could be the result of reenergizing the sensor networks in order to extend the network lifetime [9]. The new nodes added to the networks will own more energy than the old ones. Even though the nodes are equipped with the same energy at the beginning, the networks cannot evolve equably for each node in expending energy, due to the radio communication characteristics, random events such as short-term link failures or morphological characteristics of the field [9]. Therefore, WSN are more possibly heterogeneous networks than homogeneous ones. The protocols should be fit for the characteristic of heterogeneous wireless sensor networks. Currently, most of the clustering algorithms, such as LEACH [10], PEGASIS [11], and HEED [12], all assume the sensor networks are homogeneous networks. These algorithms perform poorly in heterogeneous environments. The low-energy nodes will die more quickly than the high-energy ones, because these clustering algorithms are unable to treat each node discriminatorily in term of the energy discrepancy. In [9], SEP scheme is proposed for the two-level heterogeneous wireless sensor networks, which is composed of two types of nodes according to the initial energy. The advance nodes are equipped with more energy than the normal nodes at the beginning. SEP prolongs the stability period, which is defined as the time interval before the death of the first node. However, it is not fit for the widely used multi-level heterogeneous wireless sensor networks, which include more than two types of nodes.

In this paper, we propose and evaluate a new distributed energy-efficient clustering scheme for heterogeneous wireless sensor networks, which is called DEEC. Following the thoughts of LEACH, DEEC lets each node expend energy uniformly by rotating the cluster-head role among all nodes. In DEEC, the cluster-heads are elected by a probability based on the ratio between the residual energy of each node and the average energy of the network. The round number of the rotating epoch for each node is different according to its initial and residual energy, i.e., DEEC adapt the rotating epoch of each node to its energy. The nodes with high initial and residual energy will have more chances to be the cluster-heads than the low-energy nodes. Thus DEEC can prolong the network lifetime, especially the stability period, by heterogeneous-aware clustering algorithm. Simulations show that DEEC achieves longer network lifetime and more effective messages than other classical clustering algorithms in two-level heterogeneous environments. Moreover, DEEC is also fit for the multi-level heterogeneous networks and performs well, while SEP only operates under the two-level heterogeneous networks.

The remainder of the paper is organized as follows. In Section 2, we briefly review related work. Section 3 describes the heterogeneous network model. Section 4 presents the detail of DEEC algorithm and argues the choice of its parameters. Section 5 shows the performance of DEEC by simulations and compares it with LEACH and SEP. Finally, Section 6 gives concluding remarks.

Section snippets

Related work

There are two kinds of clustering schemes. The clustering algorithms applied in homogeneous networks are called homogeneous schemes, and the clustering algorithms applied in heterogeneous networks are referred to as heterogeneous clustering schemes. It is difficult to devise an energy-efficient heterogeneous clustering scheme due to the complicated energy configure and network operation. Thus most of the current clustering algorithms are homogeneous schemes, such as LEACH [10], PEGASIS [11],

Heterogeneous network model

In this section, we describe the network model. Assume that there are N sensor nodes, which are uniformly dispersed within a M × M square region (Fig. 1). The nodes always have data to transmit to a base station, which is often far from the sensing area. This kind of sensor network can be used to track the military object or monitor remote environment. Without loss of generality, we assume that the base station is located at the center of the square region. The network is organized into a

The DEEC protocol

In this section, we present the detail of our DEEC protocol. DEEC uses the initial and residual energy level of the nodes to select the cluster-heads. To avoid that each node needs to know the global knowledge of the networks, DEEC estimates the ideal value of network life-time, which is use to compute the reference energy that each node should expend during a round.

Simulation results

In this section, we evaluate the performance of DEEC protocol using MATLAB. We consider a wireless sensor network with N = 100 nodes randomly distributed in a 100m × 100m field. Without losing generalization, we assume the base station is in the center of the sensing region. To compare the performance of DEEC with other protocols, we ignore the effect caused by signal collision and interference in the wireless channel. The radio parameters used in our simulations are shown in Table 1. The protocols

Conclusions

We describe DEEC, an energy-aware adaptive clustering protocol used in heterogeneous wireless sensor networks. In DEEC, every sensor node independently elects itself as a cluster-head based on its initial energy and residual energy. To control the energy expenditure of nodes by means of adaptive approach, DEEC use the average energy of the network as the reference energy. Thus, DEEC does not require any global knowledge of energy at every election round. Unlike SEP and LEACH, DEEC can perform

Li Qing received his B.S. degree from Sichuan Normal University, Chengdu, China, in 1996. He received his M.S. degree from the University of Electronic Science and Technology of China in 2003 and now is a Ph.D. student at the school of Computer Science and Engineering, in the University of Electronic Science and Technology of China. His major research interests include sensor networks, mobile communications, and network security.

References (16)

  • V. Mhatre et al.

    Design guidelines for wireless sensor networks: communication, clustering and aggregation

    Ad Hoc Network Journal

    (2004)
  • F. Akyildiz et al.

    A survey on sensor networks

    IEEE communications magazine

    (2002)
  • H. Chan, A. Perrig, ACE: an emergent algorithm for highly uniform cluster formation, in: Proceedings of the First...
  • P. Krishna et al.

    A cluster-based approach for routing in dynamic networks

    ACM SIGCOMM Computer Communication Review

    (1997)
  • B. McDonald, T. Znati, Design and performance of a distributed dynamic clustering algorithm for Ad-Hoc networks, in:...
  • V. Mhatre, C. Rosenberg, D. Kofman, R. Mazumdar, N. Shroff, Design of surveillance sensor grids with a lifetime...
  • W.R. Heinzelman, A.P. Chandrakasan, H. Balakrishnan, Energy-efficient communication protocol for wireless microsensor...
  • Sze-yao Ni, Yu-chee Tseng, Yuh-shyan Chen, Jang-ping Sheu, The broadcast storm problem in a mobile Ad Hoc network, in:...
There are more references available in the full text version of this article.

Cited by (0)

Li Qing received his B.S. degree from Sichuan Normal University, Chengdu, China, in 1996. He received his M.S. degree from the University of Electronic Science and Technology of China in 2003 and now is a Ph.D. student at the school of Computer Science and Engineering, in the University of Electronic Science and Technology of China. His major research interests include sensor networks, mobile communications, and network security.

Qingxin Zhu received his Ph.D. degree from Ottawa University, Canada, in 1993. In 1995, he did postgraduate research in Electronic Engineering Department of the Ottawa University and the Computer Department of the Carlton University. He is currently the professor with the University of Electronic Science and Technology of China. His major research interests include computer communication technology and network security, algorithm design, system stimulation, optimizing control and search.

Mingwen Wang received his B.S. degree from Zhejiang University, China, in 1994. He received his M.S. degree from the University of Electronic Science and Technology of China in 2003 and now is a Ph.D. student at the school of Computer Science and Engineering in the UEST. His current research interests are in network security and next-generation wireless networks.

View full text