Redundancy Level Impact of the Mean Time to Failure on Wireless Sensor Network

Transcription

1 (IJACSA) International Journal of Advanced Coputer Science and Applications Vol. 8, No. 1, 217 Redundancy Level Ipact of the Mean Tie to Failure on Wireless Sensor Network Alaa E. S. Ahed 1 College of Coputer and Inforation Sciences, Al Ia Mohaad Ibn Saud Islaic University (IMSIU) Riyadh, Kingdo of Saudi Arabia 2 Departent of Electrical Engineering, Faculty of Engineering (Shoubra), Benha University, Egypt Mostafa E. A. Ibrahi 1 College of Coputer and Inforation Sciences, Al Ia Mohaad Ibn Saud Islaic University (IMSIU) Riyadh, Kingdo of Saudi Arabia 2 Counication and Coputer Engineering Dept, Benha Faculty of Engineering, Benha University, Egypt Abstract Recently, wireless sensor networks (WSNs) have gained a great attention due to their ability to onitor various environents, such as teperature, pressure sound, etc. They are constructed fro a large nuber of sensor nodes with coputation and counication abilities. Most probably, sensors are deployed in an uncontrolled environent and hence their failures are inevitable all ties of work. Faulty sensor nodes ay cause incorrect sensing data, wrong data coputation or even incorrect counication. Achieving a reliable wireless sensor networks is a ost needed goal to ensure quality of service whether at deployent tie or during noral operation. While Nodes redundancy is considered as an effective solution to overcoe nodes failures, it ay negatively affect the WSN lifetie. Redundancy ay lead to ore energy drains of the whole syste. In this paper, the ipact of redundancy level on the Mean Tie to Failure (MTTF) of a clustered based wireless Sensor Networks (WSNs) is investigated. An expression that can be used to deterine the ost suitable redundancy level that axiizes the network MTTF is derived and evaluated. Keywords Wireless sensor network; reliability; clustering; fault tolerant; Mean Tie to Failure (MTTF) I. INTRODUCTION Wireless sensor networks (WSNs) are currently being considered for any counications applications, such as environental onitoring, agricultural onitoring, achine health onitoring, surveillance, and edical onitoring. These networks provide a great ability to better understand the surrounding physical phenoenon and translate their effect to be analyzed to give a ore accurate way to control our surroundings. However, there are any research QoS challenges that need to be studied and analyzed to achieve scalable reliable and dependable networks. Traditionally, redundancy is a technique belongs to replica anageent approaches which are designed to overcoe nodes failure and to achieve reliable syste. Redundancy ost probably leads to axiize the availability of the nodes. In sensor networks, sensor nodes can fail due to any reasons. Aong those reasons are: battery power depletion, hardware proble and cruel environental circustances in which the node is running. Nods are failed according to a variable failure rate that start by a low value and continuously increase and ay reach to value that cause a coplete failure to the syste. One solution to overcoe nodes failure is to use reliable hardware that safely send, receive and store data. This solution is expensive solution that contradicts with the objective of having cheap wireless sensor. Replicating data on spatially separated nodes is a good ethod for iniized the failure probability. Traditional solutions for replica systes are to use a pre-deterined fixed nuber of replicas [1]. Because of the dynaic behavior of failure, selecting a suitable redundancy level is an iportant issue. In soe cases sall nuber of replicas is better and on the other hand in soe situations higher nuber of replicas is ore suitable. Because of that, the process of selecting the suitable redundancy level should be adaptive to achieve less energy consuption and higher degree of availability. One way to deterine the ost suitable redundancy level is to have a coplete knowledge of the network which is not a practical solution and doesn t cope with scalability issue. Also, it requires uch inforation to be circulated around the network to keep up to date with the current situation which consues ore energy. The solution is to find a solution to deterine the ost suitable redundancy level that achieves a reliable request while axiizing the Mean Tie to Failure MTTF of the syste. The contribution in this paper is as follows: 1) Describe a syste odel that captures all stages that affect the energy consuption l of entire the network. 2) Propose an expression that is used to deterine the ost suitable redundancy level to axiize the ean tie to failure MTTF. 3) Provide a nuerical data analysis that ipleents the proposed odel to show the tradeoff between the energy consuption and the reliability iproveent. The reainder of the paper is organized as follows: In Section 2, the related work is introduced. The syste odel assuptions are presented in Section 3. Section 4 shows the analysis and nuerical results. Finally, the paper ends with the conclusion in Section 5. II. RELATED WORK The general proble of achieving reliable WSN has been covered intensively in literature. Providing the concept of redundancy is considered an efficient trend in order to achieve reliable wireless sensor network. Redundancy types can be 217 P a g e

2 (IJACSA) International Journal of Advanced Coputer Science and Applications Vol. 8, No. 1, 217 subdivided into three categories, spatial redundancy, physical redundancy and teporal redundancy. Spatial redundancy is based on geographically distribute the sensor nodes over the field area in such a way that guarantee the replication of the resources all over the coverage area. Spatial redundancy is suitable for WSN since the inforation read by any sensor node is available and stored in other different nodes. This kind of redundancy can be used in ipleenting fault tolerant approached to iprove the WSN reliability level. Spatial redundancy is drawn by large nuber of etrics [2], [3]. In [4] authors eployed the graph theory and cut sets in order to define the redundancy degree of the wireless sensor network. It is the ost possible nuber of nodes that its disappearance doesn t affect the accuracy of the easured data. Physical redundancy is a hardware redundancy and is considered one of the easiest ways to achieve reliable systes. It is ainly based on having a dense nuber of independent sensor nodes to be distributed and cover certain filed area [5], [6]. Teporal redundancy is ostly used to get better accuracy of the sensor nodes readings and to overcoe the transient failures. In Wireless sensor networks, the reading coe fro any single node ay be not precise. The solution is to iprove the whole syste reliability by using ultiple reading fro the sae sensor node [7]. In [8] authors present a ethod for evaluating the syste reliability by generating a fault tree for the whole network at the tie of failure occurrence. The provided solution is independent of the underlying network technology. However, the odel doesn t take the energy consuption into consideration. In [9], [1] a power consuption saving approach is introduced. Power consuption saving is considered one way of achieving reliable wireless sensor network. The approach adopts the use of Colored Petri Net to odel the network and provide a good way to be integrated with different applications. In [11] authors present an efficient ethod that could be used in transitting the data through the WSN. The ethod is ainly tends to iniize the nuber of lost packets by using ultiples paths for data transissions to overcoe either the node or the path failure. One proble with this ethod is the bottleneck produced as a result of having ore than one path per packets. Clustering is one of the ost used techniques in reducing the energy consuption in WSNs and consequently increase the sensor node life tie. One of the earliest WS counication algoriths is LEACH [12]. It stands for Low Energy Adaptive Clustering Hierarchy Algorith. The existence of clustering leads to iniize the energy usage with the cluster but at the sae tie accelerate the energy consuption of the cluster head CH. To solve this proble, the concept of balanced clustering is introduced in [13]. Another cluster based approach called Distributed Weight-Based Energy-Efficient Hierarchical Clustering (DWEHC) is introduced in [14]. It is based on foring well balanced clusters by adopting the hierarchical level of a node depending with the ability to iniize the energy requireents by taking into consideration two factors, the range of a cluster and the path to the CH. In [15] energy an Energy-Efficient Unequal Clustering (EEUC) echanis is introduced. It is based on dividing the nodes into unequal cluster sized. This technique is dynaic in the sense that, it picks any node randoly and checks whether it lies within the copetitive radius or not. So, the node would join the ost suitable cluster that saves its energy. A lot of energy-efficient clustering algoriths can be found is available in [16]. III. SYSTEM MODEL ASSUMPTIONS In this section, we describe the proposed syste odel and its assuptions. The network is fored by a collection of liited-energy Sensor Nodes (SNs) scattered randoly on a certain geographical area. The WSN odel used in this work is adopted fro [17]. 1) Initially each sensor node has an initial energy E init. 2) The sensor nodes are scattered randoly through a square field area. 3) It is assued that each node has failure probability f(t) where (where <f(t) <1). 4) LEACH algorith [11] is used for routing and anaging the network clustering. First the cluster head CH is elected per cluster. The CH is responsible for adinistrating the counication within the cluster, aggregate data read by sensors nodes, and set up the data sent to the sink nodes. This algorith runs in the for of rounds. Due to the extra responsibilities of the CH node, it is exposed to lose their energy quicker than other nodes. For this reason, the CH role is rotated around all eber of the cluster according to a certain probability p. the value of p depends on the cluster size. 5) In this odel it is assued that the counication channels are reliable and never fail. 6) It is assued that the reliability of the syste is expressed as follow [18]. ( ) ( ) (1) Where, N is the nuber of identical sensor nodes per cluster = the nuber of functional nodes R(t) = the reliability of individual odule. R -of-n(t) is the reliability of that (N-) or fewer nodes have failed by tie t (or at least are functional). 7) In response to a request, a sensor node sends a essage to its CH node. It is assued that the nuber of hops between 218 P a g e

3 (IJACSA) International Journal of Advanced Coputer Science and Applications Vol. 8, No. 1, 217 the sensor node and the cluster head is N clust. Any request needs sensor nodes to reach a certain redundancy level. The total energy E t for any request has two coponents E 1 and E 2. The first is the energy consued to send the data fro sensor node to the CH and the other is the energy consued by the WSN to transit sensor data fro CH to the sink node. The total energy equation can be expressed as follow [17]: E t = E 1 + E 2 (2) E t = [N hop (E receive + E transit )] + [(E receive + E transit ) + N clust (E i )] (3) 8) The energy consued in executing the clustering process includes the aount of energy due to ulticasting the leadership announceent essage and the energy due to joining the cluster. If p is the probability of becoing a CH, there will be p*n clust sensor node that ulticast the leadership announceent essage which in turn is forwarded by each sensor to the next hop until covering the whole nodes in the cluster. Assue the nuber of hops per cluster is Nhop/cluster Thus, the energy required for broadcasting is [17]: E br = pn clust (E receive + E transit )+N hop/cluster ((E receive + E transit ) (4) By substituting in the MTTF of n According to Gaa and Beta function definition [18]: So, And consequently, 9) The energy consued due to the cluster-join process includes the energy consued by any sensor node to infor the CH about their cluster ebership and the energy consued by the CH to send back an acknowledge essage to the sensor nodes. E join = N clust (E receive + E transit ). (5) 1) Since the clustering process is executed a nuber of rounds N round then the total energy consued in clustering is expressed as [17]: E clustering = N round [E br + E join ] (6) Then, Finally, IV. SYSTEM MTTF The ain objective here is to achieve the ost suitable redundancy level that achieves a reliable request while axiizing the Mean Tie to Failure MTTF. Assue it is required to process a request with required reliability Rq. Assue is the failure rate of different sensor nodes. Since we use out of N reliability odel, the MTTF can be calculated according to the following: Since, By substituting fro (1): ( ) ( ) By introducing we obtain (7) V. ANALYSIS AND NUMERICAL RESULTS In this section we introduce the nueric data analysis that ipleents the proposed odel to show the tradeoff between the energy consuption and the reliability iproveent. Also, we show the effect of changing on the MTTF of the syste. The following paragraph indicates paraeters used along with their default paraeters. These default values are siilar to the values used in [33]. The WSN odel consists of 4 sensor nodes distributed randoly in a square area of 2 by 2. Each sensor node has an initial energy of 1 Joule. The energy consued per bit due to the transitter/receiver radio process- E receive & E transit - is 5nJ/bit. A. and MTTF Analysis Fig. 1 shows the relation between the reliability and the redundancy level under the environent paraeters described in the previous paragraph. It is possible to use other paraeter but still the overall trend is the sae. In previous chapter the reliability analysis is based on adopting and applying the N out of redundant systes to achieve a reliable WSN. The N sensor nodes are assued to be identical 219 P a g e

4 MTTF Eb MTTF (IJACSA) International Journal of Advanced Coputer Science and Applications Vol. 8, No. 1, 217 in all configurations. The N out of redundant syste is considered a sort of parallel syste where represents the iniu nuber of sensors per cluster required to have a fully functional sensing operation during certain tie period T. we assue that sensor failure ay occur according to a failure rate =.1 and T = 2 unit tie. In case =1 the syste is reduced to a fully parallel syste since any single node will do the job. On the other hand when = N, this eans that the syste is act as a series syste since the all ebers in the cluster are required to have a correct operation. Fro that Fig. 1 shows the relation between the syste reliability and the value through out a certain period of tie according to (1). The graph show that as the value increases the reliability level is reduced. This result consistent with the fact that says as increase the syste tends to converge to a series one. The overall reliability of series systes is lower than the reliability of each individual coponent. At the sae tie as the value decrease the syste tends to converge to be a parallel one that ha reliability an iproved reliability level. As a consequence of the previous result, Fig. 2 shows the relation between the value and the MTTF (8). The result shows that the MTTF is decreased as increase. This result is copatible with the result achieved in Fig. 1. As the syste tends to be fully parallel, the MTTF increase since the reliability increases. The following two figures (Fig. 3 and 4) shows the relation between both the syste reliability and the MTTF and the value but with failure rate =.1 and T = 2 unit tie. The figure shows the sae trend achieved in Fig. 1 and Fig. 1. VS with =.1. B. Energy Level Consued Analysis In this section we introduce the nuerical analysis related to the level of consued energy of due to applying the out of n redundant syste. Fig. 5 shows the relation between the energy consued as a result of 24 bits size request. The figure shows three curves that represent the consued energy level with different value of the nuber of interediate nodes within the cluster (N clust ). As the value of increase the energy consued increase. This result is copatible with the fact that as increase, the syste tends to work as a series fashion that involve ore sensor nodes in the process of copleting the request. As the nuber of involved sensors increase the energy consued will increase. Also, Fig. 6 shows the relation between the value and the consued energy in the case of changing the paraeter of the nuber of nodes exists between the cluster head and the sink node (N hop ). The result indicates that as value increase the energy consued increase even with different values of N hop Fig. 3. VS with =.1. Fig. 4. MTTF VS with =.1. M MTTF 25 Nclust = 3 Nclust = Fig. 2. MTTF VS with =.1. Fig. 5. 5Eb VS with = P a g e

5 Eb MTTF Eb (IJACSA) International Journal of Advanced Coputer Science and Applications Vol. 8, No. 1, Fig. 6. E t VS with =.1. Fig. 7. MTTF VS with =.1. Fig. 8. Et vs Rq with =.1. Nhop = 25 Nhop = 2 Nhop = C. Coparison with the Basic Syste ( = 1) In this section we copare the MTTF for both proposed design and the basic syste in which there is no redundancy at R q Vs E t basic syste MTTF-Proposed Eb-Nclust 3 Eb-Nclust 2 Eb-Nclust 1 Rq all. Fig. 7 shows the change of the MTTF in both the basic syste and the proposed design. We notice that the proposed syste give higher syste MTTF copared with the basic design under the set of assued paraeter values characterizing the WSN. D. Consued Energy Effect on the Required Level One of the iportant paraeters that is useful to the WSN designer is to be able to deterine the energy consued Et when a required level of reliability R q is needed to be satisfied. Fig. 8 shows the relation between the R q and Et under the set of assued paraeter values characterizing the WSN. According to the proposed odel, as the required the reliability level increase the nuber of active nodes per cluster decrease and that would cause the energy consued to decrease. This result is shown in Fig. 8. That would consequently help the syste designer is able to have a tentative estiation about the energy consued for the required and needed syste reliability level. One ore result shown in the figure is that as the nuber of interediate nodes within the cluster N clust increase the energy consued per certain required reliability increase. VI. CONCLUSIONS In this paper, a probability odel to calculate the syste reliability, the energy Consuption of a request, and, consequently, the syste MTTF is introduced. Throughout the odel we considered energy consued by the sensor nodes due to request processing, periodic status update processing and cluster foration. The proposed odel is adopting a group counication service represented of foring, anaging and using the cluster. The effect of the redundancy level on MTTF is evaluated. A series of nuerical analysis result that show and validate the proposed odel showing the reliability, MTTF and energy consued in order to achieve a dependent WSN syste are conducted. As a conclusion, the MTTF of the syste increase as the nuber of active coponents per cluster decrease. Also, the paraeter could be used as design paraeter that helps the designer to achieve a syste with a required level of reliability. REFERENCES [1] M. A. Mohaed, W. K. Seah and I. Welch, in wireless sensor networks: A survey and challenges ahead Coputer Networks, vol 79, pp , Elsevier, 215. [2] G. Jaber, R. Kacii and Z. Maeri Exploiting redundancy for energy-efficiency in Wireless Sensor Networks in ninth Wireless and Mobile Networking Conference, IEEE, 216. [3] C. Volosencu, D. Pescaru, L. Jurca and A. Doboli Redundancy and Its Applications in Wireless Sensor Networks: A Survey, WSEAS Transactions on Coputers, Vol. 8, no. 4, pp , April 29. [4] B. Rachid, H. Hafid, k. Bouabdellah A distributed approach using redundancy for wireless sensor networks reconfiguration Models & Optiisation and Matheatical Analysis Journal vol. 1, pp. 84-9, 212. [5] D. Shi, C. Yang, Method of detecting and isolating fault in redundant sensors, and ethod of accoodating fault in redundant sensors using the sae, USPTO Application, Nov. 28. [6] Z. Bojkovic and B. Bakaz A survey on wireless sensor networks deployent, WSEAS Transactions on Counications, vol 7, no 12, pp , Dec P a g e

6 (IJACSA) International Journal of Advanced Coputer Science and Applications Vol. 8, No. 1, 217 [7] G. Jesus, A. Casiiro and A. Oliveira A Survey on Data Quality for Dependable Monitoring in Wireless Sensor Networks Sensors,vol. 17, no. 9, pp. 1-23, 217. [8] I. Silva, L.A Guedes, P. Portugal, F. Vasques and availability evaluation of wireless sensor networks for industrial applications Sensors, vol. 12, pp , 212. [9] A.V.L Dâaso, D. Freitas, N.S Rosa, B, Silva, P.R.M Maciel Evaluating the power consuption of wireless sensor network applications using odels Sensors, vol. 3, pp , 213. [1] A.V.L Dâaso, N.S Rosa, P.R.M Maciel, Using coloured petri net for evaluating the power consuption of wireless sensor network Int. J. Distrib. Sens. Netw. (IJDSN) 214. [11] W. Fang, F. Liu, F. Yang, L. Shu, and S. Nishio, Energy-efficient cooperative counication for data transission in wireless sensor networks, IEEE Transactions on Consuer Electronics, vol. 56, no. 4, noveber 21. [12] S. Xia, M. Tan, Z. Zhao and T Xiang Iproveent for LEACH Algorith in Wireless Sensor Network International Conference of Young Coputer Scientists, Engineers and Educators, pp , Springer, 216. [13] S. A. Nikolidakis, D. Kandris, D.D Vergados, C. Douligeris Energy efficient routing in wireless sensor networks through balanced clustering Algoriths, vol. 6, no. 1, pp , 213. [14] P. Ding, J. Holliday, A. Celik, Distributed coputing in sensor systes Distributed energy-efficient hierarchical clustering for wireless sensor networks; vol. 356, pp Berlin, Gerany: Springer; 25. [15] G. Chen, C. Li, M. Ye, J. Wu An unequal cluster-based routing protocol in wireless sensor networks Wireless Networks, vol. 15 no. 2, pp , 29. [16] B. M. Vallapura, G. P. Nair, K. Thangaani, A convivial energy based clustering (CEBC) solution for lifetie enhanceent of wireless sensor networks, Artificial intelligence and evolutionary algoriths in engineering systes, pp , Springer; 215. [17] A. E. S. Ahed, Analytical odeling for reliability in cluster based wireless sensor networks in Ninth International Conference on Coputer Engineering and Systes (ICCES). IEEE, 2 25, 214. [18] M. Rausand Syste Theory. 2nd edition, Wiley, P a g e