JWSN 2015, 1, 1-0009
Journal of Wireless Sensor Networks
Enhanced Security Technique for Wireless Sensor
C.O. Iwendi, A.R. Allen
WSN Consults Ltd, Aberdeen Scotland and Communications & Optical
Engineering Research Group, University of Aberdeen, Fraser Noble Building, AB24 3UE Aberdeen, Scotland; E-Mail: email@example.com;
Received: 27 May 2015/ Accepted: 07 July 2015/Published: 27 July 2015 Abstract: The lightweight computational nodes being used in WSN pose particular challenge for many security applications. This paper investigates a number of security techniques and novel implementations appropriate for WSN nodes, including various trade-offs such as implementation complexity, power dissipation, security flexibility and scalability.
Keywords: Contiki, WSN, Blom, AES, Cooja
Wireless sensor networks (WSN) are widely used by diverse communication systems . However, WSN are constrained by limited resources which include limited communication range and limited energy. Furthermore, the lightweight computational nodes that are deployed in WSN create challenges for many current security applications. This paper investigates a new security technique and suitable implementation for WSN, including various trade-offs such as implementation complexity, power dissipation , security flexibility and scalability. This enhanced security application is aimed at developing a network that has efficient and flexible key distribution scheme secured enough to prevent algorithmic complexity and denial of service attacks while conserving energy . A re-evaluation of Blom’s key generation was implemented . The aim was to provide a secured pre-allocation and distribution of keys by every single node in WSS with a piece of data that would be generating a pairwise key with every other node in the network as shown in Fig.1; implemented in Contiki operating system due to its advantages of multi threading. The resultant pairwise key generated with a piece of data was then used for generating an AES-encrypted pairwise key with every other node in the network. The novelty of this research includes:
(a) A key allocation and Key Pre-distribution smart security technique to secure Base station and the sensor nodes before deployment.
(b) A unique mesh routing mechanism that also sets up the system for encryption with trade-off in power consumption
(c) A unique combination of advanced encryption standard (AES) with key management technique.
(d) A unique alternative security mechanism with TCP/IP configuration The remainder of the paper is organized as follows:
The next section presents the related works by researchers in this field, followed by a description of the geometrical solution of key management scheme in Section 3. Section 4 describes the implementation of the enhanced security in Contiki OS. Section 5 showcases the implementation in real sensor motes and compares the solution achieved. Finally, section 6 draws the conclusion.
2 Related works
WSN is useful in the automotive industry; advanced lighting and night vision in addition to creative technology that help conform to the economy of fuel reduction, environment and road safety have so far been implemented. The combination of WSN and microcontrollers technologies have led to the production of devices such as passenger restraints, airbags, anti -lock brakes, stability control, etc . Wireless sensor networks do not differ from all other network embedded systems in the areas of confidentiality, integrity and availability of data but rather the WSN communication structure illustrates the setting under which nodes or clusters send signals to the base station in the presence of an intruder.
Moreover, Wireless sensor network facilitates many other numerous application areas for instance, tactical observation by military unattended sensor networks, elderly and patient monitoring by body area networks (BANs) and by building automation and control networks (BACnets) . Researchers are still concerned if WSN can actually be secured considering that messages can be intercepted based on what was sent, whom it was sent to, when the message was sent and where the message was sent from and to . They are also considering the level of integrity of the system in terms of the
actions of the intruders to insert, delete and/or change the message that was sent. How accurate the final information will be with consideration of denial of service is currently being investigated. Therefore, improvements in the latest electronic technology has also brought about the idea of deploying small, low power, low-cost sensor devices in the Oil and Gas Industries. Various security issues that have received consideration in ad hoc networks [7, 8] are not applicable to WSN due to architectural disparity between these two types of networks . Also, quite a number of other security schemes are already being proposed. If the cluster formation design by  currently being investigated is found to be effective in saving energy, there is need to look at the security between the individual nodes, neighbouring nodes and the base station. Moreover, the majority of current security applications proposed is not viable due to low memory to run complicated algorithm after running the Operating System (OS) and other applications. This is due to the fact that all security services are assured by cryptography; therefore, security in WSN will amount to a sizeable amount of energy consumption due to cryptographic functions .
Many WSN deployments are security sensitive and attacks against them may lead to damage to health and safety of people. Denial of service creates conditions for hardware failures, resource exhaustion, bugs, malicious attacks and environmental conditions that could reduce the functionality or totally eliminating a networks ability to perform as expected . It has been argued by Wood and Stankovic  that the possible security attack on the network layer is due to neglect, greed,
homing, misdirection, authorisation, probing, blackholes and monitoring. Further research and more detailed study by Karlof and Wagner  put specific names and methodologies to these attacks. These network attacks are classified as follows:
a) Spoofing, replaying or altering clustering messages: This is the most direct attack in the network. A malicious node by spoofing, altering or replaying routing information in the network can mount a DoS attack by making itself part of the routing formation and then creating routing loops, attracting or repelling traffic, generating false error messages, shortening or extending source routes or partitioning the network . It can also select the forward packets to reduce the probability of detection. One method to combat the threat is by authentication and antireplay protection. Another technique is by multipath routing. The multipath routing sends the same data over multiple paths to give it a higher probability of reaching its destination .
3 Geometrical Solutions for Key Management Scheme
Blom proposed a key generation system that is symmetric and allows nodes to generate a pairwise secret key with a secured communication among the uncompromised nodes.. This is the reason for using Blom’s pairwise key pre-distribution idea in the WSS base station to construct (λ+1) × N matrix G over finite field GF(q). Although, Blom’s scheme does not propose how G matrix should be generated, the only requirement is that its columns must be linearly independent in order to achieve the λ-secure property. In our implementation, we
decided to use the Vandermonde [14, 15 and 16] matrix since it is easy to implement and meets the requirement. A random symmetric (λ+1) × (λ+1) matrix D over finite field GF(q) should also be constructed. The matrix D is considered to be private information and none of the nodes in the WSS is supposed to know the D matrix while G can be known. In order to calculate the final K matrix from which we can extract the pairwise keys, the base station should first multiply the D and G matrix, then transpose and multiply the result with the G matrix as follows:
2 3 2 2 2 2 3 2 2 3 1 1 1 1 N N N s s s s s s s G s s s s s (1)
2 1 n n c n S S G S (2)
Along with G matrix the base station also generates random symmetric matrix over a finite field GF q( )but unlike G matrix, the D
matrix is kept secret and must not be disclosed to any sensor node or adversary in the wireless sensor network at any time.
The D matrix must be generated all at once and it has to be stored in the base
station the entire lifespan of the network.
is matrix that is the result of computation where is transpose of (DG). Each row of a matrix is disclosed the one sensor in the wireless sensor network in order to allow the node to generate its pairwise key with other sensors.
1,1 1,2 1,( 1) 1,1 1,2 1, 2,1 2,2 2,( 1) 2,1 2,2 2, ( 1),1 ( 1),2 ( 1),( 1) ( 1),1 ( 1),2 ( 1), T N N N D D D G G G D D D G G G A D D D G G G (4)
In practice, since the base station does not preserve the entire G matrix, a row of A matrix Ar 1,1 1,2 1,( 1) 2,1 2,2 2,( 1) ( 1),1 ( 1),2 ( 1),( 1) D D D D D D D D D D
A is generated from the column of G matrix Gc at a time when a sensor is connecting to the network as follows:
( 1) ( 1) ( 1) 1, 2, ( 1), 1 1 1 ( ), ( ),... ( ) r k k k k k k k k k A D G D G D G
is then disclosed to the node where it is used for computation of pairwise keys.
K AG DG G (7)
Since K is a N x N symmetric matrix, therefore if node n would get n : th row from K and wish to communicate with node m it will just take the m : th column of the n : th row and use the value as the pairwise key, since K is symmetrical, therefore
, , m n n m K K (8) 1,1 1,2 1,( 1) 1,1 1,2 1,( 1) 2,1 2,2 2,( 1) 2,1 2,2 2,( 1) ,1 ,2 ,( 1) ( 1),1 ( 1),2 ( 1), N N N N A A A G G G A A A G G G K A A A G G G
K is therefore a symmetric matrix. This means that where is an element of K on i-th row and j-th column. Kj i, is used as a pairwise key between
nodes i and j .
In practice when node i wants to communicate with node j , it sends the seed i
s . Node j is then able to generate the column Gc of matrix G and using the row Ar of matrix A it received from the base station compute pairwise
keyKi j, as follows:
During implementation, it was discovered that the scheme will not be effective with the original idea due to the following reasons: We would have to store information for every single node. We would have to know the exact number nodes in the network. When adding a new node we would have to update information on every single node. Therefore instead of storing n:th row of the K matrix we should store n:th of the A = (DG)T matrix. That gives us (λ+1) columns on every single node. When two nodes want to communicate, they will first exchange their seeds SN where S is a primitive element of GF(q) and n is the number of nodes in the network, therefore the K generated becomes:
1,1 1,2 1, 1,1 2,1 ,1 2,1 2,2 2, 1,2 2,2 ,2 ,1 ,2 , 1, 2, , N N N N N N N N N N N N K K K K K K K K K K K K K K K K K K
( 1) . , , 1 ( ) n m n k k m k K A G
That gives us sufficient scalability so we can add new nodes without breaking down the running network. The base station keeps a counter of number of nodes in the network and when a new node n wants to join the network, the base station increments the counter, generates a column of G for the node n
( 1) , , , 1 ( )i n ( i k k n); 1...( 1) k DG D G i
Therefore, the base station keeps track only of the matrix seed Sn where n is the number of currently connecting node and generates column Gc for the node. The
base station also sends the seed SN and the n : th row of A to the node n. The node n is now ready for distribution.
The problem with Blom’s Pairwise Key Pre-Distribution Scheme is that it provides only log q2 bits long encryption key where q is the top of the finite field
GF q so if we wanted to have 128 bits long encryption key we would have to have G, D, A and K matrices over the finite field.
(2 ) (3.40282367 10 )
GF GF (11)
So instead of generating G, D, A and K matrices over finite field Gcwe decided to generate set of 16 D matrices. Let s be the secure property
1,1 1,2 1, 1,1 1,2 1, 1,1 1,2 1, 2,1 2,2 2, 2,1 2,2 2, 2,1 2,2 2, ,1 ,2 , ,1 ,2 , ,1 ,2 1 1 1 2 2 2 16 16 16 1 1 1 2 2 2 16 16 16 , ... 1 1 1 2 2 2 16 16 16 s s s s s s s s s s s s s s s s D D D D D D D D D D D D D D D D D D D D D D D D D D D s s, (12)
Therefore we must calculate set of 16 A matrices. Let s be the secure property (λ + 1) 1,1 1,2 1, 1,1 1,2 1, 1,1 1,2 1, 2,1 2,2 2, 2,1 2,2 2, 2,1 2,2 2, ,1 ,2 , ,1 ,2 , ,1 ,2 1 1 1 2 2 2 16 16 16 1 1 1 2 2 2 16 16 16 , ... 1 1 1 2 2 2 16 16 16 s s s s s s s s s s s s s s s s A A A A A A A A A A A A A A A A A A A A A A A A A A A s s, (13)
Now, we have to provide each node with a matrix P composed of rows from each of the A matrices. Let represents matrix P composed of rows of 16 A matrices provided to nodei. ,1 ,2 ,( 1) ,1 ,2 ,( 1) ,1 ,2 ,( 1) 1 1 1 2 2 2 16 16 16 i i i i i i i i i i A A A A A A P A A A (14)
That way, it is possible to generate 16 pairwise keys with each node in the network and provide the keys to the AES algorithm. Let Gc be the column of generated
out of seed si provided by node i ( 1, 2,..., 16)A A A be the set of generated A matrices and PK be the column provided to AES as a encryption key, then:
( 1) ( 1) ( 1) , , , , , , 1 1 1 ( k c 1 ),r k ( k c 2r k),..., ( k c 16r k) k k k PK G A G A G A
This paper proposes a number of techniques for Enhanced Security Implementation as follows:
4 Description of the proposed method through example
For the implementation of the proposed enhanced security mechanism, Micaz motes were used along with Contiki operating system . Micaz has a radio frequency range between 2400MHz to 2483.5MHz with maximum data rate of 250kbps. They receive and transmit power at 19.7mA and 17.4mA respectively. Micaz uses DSSS-O-QPSK modulation with a receive sensitivity of -94dBm. The outdoor range is between 75m to 100m and operates with multichannel advantage. Given that most of the nodes have a limited energy supply, the transceiver therefore functions as a tolerable balance between data rate and energy consumption . Micaz uses a micro controller with a clock frequency of 7.37 MHz. Random access memory (RAM) is 4KB and ROM 128KB. It was originally connected with TinyOS as the operating system. But in this paper the TinyOS was
replaced with Contiki OS. All these features make Micaz an appropriate research platform suitable for this experiment. Contiki is a lightweight operational system developed for constrained platform, such as sensor nodes. Its basic configuration fits in less than 2 KiB of RAM and 40 KiB of ROM, and provides two communication stacks (µIP  and Rime ) as well as multi threading functionalities. All its modules, drivers and user application are implemented in C programming. While Although TinyOS is generally used but Contiki has a much easier learning curve for the developer and researchers, because it is based on a widely known programming language.
Key Allocation: K keys in the Base station using unicast implements a mechanism whereby new nodes can be added without breaking down the network, giving sufficient scalability. The method involves sending a request to the Base station. When base station receives a unicast request, it generates a key and sends it to the node. When node receives the key, it stops to unicast to the base station and start to unicast to all its neighbours.
Key Pre-distribution: Easy deployment of security key algorithm implements the deployment of the generated k pairwise key from the Base station. When base station receives a unicast request, it generates key and sends it to the node. When node receives the key, it stops to unicast to the base station and start to unicast to its entire neighbours. The first thing that happens is that it will request a unicast seed. When the node receives a seed request: it unicast its seed to the sender. When sender receives the seed, it says something like "Sending message to node x.y using pairwise key k". When the node receives a message it says something like "Received message from x.y, using pairwise key k". This system is an efficient security key algorithm method for pre-distribution.
Mesh preallocation, routing and set-up stage for encryption intelligently implements Mesh pre-allocation routing mechanism. The code implements pairwise key preallocation, routing, and when communication is initialized, first the sender sends it’s pairwise key seed to the receiver. The receiver responds with its pairwise key seed and then the sender sends the message. The seeds are used for initiating encryption using AES. The mesh module sends packets using
multi-hop routing to a specified receiver in the network using the Rime communication stack .
Encrypting the Pairwise Key with AES- Advanced encryption Standard: key management is a security mechanism, an added version to make the network further secured. It implements a modification to advanced encryption standard and improves the performance of AES, thereby accommodating the key management system with low memory capability and energy consumption. The new modification was implemented in Cooja simulator . 16 bytes encryption was also generated to concur with the standard of AES needed for the encryption and decryption.
Alternative security mechanisms: The paper proposes an interface with sensor nodes within an IP address: integration of IP-enabled WSN nodes and Non-IP-enabled nodes and to showcase the deployment and operational architecture during implementation. The process will involve a modification of uIP open source stack  to accommodate the sensor nodes code using the key distribution method as basis for routing mechanism. This method provides security for the IP address in sensors without creating problems of overhead, and memory capacity. Data can therefore, be easily transferred with the actualization and implementation of this contribution.
This algorithm was implemented with Micaz hardware using Contiki OS and run with Cooja, a flexible java-based simulator that is designed to simulate a network of sensors running the Contiki OS. It shows a strong key pre-allocation mechanism using linear cryptography, high performance and low memory usage. The key generated in the base station for 20 nodes is shown in Figure 3. In the experimentation, authentication, integrity and privacy were achieved. The overall energy consumption can also be reduced significantly in the hardware to encourage the massive distribution of sensor nodes in areas that cannot easily be reached like under- water sensing and hurricane monitoring.
--- [K] ---- 4670 52433 38702 55127 40113 53384 40793 52298 39267 51083 38082 50093 39640 54923 38860 53198 38817 51768 40053 54262 52433 588519 432400 614687 443222 593964 459064 589420 442914 573252 423342 558092 440660 612017 435822 599255 430517 581797 443315 608983 38702 432400 323117 456169 332051 449153 334924 436342 322853 422487 318547 422876 330344 457289 326897 441561 318635 431382 326464 450119 55127 614687 456169 638872 474561 615423 483352 609893 467200 596293 453673 575843 469211 638175 465330 630975 459291 610882 472569 633608 40113 443222 332051 474561 351548 455393 356313 442888 338403 442849 331257 431595 339244 473757 329693 455494 334509 447697 349330 464545 53384 593964 449153 615423 455393 606864 465500 599466 453505 579691 431347 569417 449305 615812 448064 613656 444249 596955 458971 617620
40793 459064 334924 483352 356313 465500 359377 458379 342803 443545 337418 440045 348842 479830 339982 463991 330656 453627 350749 470115 52298 589420 436342 609893 442888 599466 458379 595370 447106 568858 418831 561653 441855 610387 437592 597417 432371 578308 446533 606989 39267 442914 322853 467200 338403 453505 342803 447106 332000 430506 322632 426276 336035 467134 324213 452014 324456 440784 334713 458291 51083 573252 422487 596293 442849 579691 443545 568858 430506 556000 425289 539270 433575 598974 427148 584362 419010 569306 435333 595057 38082 423342 318547 453673 331257 431347 337418 418831 322632 425289 310243 406065 320039 445744 316856 438244 320874 431575 327044 446665 50093 558092 422876 575843 431595 569417 440045 561653 426276 539270 406065 532547 425410 582212 425407 583427 421689 555243 431073 571786
Fig 3 Implementation of pre-distribution keys in Micaz using Contiki OS and run with Cooja
99 ID:3 Starting 'Receive key' 1383 ID:1 unicast message_sent to: 1 2384 ID:1 unicast message_sent to: 2 3386 ID:1 unicast message_sent to: 3
3386 ID:3 Key received: 233546169 881213426 960794652 3312744758 3960419106 1528885228 2744654277 2898887135 2096983033 2017393389
5389 ID:1 unicast message_sent to: 5 6390 ID:1 unicast message_sent to: 6
6390 ID:6 Key received: 3435371504 3473273106 3570396769 3272078321 3771928141 3975712307 4029857203 887080301 1194018876 975187624
7392 ID:1 unicast message_sent to: 7 8393 ID:1 unicast message_sent to: 8 9395 ID:1 unicast message_sent to: 9
9395 ID:9 Key received: 2290326245 429090836 1763846052 1988035590 953447484 3403739712 2573806782 3387090282 487355124 3951102926
10396 ID:1 unicast message_sent to: 10
10396 ID:10 Key received: 354192334 3176386636 3966658186 3346232414 1655610952 1609361703 623970377 3769441602 3274226730 1209534441
Fig. 4 Key distribution
Fig. 6 Key Management Encrypted with AES
The experimental results are encouraging. They are depicted in Figure 3, 4, 5 and 6 showing different stages of security implementations which includes pairwise Key generation, Key distribution in the Base station, simulation with Cooja, sensor node placement and Key management encrypted with AES. It shows that a wireless sensor network can be secured from denial of service attack with a minimal usage of memory capacity and low energy consumption with higher reliability and better authentication technique. The key management technique also shows a great scalability advantage and maintains a significant impact on the performance metrics. The enhanced security method is therefore expected to
demonstrate even better secured network with the integration of IP-enabled WSN nodes
We tested our two applications using Cooja that comes preinstalled on the Instant Contiki distribution. It provides various tools to debug ContikiOS applications without having physical sensors by emulating them. Firstly we have chosen strategic position of one base station and three nodes. The programs were designed to communicate to the node of ID 3, so we placed node 3 next to the base station, node with ID 2 on a place so it can reach node with ID 3 to test single-hop communication and the last node with ID 4 so it can reach node with ID 2 but cannot reach node with ID 3 to test the multi-hop communication. Table 1 shows the positions of nodes 1, 2, 3, 4
Table 1 Sensor node positioning
Sensor ID X Y Z 1 2 3 4 104.8 66.33 83.55 34.25 53.3 66.37 31.93 82.17 0 0 0 0
After the positioning of the sensor nodes, Fig. 7 below shows the illustrative positioning of the wireless sensors nodes in the network.
Fig. 7 Node Placements in the Network
Also, the following screenshots as shown in Figure 8 illustrates the visibility of each node in the wireless sensor network.
Fig. 8 Node Placement Simulations
The table below shows speeds of single-hop and multi-hop communication for key pre-distribution and communication phase for both uIP and Rime protocols.
Table 2 Key Pre-distribution and Communication Phase
Pre-Allocation Communication Pre-Allocation Commu
nicatio n Rime uIP 46ms 91ms 7ms 7ms *NA *A 1845m s 15ms
* Pre-allocation in Rime's example is made using unicast that correctly does not allow pre-allocation on multi-hop connections. Because uIP communication uses Rime's mesh connection by default, it automatically allows pre-allocation phase on multi-hop and further steps are needed to forbid this behaviour.
Without a clear perspective of the risk involved in WSS and options available to manage the risks by intruders, it is infeasible to have a defensible network. Therefore, developing a security protocol can be quite challenging. The enhanced security application for WSS if properly implemented can bring total significant change to the scope of wireless sensor networks and increase its usefulness. Faster clustering routing algorithm with TCP/IP integration is also under consideration.
1. Tavares, J.; Velez, F. J.; Ferro, J.M. Application of Wireless Sensor Networks to Automobiles. Measurement Science Review, vol. 8, No. 3, 2008, pp. 65-70. DOI:10.2478/v10048-008-0017-8
2. Iwendi, C.O. Power consumption trade-off analysis in an OFDM
communication chain. IEEE AFRICON, 2011, vol., no., pp.1-6, 13-15 Sept. 2011
3. Iwendi, C.O. & Allen, A.R. Wireless Sensor Network Nodes: Security and Deployment in the Niger-Delta Oil and Gas Sector. International Journal of Network Security & Its Applications, vol 3, no. 1, pp. 68–79 Jan. 2011
4. Blom, R. An optimal class of symmetric key generation systems. Proceedings of EUROSRYPT’84 pp. 335-338 1984
5. Pathan, A.S.K.; Lee, H.; Hong, C.S. Security in wireless sensor networks: Issues and challenges. The 8th International Conference on Advanced Communication Technology, ICACT 2006. Volume 2, 20-22 Feb. 2006. Page(s):6 pp.
6. Undercoffer, J.; Avancha, S.; Joshi, A.; Pinkston, J. Security for sensor networks, CADIP Research Symposium, 25-26 Oct. 2002
7. Zhou, L.; Haas, Z.J. Securing ad hoc networks. IEEE Network, Volume 13, Issue 6, Nov. – Dec. 1999, pp. 24 – 30.
8. Strulo, B.; Farr, J.; Smith, A. Securing mobile ad hoc networks – A motivational approach. BT Technology Journal Volume 21 Issue 3, 2003, pp. 81 -89
9. Murshed, G. Energy efficient data gathering in wireless sensor network, PhD transfer report, University of Aberdeen, Scotland, February 2009 10. Kaplantzis, S. Security models for wireless sensor networks. PhD
Conversion Report, Monash University, Australia, March 2006 11. Raymond, D.R.; Midkiff, S.F. Denial-of-Service in wireless sensor
networks: Attacks and Defences. IEEE Pervasive Computing Magazine. Volume 7, Issue 1, Jan. – Mar. 2008. Page(s):74 – 81
12. Xu, W.; Trappe, W.; Zhang, Y.; Wood, T. The feasibility of launching and detecting jamming attacks in wireless networks. Proc. 11th Annual Int’l Conf. Mobile Computing and Networking, ACM Press, 2005, pp. 46 – 57 13. Wood, A.D.; Stankovic, J.A. Denial of Service in Sensor Networks.
Computer, vol. 35, issue. 10, Oct. 2002, pp. 54 – 62. DOI: 10.1109/MC.2002.1039518
14. Stajano, F.; Anderson, R. The Resurrecting Duckling: Security Issues for Ad Hoc Wireless Networks. Proc. 7th Int’l Workshop Security Protocols, Springer, 1999, pp. 172 – 194.
15. Karlof, C.; Wagner, D. Secure Routing in Wireless Sensor Networks: Attacks and Countermeasures. Proc. 1st IEEE Int’l Workshop Sensor Network Protocols and Applications, IEEE Press, 2003, pp. 113 – 127. DOI: 10.1109/SNPA.2003.1203362
16. Yu, Y.; Govindan, R.; Estrin, D. Geographical and Energy Aware Routing: A Recursive Data Dissemination. Protocol for Wireless Sensor Networks, tech. report UCLA/CSD-tr-01-0023, Computer Science Dept., Univ. of California, Los Angeles, 2001.
17. Newsome, J.; Shi, E.; Song, D.; Perrig, A. The Sybil attack in sensor networks: Analysis & Defenses. Third International Symposium on Information Processing in Sensor Networks, IPSN, 26 – 27 April 2004 Page(s): 259 – 268
18. Nakayama, H.; Ansari, N.; Jamalipour, A.; Nemoto, Y.; Kato, N. On data gathering and security in wireless sensor networks. IEEE Sarnoff
Symposium, April 30 2007 Page(s): 1-7
19. Deng, J.; Han, R.; Mishra, S. Defending against Path-Based DoS Attacks in Wireless Sensor Networks. Proc. 3rd ACM Workshop Security of Ad Hoc and Sensor Networks, ACM Press, 2005, pp. 89 – 96.
20. Hui, J.W.; Culler, D. The Dynamic Behaviour of a Data Dissemination Protocol for Network Programming at Scale. Proc. 2nd ACM Conf. Embedded Networked Sensor Systems, ACM Press, 2004, pp. 81 – 94. 21. Dutta, P.K.; Hui, J.W.; Chu, D.C.; Culler, D.E. 2006. Securing the deluge
Network programming system. In Proceedings of the 5th international conference on Information processing in sensor networks (IPSN '06). ACM, New York, NY, USA, 326-333. DOI=10.1145/1127777.1127826 22. Neagoe, V.E. Inversion of the Van der Monde Matrix. IEEE Signal
Processing Letters, Volume 3, Issue 4 August 2002, Page(s): 119 – 120. 23. Oruk, H. LU Factorization of the Vandermonde Matrix and its
Applications. Science Direct: Applied Mathematics Letters 20 (2007) 982-987
24. Merouane, D.; Øyvind, R. Applications and Fundamental Results on Random Vandermond Matrix. Journal for Computing Research Repository. CORR. Vol. Abs/0802.3 July 2008.
25. Iwendi, C.O.; Allen, A.R. Wireless sensor network nodes: Security and deployment in the Niger-Delta oil and gas sector” International Journal of Network Security & Its Applications (IJNSA), vol. 3, no. 1, January 2011 pp 68–79
26. Heinzelman, W.R.; Chandrakasan, A.; Balakrishnan, H. Energy-efficient communication protocol for wireless microsensor networks. Proceedings of the 33rd Annual Hawaii International Conference on Systems Sciences Jan 4-7, 2000 Page(s): 10pp. vol 2
27. Dunkels, A. Full TCP/IP for 8-bit architectures. Proc of the 1st
international conference on mobile systems, applications and services, MobiSys ’03, New York, NY, USA: ACM, 2003, pp. 85-98
© By the authors; licensee JWSN. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license.