SECURITY ENHANCEMENT OF E-VOTING SYSTEM CHAPTER-4 Security is one of the most important issues for a reliable and trusted Electronic Voting system. The term E-Voting is used in different ways and it encompasses all voting techniques involving electronic voting equipments like internet voting, polling stations with electronic booths etc. This chapter presents a model for security enhancement of e- voting system. Rest of the chapter is organized as follows: Section 4.1 presents the basic introduction related to security in E-voting followed by the proposed E-voting system for security enhancement in Section 4.2. Results and discussion has been presented in Section 4.3 based on the algorithm defined in Section 4.2. Finally chapter has been summarized in Section 4.4. 4.1 INTRODUCTION A voting system that can be proven correct has many concerns. The basic reasons for a government to use electronic systems are to increase election activities and to reduce the election expenses. Still there is some scope of work in electronic voting system in terms of checking the authenticity of voters and securing electronic voting machine from miscreants. Biometrics is used to verify the identity of a an individual person based on a physiological or behavioral behavior. It has the capability to distinguish between a valid and invalid person. Biometrics features are distinct and cannot be lost. Moreover, since individual needs to be verified should be available physically during the verification process, biometrics proved to be more reliable and more capable than traditional knowledge-based and token-based techniques. Electronic Voting Machine is a simple device used to store votes in place of ballot papers which were used earlier in conventional voting system. It is a very simple electronic machine that can be handled easily by the officials on election duty as well as 34
by the voters. Being a autonomous and standalone machine, no one can interfere with its programming and manipulate the results. Advantages of EVM [59] over the traditional ballot paper/ box system are: It completely removes the possibility of vote casting by an unauthorized voter leading to unpleasant situation like disputes and election petitions. It makes the procedure of counting the votes faster than the traditional system. It minimize to a great scope the quantity of paper used thus saving a large number of trees making the procedure eco-friendly. It minimizes the cost of printing. In this research work, a model to enhance the security of electronic voting system has been presented by incorporating fast and accurate biometric technique to prevent an unauthorized person to vote. 4.2 PROPOSED E-VOTING SYSTEM A model for e-voting based on biometric technique is presented in this section. Biometrics has been widely used in various applications such as criminal identification, prison security electronic banking, e-commerce [14]. Biometric authenticity requires comparison between registered biometric sample and a new captured biometric sample. During enrollment, a sample of the biometric trait is encapsulated, managed by a computer and stored for future comparison. In identification mode, the biometric system identifies a person from the complete enrolled population by scrutinizing a database for a match based merely on the biometric. A system can also be used in confirmation mode, where the biometric system validates a person s declared identity from their formerly enrolled pattern. This is also known as one-to-one matching [43]. The proposed model uses biometrics in the confirmation mode during electronic voting. Fingerprint systems can also be used in mode. Fingerprints are the ridge and furrow patterns shown in Figure 4.1 on the tip of the finger used to find the identity of individual s. 35
The proposed model assumes that the unique id has been assigned to each voter (Citizen of India). Each perspective voter must have a unique fingerprint registered in the database. Working Algorithm: System implementation has been done with the underlying algorithm: Step-1: Open login window. Step-2: Enter P.O.id and password. /* id and password filled by Presiding Officer(P.O.) */ Step-3: If P.O. id and password verifies Main window /* main window will appear automatically */ Else Print- wrong id and password Go to step 2 Ridge Bifurcation Figure 4.1: Ridges and Furrows 36
Step-4: Enter user id and password. /* for voter, if the system has been successfully started by P.O. */ count=0; Step-5: Verify userid and password If user id and password matches If votedflag=false Print- enter thumb impression count++ If thumb impression valid go to step 6 Else Print-This time your thumb impression didn t match Go to Block 1 Else Print- you have already voted Go to step 7 Else Print Userid and password don t match Go to step 7 Block 1 Print (chances left, 3-count ) If(count <=3) Go to step 5 Print- You are not authorized Go to step 7 Step-6: Print- please enter your vote Set votedflag = true Print-thank you 37
Step-7: Check current time (It should be less than closing time 05:00PM.) If (current time < 05:00 PM) Go to step 4 /* another user can vote*/ Else Go to step 8 /*After 5:00 PM, no person is eligible to restart the system for voting procedure.*/ Step-8: Exit 4.3 RESULT AND DISCUSSION This proposed model has been simulated successfully on Java platform. Figure 4.2 to Figure 4.5 are few snapshots of steps involved in the implementation of the presented model. 38
Figure 4.2: Home page of Biometric Electronic Voting System 39
Figure 4.3: Login Window of Presiding Officer 40
Figure 4.4: Voting Procedure 41
Figure 4.5: Vote Counting 42
We have implemented the proposed model by integrating it with FTA 5454(A10) - Fingerprint Time and Attendance system as shown in Figure 4.6. It can store 3000 fingerprint templates and 50000 transaction records. Figure 4.6: Device used for Fingerprint 4.4 SUMMARY In this chapter, a model for security enhancement of electronic voting is presented wherein fingerprint is embedded as biometrics for voter identification. Performance of the system model has been validated on the test data set and it is observed that system is able to authenticate the individual s identity successfully and also ensures that no voter can cast his/her vote more than once. In the next chapter, a model for designing a secure electronic voting system using finger print has been presented. 43