See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/224208432

A Roadmap Towards the Implementation of an Efficient Online Voting System

in Bangladesh

Conference Paper · January 2011

DOI: 10.1109/CISE.2010.5676848 · Source: IEEE Xplore

CITATIONS READS

0 5,289

3 authors, including:

Mohammad Shabbir Hasan Abdullah Al Mahmood

Virginia Polytechnic Institute and State University UNSW Sydney
32 PUBLICATIONS   225 CITATIONS    29 PUBLICATIONS   96 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Microrecycling of packaging waste View project

Genesis-indel View project

All content following this page was uploaded by Mohammad Shabbir Hasan on 23 May 2014.

The user has requested enhancement of the downloaded file.

 A Roadmap towards the Implementation of an
Efficient Online Voting System in Bangladesh

Mohammad Shabbir Hasan1, Abdullah Al Mahmood2, Quazi Farhan3

Panacea Research Lab
Dhaka, Bangladesh.
E-mail: 1shabbir_cse03@yahoo.com, 2sajibcseofkuet@gmail.com, 3quazifarhan@gmail.com

Abstract—Online voting system has become an interesting topic purchased by a candidate. The voting system must also be
in recent years and most governments in Europe and elsewhere tamper-resistant to thwart a wide range of attacks, including
are taking steps to experiment as well as implement it to a great ballot stuffing by voters and incorrect tallying by insiders.
extent. Recent election problems in Bangladesh have sparked Voting systems are hard to make trustworthy because they
great interest in managing the election process through the use of
have strong, conflicting security requirements: Integrity and
internet. The present form of voting in general elections is
founded upon entirely paper based and largely manual voting confidentiality [2]. Integrity means election results must be
procedures. New technology for managing elections may entail assured so that all voters are convinced that votes are counted
several advantages. It may enhance the voters’ scope for correctly. Any attempt to corrupt the integrity of an election
participating in the election as well as create scope for more must be detected and correctly attributed. On the other hand,
rapid tallying of votes and distribution of seats. This enables the confidentiality means voters must be assured about the privacy
election commission to promptly announce the election results of their votes, prevent selling of votes, and defend voters from
within a short time. The risk of error in vote-tallying can also be coercion. Integrity is easy to obtain through a public show of
largely eliminated and through all these steps, the elections can hands, but this destroys confidentiality and confidentiality can
be made fair. In this paper we have proposed an online based
be obtained by secret ballots, but this fails to assure integrity.
voting system to eliminate the problems and bottlenecks of the
traditional voting system in Bangladesh. Because of the civic importance of elections, violations of
these requirements can have dramatic consequences. A voting
Keywords- Online Voting System, Voting in Bangladesh, system must be comprehensible to and usable by the entire
Electronic Voting System. voting population, regardless of age, infirmity, or disability.
Providing accessibility to such a diverse population is an
I. INTRODUCTION important engineering problem and one where, if other
security is done well, online voting system could be a great
Online voting system is considered as a particularly
improvement over current paper systems.
interesting topic in the field of information security research.
Previous researches on electronic voting system [3, 4, 5, 6,
This may be due to a number of reasons. Election is the way
and 7], argue that assurance in electronic voting systems is too
through which people choose their representatives and express
hard to obtain and that their deployment creates unacceptable
their preferences for how they will be governed. Naturally, the
risks. Our work, however, was inspired by the possibility that
integrity of the election process is fundamental to the integrity
the proposed online voting systems could be more trustworthy
of democracy itself [1]. Again, elections usually have high
and also efficient in terms of cost and time than the traditional
media coverage, especially if something goes wrong.
voting system in Bangladesh.
Furthermore, online voting system seems to have a unique
The organization of the paper is as follows: section 2
combination of security requirements: voters need to be briefly discusses the steps of the proposed system. Section 3
authenticated as well as results need to be verifiable. So the describes various features of the proposed system. A
online voting system should be sufficiently robust to withstand comparison between the present voting system in Bangladesh
a variety of fraudulent behaviors and also must be transparent and the proposed system is mentioned in Section 4 and Section
and comprehensible enough so that voters and candidates can 5 includes conclusion.
accept the results of an election. Unsurprisingly, previous
election history in Bangladesh is littered with examples of II. SYSTEM OVERVIEW
manipulated elections in order to influence their outcome.
A good voting system must satisfy a number of sometimes The proposed system comprises of several steps. The
competing criteria. The anonymity of a voter’s ballot must be system is accessible from two sides: (a). Election Commission
preserved, both to guarantee the voter’s safety when voting who is the administrator and (b). The voter. There are some
against a malevolent candidate, and to guarantee that voters steps which are automated i.e. not accessible from any side.
have no evidence that proves which candidates received their Figure 1 shows the possible input output scenario of the
votes. The existence of such evidence would allow votes to be proposed system and total system architecture is presented in

978-1-4244-5392-4/10/$26.00 ©2010 IEEE

detail in Figure 2. Steps of the proposed system are described
in the following sub sections.
A. Adding Voter Information
In this proposed system, information of each voter is added
according to their National Identity Number. This National
Identity Number is unique for each voter and this number is
also used to identify the constituency of the voter. After adding
information, an auto generated e-mail is sent to the e-mail
address of the voter notifying him/her about the information
and this e-mail also contains a computer generated password
which can be used by the voter for login as well as for
changing password and setting verification keys. Here
verification keys are used to protect “Vote Purchase” and to
ensure security. On screen keyboard is used to take the new
password from user. The system will not take the password
typed in any keyboard other than the onscreen keyboard.
Purpose of using on screen keyboard is to prevent capturing
password through any software if the voter changes password
from any cyber café or in any public computer. Discussion
about the use of verification keys is presented later.

Figure 2. Details architecture of the proposed system

D. Start Election
This is an automated phase. During the stipulated date and
time, election is started. Voter can cast their vote within this
time period. If anyone wants to cast vote before or after the
specific time period, an alert message is shown.

E. Cast Vote
In this phase, voter has to login first. After logging in with
Figure 1. Possible Input-Output scenario of the proposed Online Voting
System the national id and password, the constituency of the voter is
determined from the information stored in the voter database.
An E-Ballot paper is created automatically for that
B. Adding Candidate Information constituency from information stored in candidate database.
Candidate information is added according to the This ballot paper contains the candidate name along with their
constituency. Here each candidate is assigned an auto profile picture, party name with party logo and a radio button
generated code to identify uniquely. Party symbol and to select the candidate for casting vote. There is also a “No”
candidate profile image are also added with other information option if the voter is not interested to cast vote to any of the
in this phase. available candidates. Figure 3 shows a sample ballot paper for
a constituency.
C. Setting Election Date and Time Voter selects radio button of the corresponding candidate
and finally press the “Cast Vote” button. A security checking is
In this phase, starting time and ending time of election
done to verify whether the vote is casted actually by the voter
along with election date are set by the election administrator.
himself. A discussion about the security checking is presented
Bangladesh is in the “Dhaka” time zone which has +6.00 offset
in the next section.
from GMT. The voting server is configured according to this
time zone. Voters who reside outside of Bangladesh can cast F. Security Checking
their vote according to the local time in Bangladesh during When a voter presses the “Cast Vote” button in the e-ballot
Election Day. paper after selecting the suitable candidate, a security checking
is done internally. This security checking is one of the most
striking points of this proposed system as it is required to
protect “Vote Purchase” by any candidate. Figure 4 shows the Voter can enter verification key before election by logging
steps of the security checking process. in his/her account as we mentioned in section A. Number of
verification keys to be used are selected by the voter, but
among them, only one key is used as “Real Key” and rest of
them are treated as “Fake Key”. Among these keys voter will
select which one will be used as “Real Key”. In case of any
attempt to “Purchase Vote” by any candidate, the voter has the
option to hide “Real Key” and supply only “Fake Keys” to the
candidate. If vote is casted by using fake keys, notification will
be shown as if vote is casted successfully, but there will be no
update in database. So if the voter wants to cast his/her vote,
he/she can cast vote using the real key in any time within the
voting period. This process will reduce the tendency of
purchasing vote by any candidate and hence make the election
process fair.
G. Auto Tabulation and Publication of Result
During the election period, result is tabulated automatically
and after the election period is over, winner in each
constituency is declared automatically. In this proposed
system, results can be published immediately which leads to a
huge saving of time than the existing method.
III. FEATURES OF THE PROPOSED SYSTEM

The features of the proposed system are:

• No need to stay in the queue for a long time to cast
vote, anyone can cast his/her vote from anywhere
through internet.
• Increased voter turnout.
• The cost for arranging election will be reduced as there
is no need to prepare any voting center or no need to
Figure 3. A sample E-Ballot paper for a constituency
manage huge manpower.
• Impervious security checking to verify whether vote is
casted by the voter himself.
• Coercion resistant voting system.
• Result can be published within a very short time as the
system automatically tabulate the results and declare
winners.
• Overall, the election will be fair enough and
transparent.

IV. A COMPARATIVE STUDY BETWEEN THE PROPOSED

SYSTEM AND THE EXISTING SYSTEM
To compare with the existing system, the proposed system
was implemented in the trade union election of Technocrats
BD on February 2010. The election system was totally online
based and the previous elections were paper ballot based like
other traditional elections. Comparative statistics between
these two election methods are shown in Table 1.
Table 1 shows that the proposed system increases voter
turnout than the previous method as online voting system is
very much convenient than the traditional voting system. In
this proposed system voter cast their vote from any place
through internet rather waiting in a queue for long time and
Figure 4. Flow Diagram of the Security Checking Process
 Figure 5. Comparison of costing between traditional and proposed voting system

this is the main reason for making the voting system in the proposed system decreases rapidly than the traditional
interesting which leads to the increase in voter turnout. system which is shown in Figure 5.
From Table 1 it is also clear that there is no way of invalid Finally, after this comparative study, it can be said that
vote in the proposed system which is a case in the traditional better turnout and strong security system with low setup cost
one. As the e-ballot paper contains radio button for each have made the proposed voting system better than the
candidate, voter can select only one candidate at a time for the traditional system.
same post. If the voter doesn’t select corresponding radio
button of any candidate, “NO” vote is casted which means the V. CONCLUSION
casted vote doesn’t correspond to any candidate. So there is no This paper describes the design, implementation and
way of either casting multiple votes in the same ballot paper at evaluation of an internet based efficient voting system in
a time or cast vote without selecting any radio button. That’s Bangladesh. To our knowledge, this has not been done before.
the reason of 100% valid vote in the proposed system. The proposed system is coercion resistant and provides
stronger security than the traditional voting system.
TABLE I. COMPARATIVE STATISTICS BETWEEN TWO ELECTION
Experimental results show that cost, tabulation time and
METHODS security can be practical for real-world elections.
Deriving from the previously-known voting scheme, the
Paper Based Election on Online Based Election
Criteria
January 2006 System on February 2010 implemented system ensures security and efficiency through
Total 1054 1173 technical advances. But perhaps the most important
Voter contribution of this work is evidence that secure online voting
Turnout 797 (75.6%) 1053 (89.8%) system could be made possible and we are optimistic about the
Valid 753 (94.47%) 1053 (100%) future of the proposed system which abides by the principals of
Votes electoral system.
Total 84,320 82,110
Cost
(in REFERENCES
Banglad [1] Tadayoshi Kohno, Adam Stubblefield, Aviel D. Rubin, Dan S. Wallach,
eshi “Analysis of an Electronic Voting System”, Johns Hopkins University
Taka) Information Security Institute Technical Report, TR-2003-19, July 23,
2003.
Another striking point of the proposed system is the [2] Michael R. Clarkson Stephen Chong Andrew C. Myers, “Civitas:
reduction of cost to arrange an election. Lots of manpower and Toward a Secure Voting System”, Computing and Information Science
funds are required to arrange the election in the traditional Technical Report TR 2007-2081, May 2007.
manner. But in this online voting system, there is no need of [3] David L. Dill, Bruce Schneier, and Barbara Simons, “Voting and
setup cost for any voting center as voters can cast vote from technology: Who gets to count your vote?”, Communications of the
ACM, vol. 46(8), Aug. 2003, pp. 29–31.
their personal computer and hence no manpower is required to
[4] David Evans and Nathanael Paul, “Election security: Perception and
maintain any voting center. Few people are enough to manage reality” , IEEE Security & Privacy, vol. 2(1), Jan. 2004, pp. 24-31.
the complete process of the proposed system as many steps are [5] David Jefferson, Aviel D. Rubin, Barbara Simons, and David Wagner,
fully automated. Huge reduction of cost is clear from Table 1. “Analyzing Internet voting security”, Communications of the ACM, vol.
In the traditional system, 80 Bangladeshi Taka is required 47(10), Oct. 2004, pp. 59–64.
for each voter where as only 70 Bangladeshi Taka is required [6] Rebecca Mercuri. “Statement on electronic voting”.
in the proposed system when total voter is 1000. With the http://www.notablesoftware.com/ RMstatement.html, 2007.
increase in total number of voters, total cost to arrange election [7] Aviel D. Rubin, “Security considerations for remote electronic voting”,
Communications of the ACM, vol. 45(12), Dec. 2002, pp. 39–44.

View publication stats