SISY 2010 • 2010 IEEE 8th International Symposium on Intelligent Systems and Informatics • September 10-11, 2010, Subotica, Serbia
A Distributed Smart Card Based Access Control System M. Popa*, A. S. Popa** and M. Marcu* *
“Politehnica” University of Timisoara, Faculty of Automation and Computers, Timisoara, Romania ** “Politehnica” University of Timisoara, Faculty of Mechanics, Timisoara, Romania
[email protected];
[email protected];
[email protected]
Abstract— An access control system has the role to verify and mediate attempts made by users to access resources in a system. An access control system maps activities and resources to legitimate users. This paper presents a distributed smart card based access control system for a building. The system is Internet centered. Each door has a control access point and all these are connected to a server which will grant or not the access. A user requests access by using a smart card which memorizes the user’s unique identity code.
I.
INTRODUCTION
An access control system has the role to verify and mediate attempts made by users to access resources in a system. An access control system maps activities and resources to legitimate users. The main objective of an access control system is to protect system resources against inappropriate or undesired user accesses. The access control systems can be mainly divided in: capability-based and ACL (Access Control List)-based. In the first case, the possession of a reference specific to an object (resource, activity) offers the access to that object. In the second case, the access is granted if the user is on a list associated to an object. The techniques to controlling the access are discretionary and non-discretionary, [1]. The discretionary access control (DAC) leaves the decision to the discretion of the object’s owner or another person who is authorized to control the object’s access. All other access control techniques are NDAC. Two examples are: Mandatory Access Control (MAC): access control decisions are made by a central authority, not by the individual owner of an object and cannot be changed by this. The solution is used in military security. Role-Based Access Control (RBAC): access decisions are based on the roles that individual users have as part of an organization and access rights are grouped by role name and the use of resources is restricted to individuals authorized to assume the associated role. Access control can be implemented using IT infrastructure. The implementation can be at different levels and in many places. For example operating systems protect files and directories by using access control, database management systems control the access to tables through access control systems etc.
978-1-4244-7395-3/10/$26.00 ©2010 IEEE
A well-known implementation refers to door access control systems. Such a system controls and validates the access in the rooms of a building (such as a company, an institution) according to some policies or access rights. This paper presents a distributed smart card based access control system for a building. The system is Internet centered. Each door has a control access point and all these are connected to a server which will grant or not the access. A user requests access by using a smart card which memorizes the user’s unique identity code. The next section presents similar work, the third section describes the proposed solution and the last section outlines the conclusions. II.
RELATED WORK
Access control systems have been described in many papers. Reference [2] describes a solution for controlling the access to private areas by using a low cost embedded controller based smart card consisting in a card reader and a contact less card. There are predefined codes memorized in the internal memory of a microcontroller situated in the card reader. Each microcontroller contains a specific code for a specific user. In reference [3], the aim was to obtain a flexible modular system based on integration of access sensors and on an arbitrary number of stand-alone modules. Such a module consists of a door equipped with an open/close sensor, an identity card reader, a fingerprint reader and a camera. The door unlocks after the two readers have successfully identified and verified the user. The entry process is recorded by the camera and the outputs of the sensors are memorized in a database. Reference [4] proposes a voice based door access control system for building security. The access may be authorized by means of an enrolled user speaking into a microphone attached to the system. The approach is justified by the fact that a person’s voice cannot be stolen, lost, forgotten, guessed or impersonated with accuracy. In reference [5], the smart card approach is applied in connectivity. The authors start from the observation that the traditional approach for smart card based web solutions provide security but has the drawbacks of usability and flexibility. The new approach focuses on
- 341 -
M. Popa et al. • A Distributed Smart Card Based Access Control System
communications and allows web applications to fully utilize smart card capabilities. The solution from reference [6] is a biometrics-based access control system. It provides higher security than other similar solutions because of the Fingerprint Vault scheme which ensures the security of the fingerprint template. The system does not store the user’s fingerprint template in the IC card but a secret k locked by this fingerprint. Only the user who owns the matching fingerprint can retrieve k successfully and therefore validate his identity. Reference [8] describes a security scheme to control access to a building based to the fact that the access is forbidden to a single person but allowed to a group of authorized people whose number satisfies a prescribed threshold. The solution described in this paper consists in a web connected system in which the communication is achieved through the TCP/IP protocol. Each control access point is in fact an IP reader and all are connected to a server. III.
THE ACCESS CONTROL SYSTEM
A. General Presentation The proposed system was aimed to control the access in the rooms of a building. Each door is connected to a control access point (CAP) which unlocks or not the door. A user has a smart card, containing a unique identity code. In order to enter in a room, the user inserts the card in the card reader from the CAP and is allowed or not to enter in the room. The CAPs are connected to a server, through Internet. The server receives the data read from the CAPs and grants or not the accesses according the to access policies implemented. Messages are sent to the users according to the decisions taken. The server implements also the administrator function, which has the possibility to modify the access rights of the users and to add or delete users. The communication is made using the TCP/IP protocol and each CAP has an IP conforming to the network rules. Consequently, each CAP is also an IP reader. The system can be easily moved to a LAN, simply by respecting the requirements of the IP addresses. A main characteristic of the system is its distributivity. Unlike the centralized access control systems, the distributed ones allow multi-access and have higher reliability. The proposed system is an ACL type, the control list being the database of the server which contains all the data regarding the users and the privileges granted to each one. The technique is RBAC because the rights are established according to the roles of the users. The system was thought as being multilevel, where each user has certain rights. According to it, a user may access more or less rooms. A user is recognized by a CAP through the identity code found on the card. This code is sent to the server which determines the user rights.
B. The Hardware A CAP is made by a microcontroller based board connected to Internet. The board has the following features: - LPC 2148 microcontroller with ARM7TDMI core; - LCD display for messages sent to the user; - SD/MMC interface; - UART interface for in-system programming; - SPI interface; - buzzer; - Ethernet interface based on the ENC28J60 controller connected to the microcontroller through the SPI interface; it conforms to the IEEE 802.3 protocol; it is connected to the medium through a RJ-45 connector. The LPC 2148 microcontroller is a 32 bit, low consume, RISC machine. Most of the instructions are executed in a single machine cycle due to the existence of the internal 3 stage pipeline. The microcontroller has rich internal resources: RAM and flash memory, general purpose and fast general purpose I/O, vectored interrupt system, test/debug interface, emulation test module, timers with capture/compare facilities, A/D and D/A converters, PWM, real time clock, watchdog timer and several serial interfaces: USB 2.0 full speed, I2C, SPI and SPP, 2 UARTs. C.
The Software
The general diagram of the software is presented in fig. 1. The CAPs run a TCP/IP stack for embedded systems, [8], and the application through which the user interacts with the system runs on that stack. The application ensures the reading and the sending of the data from the card and the display of the messages received from the server. The server side application is written in Java and runs on a PC connected to Internet. The server side application establishes the connections with all the CAPs, accepts the data read by the CAPs and takes a decision according to the access policies. The decisions are sent to the CAPs in order to be displayed. The client side application From the stack’s point of view, the application is a function called when events appear on the network. The state diagram of the application is presented in fig. 2. The main state is NO_CARD. When the user inserts the card the application will read the data, will send the data to the server and will wait an answer. The state the application is in is DATA_SENT. When the answer arrives from the server, the application will verify it and if it is APP will pass in the APP_CARD. If the answer is DIS it will pass in the DIS_CARD state. In both situations the user is informed through a “Access Granted/Denied” message displayed on the LCD and is required to take its card. The user is announced by the buzzer too that he has to extract his card from the card reader. The application will stay in this state until the card is pulled out when it will enter the NO_CARD state.
- 342 -
SISY 2010 • 2010 IEEE 8th International Symposium on Intelligent Systems and Informatics • September 10-11, 2010, Subotica, Serbia
Figure 1. General diagram of the software
The next sample shows the implementation of the NO_CARD state: case(NO_CARD): write(0x00,” “); write(0x40,” “); if(mmc_init()!=0) { hs->state=NO_CARD; printf(“No card detected\n”); return; state=APP_CARD return; } uip_send(send,1); hs->state=DATA_SENT; return;state=NO_CARD; return;
