Sungsoo Kim

PUF-based Encryption Processor for the RFID Systems

The attacker can access the RFID systems illegally because authentication operations on the RFID systems are performed in wireless communication. Authentication methods based on the PUF were presented to defend attacks. Because these utilize Hash and AES, the cost is expensive for the low-cost RFID tag. In this paper, the PUF-based encryption processor and the low-cost RFID authentication protocol are proposed. The challenge-response pairs (PUF’s input and output) are encrypted by the PUF’s characteristics, not by Hash and AES. Besides, the encryption method is changed by the instruction code. Therefore, it is difficult for the attacker to analyze challenge-response pairs. The proposed scheme defends several kinds of attacks, which are physical, modeling, spoofing and location tracking attacks. We implemented the PUF-based encryption processor at low cost with small footprint and low power.

Improved 4-ary Query Tree Algorithm for Anti-Collision in RFID System

In RFID system, because a collision occurs while a reader attempts to identify multiple tags at the same time, the anti-collision algorithm is necessary. This paper uses the tree based algorithm expressed in the form of a B-ary tree, in which the time required to indentify all tags varies according to the value of B(B≥2). An increase in the value of B will decrease the number of collision cycles, while increasing the number of idle cycles. In order to solve this problem in previous algorithms, an Improved 4-ary Query Tree Algorithm (I4QTA) is suggested. If a collision occurs, the I4QTA uses the Bits Change Method (BCM) to remove idle cycles by only detecting the tag IDs that actually exist. In other words, it can quickly identify all tags by reducing the number of query-responses.

An Enhanced Slotted Binary Tree Algorithm with Intelligent Separation in RFID systems

In RFID systems, identifying the tag attached to the subject begins with the request from a reader. When the reader sends a request, multiple tags in the readers interrogation zone simultaneously respond to it, resulting in collision. The reader needs the anti collision algorithm which can quickly identify all the tags in the interrogation zone. We propose the Enhanced Slotted Binary Tree Algorithm with Intelligent Separation (IS-ESBT). In IS-ESBT, a collision can be prevented because the tags which match with the prefix of the readers request respond as followings; the group of tags with an even number of 1s in the bits to the prefix + 2nd bits responds in slot 0, while the group of tags with an odd number of 1s responds in slot 1. The IS-ESBT generates the request prefix so that the only existing tags according to the response in the corresponding slot. That is, it decreases the tag identification time by reducing the overall number of requests.

Regression of the Recovery Rate of ACM Rubber Gasket for Long-Term Performances

An engine head cover gasket in automobile engines is constantly exposed to high temperature and pressure conditions. The life time of an engine head cover gasket is important due to the increasing demand for more reliable automobile parts. The main failure mechanism of rubber gaskets, which is the aging due to temperature, can be determined using standard test equipment. The highly accelerated life test (HALT) is generally used for predicting the life time of rubber materials for a long period. Polyacrylate (ACM) with chlorine cure sites has been generally used for engine head gaskets. In order to improve the life time of such gaskets, ACM with carboxyl cure sites has been developed recently. In this study, the authors applied a mathematical model to obtain the recovery function of a rubber gasket, and utilized the FORTRAN program in order to obtain the optimized recovery function by using the successive zooming genetic algorithm (SZGA). Comparisons of the mean squared error (MSE) between the function data and the experimental data have been made. The life times of ACM with carboxyl cure sites rubber gaskets for different compression rates were obtained using the Arrhenius model. More accurate quantitative predictions considering the maximum compression rate would be realized.

An anti-collision algorithm without idle cycle using 4-ary tree in RFID system

This paper proposes an anti-collision algorithm needed in the process of recognizing data of the tags when one reader and multiple passive tags simultaneously have data communications in RFID system. QTA And 4-ary QTA which are based on the Tree algorithm, have the problem that the numbers of query - response are many and the idle cycle (requesting about non-existent tag) exists in the process of recognizing the tags. Advanced 4-ary Query Tree Algorithm(A4QTA) has a 4-ary tree structure and when tag responses according to query of the reader, it applies a digital coding method, the Manchester code, in order to detect the collision point. Also, this algorithm reduces the number of query-response compared to QTA and 4-ary QTA by devising a method for one bit collision and series two bit?s collision in these bits as well as improves the performance by removing the idle cycle.

A Symmetric Key Based RFID Authentication Protocol Using Encrypted Tag ID

The RFID technique is spot-lighted as the main technology in Ubiquitous era to identify the information of the objects. Because the communication on the RFID system is operated in the wireless, the secure and privacy problems always exist. In this paper, we propose mutual authentication protocol which is good for the low-cost tags based on ECC and AES. By using AES, the protocol could minimize tag resources. Even if the shared key might be opened to the world, the actual tag IDs encrypted by ECC are secure. We verified the security of the protocol using AVISPA.

A tag prediction anti-collision algorithm using extra bits for RFID tag identification

Tags have their own unique identifiers and information, and send such information to a reader; thereby the reader identifies the ID. The proposed algorithm predicts the tag ID by using the number of collided bits and the number of 1s in tag response signal. Even if tag collision occurs in the identification process, the reader identifies the RFID tag through prediction based on information provided by the extra bits. If the number of 1s to be included in collided bits is 1 or collided bits -1, the reader identifies tags as many as the number of collided bits.

An Efficient Tag Identification Algorithm using Bit Pattern Prediction Method

The procedure of the arbitration which is the tag collision is essential because the multiple tags response simultaneously in the same frequency to the request of the Reader. This procedure is known as Anti-collision and it is a key technology in the RFID system. In this paper, we propose the Bit Pattern Prediction Algorithm(BPPA) for the efficient identification of the multiple tags. The BPPA is based on the tree algorithm using the time slot and identify the tag quickly and efficiently using accurate bit pattern prediction method. Through mathematical performance analysis, We proved that the BPPA is an O(n) algorithm by analyzing the worst-case time complexity and the BPPA`s performance is improved compared to existing algorithms. Through MATLAB simulation experiments, we verified that the BPPA require the average 1.2 times query per one tag identification and the BPPA ensure stable performance regardless of the number of the tags.

A Short Distance Data Dissemination Protocol for Wireless Sensor Networks

We propose SDDD (short distance data dissemination) routing protocol. In SDDD, the transmission course uses the short distance path between the source node and sink node. SDDD can reduce total hop counts from the source to the sink as a result. That is, it extends the lifetime of the entire sensor network.