Ultima Computing : Jurnal Sistem Komputer, Vol. 14, No. 2 | December 2022 48 ISSN 2355-3286 Black Hole Detection Using Modified Sequence Number in Vehicular Ad-hoc Network Rakha Fikran Julda 1 , Dany Primanita Kartikasari 2 , Rakhmadhany Primananda 3 1,2,3 Informatics Engineering, Brawijaya University, Malang, Indonesia 1 rakhafikran@gmail.com, 2 dany.jalin@ub.ac.id, 3 rakhmadhany@ub.ac.id Accepted 08 September 2022 Approved 20 October 2022 Abstract— Vehicular Ad-hoc Network (VANET) is a type of wireless network with Dedicated Short-Range Communication (DSRC) that enables communication between vehicles (V2V) and communication between vehicles to infrastructure around them (V2I). VANET has several security requirements to consider in order to maintain the network functionality. Availability is the most important security requirement due to its responsibility of maintaining the functionality of the network, attack on availability may cause the lack of availability and reduce the efficiency of VANET. One of the attack that threat the availability of VANET is black hole. In this paper, we address the problem of black hole attack in VANET, using Modified Sequence Number (MSN) as a detection method. The simulation is performed using NS-2 as a simulator and AODV as a routing protocol. Detection Rate (DR) and False Alarm Rate (FAR) are used to evaluate the performance of MSN algorithm in detecting black hole attack. Evaluation with variation in the number of CBR packets shows that MSN algorithm successfully detects black hole attacks with DR values reaching 69.0909% at 10 CBR packets and FAR values reaching 0.0037 at 20 CBR packets. We also evaluate the performance of MSN algorithm with variations of node density. The evaluation shows that MSN algorithm successfully detects black hole attack with DR values reaching 100% with a density of 10 and 20 nodes, with the percentage of FAR values reaching 0% in all numbers of node density. Index Terms— AODV; Black Hole Attacks; VANET I. INTRODUCTION VANET is a dedicated short-range communication (DSRC) wireless network technology that works on the 5.9 GHz frequency spectrum for communication between vehicles. Unlike MANET which uses IEEE 802.11a/g/n, the physical layer of VANET uses the IEEE 802.11p standard protocol with its focus on communication in Intelligent Transportation Systems (ITS) environment [1]. In VANET, there are two types of communication that take place. Vehicle-to-Vehicle (V2V) is a communication between vehicles, whereas Vehicle-to-Infrastructure (V2I) communication is between vehicles and infrastructure such as Roadside Units. (RSU) [2]. VANET has several parameter to ensure its network security, including integrity, authentication, availability, privacy, and non- repudiation. Attackers exploit these security parameter to carry out their attack. The primary role of availability to guarantee the network services continue to work in the event of malicious attacks is what makes availability one of the most important parameter in VANET security, as its lack of availability may diminish the efficiency of VANET [3]. One of the attack that threat the availability of VANET is black hole [4]. Black hole use the routing protocol’s route discovery process which searches for the nearest route, by pretending to be the node with the newest and shortest path to destination. The black hole node that receives the packet drops the packet so that the packet is not received by the destination. The routing table of the routing protocol is disrupted by the behavior of black hole attacks, reducing the performance efficiency of VANET [5]. Therefore, a routing protocol that can works on network with dynamic topologies, high mobility, and able to protect against attacks such as black hole is required. In attempt to secure VANET from black hole attack, several research has been done. Among these is using a protocol called Secure AODV (SAODV). The proposed routing protocol changes the destination IP address of AODV packet request by applying Cyclic Redundancy Check 32 bits (CRC-32) as a hash function, resulting a secure AODV RREQ packet without any extra overhead. The test results in this study explain that the proposed protocol successfully addresses black hole attack with PDR performance, end-to-end delay, routing overhead, and throughput that is nearly identical to traditional AODV. Additionally, the proposed protocol has a high detection rate at both high and low node density [6]. However, the research has not shown the performance of SAODV in dealing with black hole attack with variation in number of packets sent. The variation in number of packets sent can determine how well a method to detect black hole attack [7]. Therefore, we use a scenario with different number of Constant Bit Rate (CBR) packet sent to measure the performance of our propose method.