آنالیز تهدید شبکه IoT توسط سیستم تشخیص نفوذ شبکه عصبی مصنوعی Threat analysis of IoT networks Using Artificial Neural Network Intrusion Detection System

I. INTRODUCTION The internet of things (IoT) is a network of distributed (sensor) nodes, (cloud) servers, and software. This paradigm permits measurands to be sensed and processed at in real-time creating a direct interaction platform between cyber-physical systems. Such an approach leads to improved efficiency in the generation and usage of data leading to economic benefits [1]. Research conducted by Cisco reports there are currently 10 billion devices connected, compared to the world population of over 7 billion and it is believed it will increase by 4% by the year 2020 [2]. Threats to the IoT paradigm are on the rise; however patterns within recorded data can be analyzed to help predict threats [3]. Intruder types are categorized into two: 1) External Intruders – these are people who fall outside the network and hence do not have permissions on the network. They operate by sending malware, or by using exploits to gain access to systems [4]. 2) Internal Intruders – these people have rights and privileges to access the network, but misuse them malevolently. These types of attack include changing important data content or theft of confidential data. All these threats can be done physically by hacking into the computer system or by accessing a network remotely without permission [4]. IoT threat can be classified into four types [5]: 1) Denial of Service (DoS) – This threat denies or prevents user’s resource on a network by introducing useless or unwanted traffic 2) Malware – Attackers use executable code to disrupt devices on the IoT network. They may gather sensitive information, or gain unauthorized access to the devices. The attacker can take advantage of flaws in the firmware running on the devices and run their software to disrupt the IoT architecture. 3) Data breaches – This is a security incident where sensitive, protected or confidential data is retrieved from the network. Attackers can spoof ARP packets to listen on the communication between peers on the network. 4) Weakening Perimeters – IoT network devices are currently not designed considering the pervasive security. Network security mechanisms are not often present in the devices making the network a vulnerable one for threats [5][6].