Development of AI-integrated Randomly Hopped LFM Photonic RADAR for Mitigating Interferences and Attackers in Autonomous Vehicles

  • Ukaegbu, Ikechi (PI)
  • Mustafa, Manat (Co-PI)
  • Pavlenko, Vladimir V. (Co-PI)
  • Nakarmi, Bikash (Co-PI)
  • Muthaiyah, Saravanan (Other participant)
  • Pan, Shilong (Other participant)
  • Wang, Xiangchuan (Other participant)

Project: Monitored by Research Administration

Project Details

Grant Program

Collaborative Research Program for 2022-2024

Project Description

The use of sensors is increasing in every aspect of our daily necessities such as communications, transportation, agriculture, engineering, data science, education, gaming and others. The applications of sensors have created massive connectivity and various sensors, enough computing powers, big data and artificial intelligence [1-3]. The evolution of integrated circuits followed Moors law (the number of transistors in integrated circuits doubles every two years) for many decades. The computing power also evolved accordingly. According to Butters’ law [3], the amount of data coming out of an optical fiber is doubling every nine months. Data traffic has been increased dramatically by connected devices (personnel computers, smartphones, smart pads, sensors) to the internet. So, we are in the era of the internet of things. For sensors that convert physical quantities to digital data, we have cameras, Radio detection and ranging (RADARs), Light detection and ranging (LiDARs), environmental sensors, velocity and acceleration sensors, gyroscopes and many others. The recent forecast shows that number of connected devices to mobile networks will be 75 billion in 2025 [3]. Hence, massive data sets generated with the increased connected devices. Researchers focus on managing these data and getting benefits from it for the general public [3]. The increasing research trends and maturity on computing power, connectivity and sensors and AI have become enablers for many new applications, including autonomous vehicles.
The Autonomous vehicle is one of the prominent technologies that will shape the future of transportation, whether public or private because 93% of car accidents are caused by human errors, with the deaths of 1.3 million people every year and costs of about USD 518 billion. Not only it loss of people’s life and economic loss, but it also causes a loss of almost 5.5 billion hours due to traffic jams. The Autonomous vehicle may be one of the solutions that solve these issues and be of great help to senior citizens and disabled people. To have a real autonomous car, we require level 4 and level 5 autonomous cars, which is possible only recognize all the parameters that influence decisions correctly, timely and effectively. For this, we have to mimic the natural eyes and make a quick and fast decision. In order to mimic the natural eye and brain of humans on decisionmaking in the autonomous vehicle, several sensors such as cameras, LiDARs and RADARs have been the essential components and have been implemented separately or together in autonomous vehicles. These sensors have their advantages and disadvantages, as illustrated in Table 1. No matter which sensors are used, with an increase in the number of sensors used, data processing and advancement in AI technology, threats on safety and security become alarming issues due to interferences and the intruder. Recently serious accidents occurred in the USA and in CHINA (Tesla Model 3) which may be due to the false recognition of white trailer under bright sunlight and sensor and wrong decision of control system, respectively.
It is clear from the above-mentioned scenarios that the safe and secure movement of vehicles without intervention is the critical issue to be addressed for the successful implementation of level 4 and level 5 autonomous vehicles. To achieve this, several sensors are needed for the real mimicking of the eyes, ears, nose and brain of the human being and should work more efficiently. There are several works where Cameras, LiDARSs and RADARs have been implemented either as separate sensor units or as a combination of different sensor types. With the increase in the use of different sensors, interference between LiDARs [6] and RADARs [7,8] becomes a critical issue for safety. Hence interferences and attacks in autonomous vehicles should be addressed to make the autonomous vehicle in reality. However, there has been less focus on the interferences from the sensors or intended attacks that are detrimental to the safety and security of autonomous vehicles. Hence, the primary goal of this project is to make autonomous vehicles environment safe and secure by research in physical layers via photonic technologies, especially implementing Photonics RADARs (Ph-RADARs) for mitigating interference and attackers. In this project, we choose RADAR as a primary sensor based on Table 1 as it outperforms others in velocity measurement, proximity detection, good range resolution, and works in all weather [9]. The primary goal of this project is achieved by carrying out the following tasks:
Develop Ph-RADAR with LFM waveforms, which is the most widely used in autonomous vehicles.
● Investigate interferences among the sensors, LFM Ph-RADARs, among which one works as a Victim RADAR and
another as Intruder RADAR.
● With an in-depth understanding of interferences, this project aims to develop a random hopped LFM Ph-RADAR to
mitigate the interference and attacks in an actual real autonomous vehicle scenario.
● Implement AI technology to improve the range resolution and RADAR image-resolution.
StatusActive
Effective start/end date1/1/2212/31/24

Keywords

  • Photonic RADAR
  • Microwave Photonics
  • Artificial Intelligence
  • Optoelectronics

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.