توصيفگر ها :
رادار اتومبيل , مكان‌يابي و كاهش تداخل , رادار FMCW , مديريت تداخل , بازسازي سيگنال , تكميل ماتريس
چكيده فارسي :
مردم در سراسر جهان، روزانه وقت زيادي صرف جابه‌جايي با اتومبيل از يك مكان به مكان ديگر مي‌كنند. در سال‌هاي گذشته به ويژه در دهه اخير مطالعات فراواني در زمينه اتومبيل‌هاي خودران انجام شده است. دانشمندان و شركت‌هاي فناوري، تلاش زيادي براي توسعه تكنولوژي خودروهاي كاملا خودران انجام داده‌اند تا در آيندهاي نزديك روياي اتومبيل خودران به واقعيت بپيوندد.
اين خودروها، از تركيب فناوري‌هاي سنسورهايي همچون رادار، ليدار و دوربين براي درك محيط اطرف خودرو در سيستم كمك راننده پيشرفته و يا ADAS استفاده مي‌كنند. رادارها با توجه به عملكرد مناسب در انواع شرايط محيطي از اهميت بسزايي در سيستم كمك راننده پيشرفته برخوردارند. بنابراين مطالعه اين زمينه كمك شاياني براي حضور هرچه سريع‌تر اين وسيله نقليه هوشمند در دنياي واقعي مي‌كند.
با توجه به افزايش تقاضا براي اتومبيل‌هاي خودران در آيندهاي نزديك و همچنين افزايش تعداد رادارهاي موجود در يك اتومبيل خودران، احتمال رخداد تداخل در سيستم رادار اتومبيل افزايش مي‌يابد. بنابراين يكي از مهم‌ترين مسائلي كه در حوزه رادارهاي اتومبيل‌هاي خودران بايد بررسي شود، موضوع كاهش تداخل و تاثيرات ناشي از آن مي‌باشد كه مي‌توان به تداخل با سيگنال ديگر رادارهاي خود اتومبيل، تداخل با سيگنال ديگر اتومبيل‌ها و تداخل با امواج سيستم‌هاي مخابراتي اشاره كرد.
در اين پايان‌نامه برآنيم به بررسي تداخل سيگنال رادار اتومبيل با رادارهاي ديگر اتومبيل‌ها بپردازيم. در قدم اول با بررسي روش‌هايي سعي بر مكان‌يابي تداخل مي‌شود، سپس به حذف تداخل در مكان‌هايي كه تداخل آشكارسازي شده است، مي‌پردازيم. در نهايت با به كارگيري رويكرد تكميل ماتريس تلاش مي‌كنيم سيگنال آسيب‌ديده در اثر تداخل را ترميم كنيم. طرح پيشنهادي سبب مي‌شود تداخل مديريت شده و عملكرد سيستم رادار اتومبيل در برابر تداخل بهبود بخشيده شود.
چكيده انگليسي :
Nowadays, people spend so much time traveling by car, so providing comfort by implementing Advanced Driver Assistance Systems (ADAS) in vehicles is a crucial topic to study. There are a lot of advanced technologies that have been implemented in the ADAS, such as cameras, radars, and lidars. Also, ADAS sensors, especially radars, must perform safely and reliably. The importance of radars in the ADAS is because of their ability to work under different weather conditions and during the night. However, some industrial bottlenecks in automotive radar should be solved to provide vehicle ADAS reliability.
In the Advanced Driver Assistance Systems of the car, there is a bottleneck regarding the automotive radar, which can cause the car to be blind in certain situations or even detect ghost targets in some situations. In fact, the radar system is one of the most important sensors in an autonomous car. This is because of its ability to work in different weather conditions and in places with no light. Considering there are multiple radars in a car that provide the information that is needed, it is crucial that there is a radar system that provides a reliable and accurate signal to ensure that the information we require is delivered to us. However, due to the outstanding abilities of radars, not only will the number of radars in a car increase because of their outstanding capabilities, but also the number of cars with ADAS will increase due to their high demand in the near future. This means that the radars’ signals may interfere with each other, so the performance abilities of the ego radars which receive the signals of interferer radars will decrease considerably.
A study on radar-to-radar interference is essential since it affects the performance of the Advanced Driver Assistance System (ADAS) during the detection of cars. Furthermore, interference management techniques are mainly divided into three categories: interference avoidance, interference detection, and repairing interfered signals. Although interference avoidance techniques can decrease the probability of interference accordance, it is vital to implement some methods to detect the interference region and repair the interfered samples because of the numerous automotive radars in the future. As soon as the interference region has been detected, it is time to repair the samples that have been missed because of the interference.
Concerning signal reconstruction methods, it is important to increase the signal-to-noise-and-interference ratio (SINR) since when a waveform experiences interference, the SINR drops significantly. A novel method is a matrix completion approach, which can estimate the missing samples. There are two matrix completion methods, the Naive Approach and Low-Rank Matrix Fitting (LMaFit), that we have implemented to reconstruct the interfered samples. A Naive Approach uses Singular Value Decomposition to solve the problem, while LMaFit uses an optimization problem to solve the problem. Also, these methods use the information from unaffected samples to reconstruct the interfered samples and increase the SINR considerably, so it provides fertile ground for the car to extract the necessary information. As a result, one of the important technological bottlenecks in the vision system of a car with ADAS can be solved, and people use these cars safely.
