Most of the currently developed passive radar systems determine target position by computing cross-correlation between signals arriving from different paths. Achieving decent range and resolution using this technique is a major challenge due to the high processing power required. This obstacle is often overcome by simply applying brute-force solutions, such as using high-performance multicore processors. However, this dramatically increases price and power consumption while making the radar less portable.
In this thesis, an alternative method will be analysed. Instead of using pure software, the solution proposed includes the application of a hardware-accelerated device. The main concept is to create hardware that is dedicated to the given task while using software to aid the custom logic.
For general-purpose processors, serial operation and overhead arising from instruction fetch and other administrative tasks create a potential bottleneck. Pro-grammable logic can solve this issue, but it is relatively difficult with these devices to serially perform a set of different tasks, since a separate piece of hardware has to be synthesized for each of those. By using components that provide support for software and hardware solutions alike, the benefits of both methods can be exploited.
The basic building block of the design presented here is the Zynq 7020 All Programmable SoC, which is capable of performing computationally demanding procedures while ensuring low cost and power consumption. To demonstrate this, the design steps of the basics of a system running the required signal processing algorithm will be presented. Correct operation will be proven using simulation, while calculations will show that the appropriate speed on the hardware can be achieved.