Luis Duarte https://luisduarte.eu Mon, 20 Jan 2025 11:50:31 +0000 en-US hourly 1 https://wordpress.org/?v=7.0.2 https://luisduarte.eu/wp-content/uploads/2023/09/cropped-Untitled-design-2-32x32.png Luis Duarte https://luisduarte.eu 32 32 PhD FCT grant https://luisduarte.eu/phd-work/ Fri, 15 Sep 2023 08:00:56 +0000 https://luisduarte.eu/t-shirt-design-is-the-part-of-design-1-copy/ Studentship for Doctoral Research was awarded by "Fundação para a Ciência e Tecnoologia" (FCT) in Portugal.

]]>

This PhD work’s main objective is to accomplish a dual-polarisation STDCC radar with fully reconfigurable capabilities. It explores the auto-correlation properties of PN sequences to detect targets taking advantage of their high immunity to interference characteristics.

This novel radar uses the Swept Time-Delay Cross-Correlator (STDCC) technique that presents high-resolution and multi-user operation with its good interference immunity.

STDCC reconfigurable baseband:

A key part of this thesis was the development of a radar baseband with fast prototyping capability to improve radar detection performance on the fly. Thus, an agile radar baseband is required to generate radar waveforms and tune them, allowing to study of Key Performance Indicators (KPI) for different radar waveforms. The proposed reconfigurable baseband architecture, depicted below, uses an FPGA to change, on-the-fly, the all-digitally outputted PN sequences and their bandwidth.

Block diagram of reconfigurable STDCC radar baseband.
mmWave reconfigurable RF front-end operating between 24 and 28GHz:

The 24GHz band has a license-free band of 250MHz bandwidth that was widely used in the last decade in radio location applications, especially for short-range radar (SRR) scenarios in the automotive and drone sectors. We developed an RF stage using X-microwave technology that provides a complete modular building block eco-system for microwave components as shown in the following figure. Two transmitting and receiving RF stages operating between 24 and 28GHz were achieved in the end.

mmWave reconfigurable RF front-end
Radar App development:
PDP plot of two close target detection
PPI plot of tree and wall detection

 

 

 

 

 

 

A MatLab app was fully developed to automate the FPGA configuration and to visualise the radar captured data. Three different data visualisation tabs, each responsible for a different capture mode with the antennae still or moving. The user can select between a PDP that is a range profile mode with received signal strength as a function of distance, a waterfall mode, which shows the radar signal’s power spectrum as a function of time and PPI that uses motors to rotate the radar to accomplish a comprehensive view of the radar surrounding area.

Waterfall tracking of moving target at 3 meters (VV – left, HH -right).
SAR algorithm development:

Typical application usage for this technique is high-resolution screening and mapping of difficult access areas like heavy vegetation forests or other critical systems used in aviation systems. Using SAR algorithm we can reconstruct the intended target by performing various acquisitions with different perspectives accomplished by the radar movement as shown below.

The SAR algorithm can also improve radar spatial resolution. The STDCC radar has a maximum of 500 MHz, which corresponds to a 30 cm resolution. By using the synthetic aperture technique we can accurately reconstruct smaller targets below our resolution like is shown in the results below from a PCB that has the physical dimensions of 12x8cm with three shapes of 3 cm.

SAR STDCC radar on a PCB target with shapes (12 x 8 cm)

]]> PURE5GNET https://luisduarte.eu/pure5gnet/ Mon, 10 Dec 2018 08:30:09 +0000 https://luisduarte.eu/t-shirt-design-is-the-part-of-design-copy/

The demonstration scenario, illustrated in Figure 1, is composed of a Primary User (PU) transmitting to a Primary Base Station (BS) and two Secondary Users (SU1 and SU2) transmitting to a Secondary Base Station. Since all users will utilize the same frequency and be transmitted at the same time, the two Secondary Users will be perceived as interference by the Primary BS. Therefore, a precoding algorithm is required to code all users’ data before transmission to enable the proper data decoding at the Base Station. The precoding and decoding algorithms were implemented in Xilinx System Generator in the HETCOP project.

PURE5GNET Hardware mapping

The proposed scenario has three users that need, in total, 5 transmitter (Tx) and 3 receiver (Rx) antennas. Therefore, 1 FMCOMMS5 and 1 FMCOMMS3 RF boards were required to accomplish all the RF chains required by the three users on the User Equipment side (UE). 

PURE5GNET full system (UE – left side ; BS – right side)

The hardware project, responsible for configuring the RF boards, was designed from scratch with the Xilinx KC705 development kit in mind. The KC705 only has two FMC connectors, thus another FPGA was required to have the FMCOMMS3 and FMCOMMS5 RF boards. Thus, a high-speed connection between both FPGAs was required to ensure the synchronisation between all three users’ transmissions. Therefore, both boards were connected via fibre optics (SFP+) using a Aurora encoder that was integrated into the OFDM engine and MU-MIMO precoding developed in Xilinx System Generator.

User Equipment side – hardware
User Equipment side – diagram

 

 

 

 

 

 

 

The Base Station side onlyrequires one transmitter and two receiver antennas. To implement it a KC705 FPGA development kits and one FMCOMMS3 RF card was used.

Base Station side – diagram
Base Station side – hardware

 

 

 

 

 

 

Live demonstrator:
]]> HETCOP https://luisduarte.eu/hetcop_proj/ Fri, 01 Jan 2016 17:51:37 +0000 https://luisduarte.eu/t-shirt-design-is-the-part-of-design-2-copy/

The demonstration scenario, illustrated in Figure 1, is composed of a Primary User (PU) transmitting to a Primary Base Station (BS) and two Secondary Users (SU1 and SU2) transmitting to a Secondary Base Station. Since all users will utilize the same frequency and be transmitted at the same time, the two Secondary Users (SU1 and SU2) will be perceived as interference by the Primary BS.

Demonstration scenario

The precoding and decoding algorithms were implemented in Xilinx System Generator environment for later compilation and deployment to the FPGA kits, as cn be seen in the next figure. 

System Generator top view

 

The Tx block is comprised by a Precoding and Framing part, OFDM engine and UE demultiplexer. The first part pre-codes all users data, introduces pilots for channel estimation and serializes the information. Such serializer is a resource saving technique that allows the usage of a single OFDM Tx Engine. Then, the UE demultiplexer receives the modulated information and feeds it to the correct transmitting antennas for Over-The-Air transmission. 

User equipment transmitter

The algorithm running on the Base Station side is responsible for channel estimation recurring to the sent pilots, data demodulation with the OFDM Rx Engine, as well as, estimating the PU data with the decoding algorithm.

Base Station receiver

The MU-MIMO OFDM had such a frame structure. The sent payload is represented by the white tiles that are common in all users, each being an interference to each other. In the end the decoding will be responsible of recovering the original data.

MU-MIMO OFDM structure
]]>