Design and implementation of a high speed interface system over Gigabit Ethernet based on FPGA for use on radar acquisition systems

• Published in: 2017 Electronic Congress (E-CON UNI)
• Main Authors: Rojas, John, Verástegui, Joaquín, Milla, Marco
• Format: Conference Object
• Language: English
• Published: Institute of Electrical and Electronics Engineers 2018
• Subjects: FPGA; LVDS; Double buffering; Gigabit Ethernet; UDP; Threads of execution; http://purl.org/pe-repo/ocde/ford#1.05.01; Greenland
• Online Access: https://hdl.handle.net/20.500.12816/3065; https://doi.org/10.1109/ECON.2017.8247311

The Jicamarca Radio Observatory (JRO) is part of the Western Hemisphere chain of Incoherent Scatter Radar (ISR) observatories which extends from Lima, Peru to S0ndre Str0mfjord, Greenland. The equatorial ionosphere is studied only by JRO in the world. The Observatory is a facility of the Instituto Geofísico del PerU operated with support from the US National Science Foundation Cooperative Agreements through Cornell University. One of the main radar components is JARS (Jicamarca Acquisition Radar System), which functionality is based on CPLDs (Complex Programmable Logic Device) to configure the system and transfer the data from the digital receivers to the acquisition computer over a proprietary interface NI PCIe-6537. However due to some limitations as its high cost, driver updates dependency and the data transfer speed, it was necessary to replace this interface with the design and implementation of a high-speed hardware, embedded on FPGA devices, to transmit the data through the LVDS interface to a double buffering stage, and forward these as packets on the standard Gigabit Ethernet, based on the IEEE 802.3 protocol and using the UDP protocol. In this way, this development will allow to update the hardware of the current JARS to get a low cost portable system and to work with the required bandwidth. A prototype of this system was developed on the JRO and also a customized software was written, based on UDP socket and multiple threads of execution.