EdgeQ is sampling what it calls a “base station on a chip” that handles high-PHY through baseband. CEO Vinay Ravuri explains its architecture in a video.
Wireless telecom networks need small cells to bring mmWave signals to fruition. With small cells, consumers can get the full high-speed benefits of 5G. Because small cells are, well, small, any integration of functions reduces size and weight. To meet that need, EdgeQ has developed what it calls a “base station on a chip.” The company recently announced that it’s sampling the chip to customers. To learn about the technology behind the devices, we spoke with EdgeQ CEO Vinay Ravuri in the video above.
The EdgeQ device is based upon an ARM 8-core Neoverse processor running Linux to handle overall housekeeping. The baseband signal processing relies on a RISC-V processor that uses EDGEQ’s PHY software to produce what’s called the “high PHY” shown in the middle layer of Figure 1. The EdgeQ chip doesn’t handle RF functions, but it does perform the signal processing needed to get the demodulated signals into baseband digital form. The Layer 2 and Layer 3 functions in the chip use third-party software from Radisys to organize bits into packets for transport to and from the network. See the Figure 2 for the three major functions of the chip.
In the video, Ravuri explains the chip architecture in detail. It supports Open RAN using the 7.2 functional split, which places the high-PHY functions in an Open RAN network’s distributed unit (DU). Thus, the chip can also function as a DU for small cells, much in the way a baseband unit (BBU) functions with macro cells.
The chip has two interfaces to the radio: analog and digital. The analog interface uses a DAC and ADC to take signals and, through software-defined radio processing, bring them to and from baseband. The JESD204 digital interface lets it connect to 5G RF transceivers such as those from Analog Devices and Texas Instruments.
See Open RAN functional splits, explained and Functional Splits: the foundation of an Open 5G RAN for explanations of the Open RAN functional split.
Tell Us What You Think!