**1 Overview**
As the demand for timely, accurate, and convenient communication continues to grow, it has become increasingly important to access information quickly and efficiently from anywhere. In 2002, the IEEE introduced the 10 Gb/s Ethernet standard, known as IEEE 802.3ae [1]. This marked a significant advancement in traditional Ethernet technology, increasing transmission speeds by ten times over Gigabit Ethernet to support mobile communication services.
The official 3G license was issued in China in January 2009, marking the start of the 3G era. To meet the growing needs of mobile communications, major telecom operators began large-scale 3G network construction. At the heart of this expansion is the base station, which is both expensive and crucial to service quality. As a result, operators are continuously looking for ways to reduce costs while improving performance. Distributed base stations, with their low cost, strong adaptability, and ease of deployment, represent the future direction of next-generation base stations.
In a distributed base station system, the traditional baseband unit (BBU) and remote radio unit (RRU) are separated and connected via fiber optics. The BBU is centralized and offers high capacity, allowing for flexible coverage and capacity management. The RRU, located at the site, is more adaptable to environmental conditions. Multiple RRUs can share BBU resources, reducing baseband investment. To enable efficient communication between base stations and repeaters, companies like Ericsson, Huawei, NEC, Nortel, and Siemens developed the CPRI interface protocol, which has since become a widely accepted standard.
**2 Brief Introduction to the CPRI Protocol [2]**
The CPRI (Common Public Radio Interface) protocol consists of two layers: the physical layer (L1) and the data link layer (L2). It serves as an internal interface between the Radio Equipment Control (REC) and Radio Equipment (RE), as well as between two REs. The CPRI interface supports three types of data streams: user platform data stream (SAPIQ), control and management platform data stream (SAPCM), and synchronization platform data stream (SAPS).
**3 FPGA-Based CPRI Protocol Transmission Scheme**
**3.1 Basic Design Options**
Implementing the CPRI protocol using FPGA can be done through two main approaches.
**(1) Option 1**
This method uses the RockeTIO transceiver from Xilinx, integrated into Virtex2Pro and later chips. It provides dedicated serial communication without consuming other FPGA resources. In the Virtex5 series, RockeTIO is referred to as GTP. The advantage of this approach is its compact design, which makes PCB layout easier and improves system immunity to interference. Each Virtex5 chip includes multiple GTP transceivers, enabling 10 Gb/s transmission using four GTP cores. Each GTP core includes a receive and transmit link [4–5].
**(2) Option 2**
This approach involves using specialized chips such as the TI TLK2501 or National Semiconductor’s SCAN25100. Among these, the SCAN25100 offers the most complete features, including 8b/10b encoding, high-speed serialization, clock management, and signal conditioning. However, achieving 10 Gb/s with such chips requires multiple units, increasing complexity and board space.
**3.2 Implementation Details**
The SCAN25100 supports data rates of 2.4576 Gb/s, 1.2288 Gb/s, and 0.6144 Gb/s, while the TLK2501 operates between 1.5 Gb/s and 2.5 Gb/s. To reach 10 Gb/s, four of these chips would be needed, which increases wiring complexity and board size.
This paper adopts the first design approach, using the GTP transceiver in a Virtex6 FPGA to implement 10 Gb/s CPRI data transmission. As shown in Figure 1, the Ethernet optical interface includes four components: a 10GE optical module, a PHY transceiver, a clock module, and the FPGA.
**10GE Optical Interface**: This module includes optoelectronic components and circuits, responsible for converting electrical signals to optical signals and vice versa. It uses an XFP optical module with a 10 Gb/s serial interface, offering small size, low power consumption, and easy multi-port integration.
**PHY Transceiver**: This physical layer chip handles data transmission between the FPGA and the optical module. It processes 4 channels of 3.125 Gb/s data and combines them into a 12.5 Gb/s stream, which is then split back into 4 channels for the FPGA. With 8b/10b encoding, the effective rate reaches 10 Gb/s, meeting the design requirements.
**Clock Module**: A 61.44 MHz crystal oscillator provides the reference clock. Since each data link runs at 3.125 Gb/s, the GTP core requires high-precision timing. Instead of using DCM, an external differential oscillator is used to minimize jitter. The GTP generates internal clocks for RXUSRCLK and TXUSRCLK, ensuring reliable data transmission.
**FPGA Part**: The FPGA manages the 8b/10b encoding, high-speed serialization, framing, de-framing, and synchronization of the CPRI protocol. It also controls the optical and PHY modules, monitoring their status and configuration.
**4 Design Verification**
Usb Common Mode Choke,Magnetic Ring Inductor,Wurth Inductor,Coilcraft Inductor
IHUA INDUSTRIES CO.,LTD. , https://www.ihuagroup.com