100 Gb/s Data Link Layer - from a Simulation to FPGA Implementation

In this paper, a simulation and hardware implementation of a data link layer for 100 Gb/s terahertz wireless communications is presented. In this solution the overhead of protocols and coding should be reduced to a minimum. This is especially important for high-speed networks, where a small degradation of efficiency will lower the user data throughput by several gigabytes per second. The following aspects are explained: an acknowledge frame compression, the optimal frame segmentation and aggregation, Reed-Solomon forward error correction, an algorithm to control the transmitted data redundancy (link adaptation), and FPGA implementation of a demonstrator. The most important conclusion is that changing the segment size influences the uncoded transmissions mostly, and the FPGA memory footprint can be significantly reduced when the hybrid automatic repeat request type II is replaced by the type I with a link adaptation. Additionally, an algorithm for controlling the Reed-Solomon redundancy is presented. Hardware implementation is demonstrated, and the device achieves net data rate of 97 Gb/s.