Low power consumption and crosstalk resistant coding and decoding method and coding and decoding device for network on chip

Abstract

The invention provides a crosstalk resistant and low power consumption coding and decoding method and a codec thereof, which can effectively reduce the dynamic power consumption of data transmission and improve the quality of the data transmission. The coding and decoding method of the invention changes the encoding mode of data by considering the reduction of the number of turns between data; compared with original data, the number of turns can be reduced by 37.56%, so that the dynamic power consumption in the data transmission is effectively reduced; and by inserting a mask code in the worst crosstalk condition, the crosstalk between data can be effectively reduced on the premise of minimizing the additional power consumption. For a random data source, the crosstalk resistant and low power consumption coding and decoding method and the codec thereof provided by the invention reduce the transmission power consumption by 37.51% while avoiding the worst crosstalk.

Description

technical field [0001] The invention belongs to the technical field of microelectronics, relates to a network-on-chip (NoC) of a deep submicron process integrated circuit, in particular to a low-power codec method and a codec for suppressing crosstalk, which can be used for ultra-large-scale integration circuit design. Background technique [0002] Network-on-Chip (NoC) is a brand-new communication method of System-on-Chip (SoC), which is the main component of multi-core technology. Using routing and packet switching technology to replace the traditional on-chip bus to complete the communication task is a new design paradigm for VLSI in the future. With the increase of the number of integrated cores in a single chip, the long interconnection with heavy load becomes the main bottleneck of the power consumption of the network on chip. Low-power coding techniques are one of the most effective ways to reduce the dynamic power consumption of an on-chip network. [0003] Howeve...

AI Technical Summary

Problems solved by technology

At present, many literatures introduce weight-based bus-invert coding (Weight-Based Bus-InvertCoding), "green" self-correcting error coding (Self-Corrected Green coding) and other coding to achieve low-power coding, but do not consider long interconnections The serious crosstalk problem caused by the capacitance between lines, only using low-power encoding technology, it is very difficult to achieve high-speed and low-power data transmission for long-interconnected on-chip networks

Images