National Conference on Advances in Computing, Communication and Networking |
Foundation of Computer Science USA |
ACCNET2016 - Number 5 |
June 2016 |
Authors: Pooja Painully, Kavita Joshi, Munmun Ghosal |
e28f9e46-d40a-428a-ba1e-027273fa2e24 |
Pooja Painully, Kavita Joshi, Munmun Ghosal . Energy Saving using Data Encoding Scheme In Network-on-Chip for LDPC Application. National Conference on Advances in Computing, Communication and Networking. ACCNET2016, 5 (June 2016), 9-13.
Communication subsystems such as routers and wired or wireless network interfaces consume more power in network on chip. With the target of reduction in power consumption in network on chip, three encoding schemes are presented in this paper. These schemes work at flit level which results in reduction of self and coupling transition without making modification in the link architecture and routers. Globally to defend memories from soft error correction codes have been used which make changes in logical value of memory cubicles without destructing the circuit. As memory expands, demand of complex decoder rises due to complex codes. One of such decoder is LDPC. We can apply these encoding schemes to linear density parity check (LDPC) codes. Power dissipation is reduced. The dynamic power value reduction is verified by using x-power analyzer tool.