A Middleware Approach to Achieving Fault Tolerance of Kahn Process Networks on Networks on Chips

Algorithm 2

Receiver middleware task per incoming connector.

for 𝑖=1 to connector_bound do

𝑀𝑃𝐼_𝑠𝑒𝑛𝑑(𝑡_𝑣𝑐,𝑃_𝑅,𝑣𝑐)

𝑖←𝑖+1

end for

Loop

𝑀𝑃𝐼_𝑟𝑒𝑐𝑣(𝑡,𝑃_𝑅,𝑜𝑐)

 𝑀𝑃𝐼_𝑠𝑒𝑛𝑑(𝑡_𝑣𝑐,𝑃_𝑅,𝑣𝑐)

 output_port.write(t)

end loop