LDPC Decoding on GPU for Mobile Device

Algorithm 3

MSA kernel executing on the GPU grid.

(1) Initialize the work-group size (or number of work-item per work-group).

(2) Generating compact , from matrix H

(3) while

(4)   as the work-item on an work-group: do

(5)      for all : do

(6)        

(7)        update the message sent from to

(8)        update the message sent from to

(9)      Synchronize all threads

(10)      for offset = 0 to : do

(11)        

(12)        for all : do

(13)        update the posterior probabilities of

(14)      Synchronize all threads

(15)      perform hard decoding