Investigation On Coding Techniques With Spinal Codes

Spinal codes are widely concerned by the researchers because of its excellent performance.This paper deeply researches the coding algorithm of Spinal codes,analyzes the factors which influence the performance of Spinal codes and presents improved coding algorithms.Results can be gotten as follows:First,factors affecting the performance of Spinal codes are analyzed including the intra block length,hash function and the sequence of encoding symbols.The simulation results show that appropriate encoding symbol sequence and intra block length are helpful to improve the encoding performance.Rational selection of hash function can reduce the complexity of encoding algorithm.Then,in order to improve the performance of truncated tree decoding algorithm,this paper proposes two kinds of improved decoding algorithms.The decoding algorithms use the channel state and decoded node information to dynamically adjust the algorithm parameters.The simulation results show that the proposed decoding algorithms can improve the decoding performance compared to the truncated tree decoding algorithm.Furthermore,a concatenated code which employs Spinal codes as inner codes and polar codes as outer codes is proposed.Polar codes are used to correct the errors in the decoding process of Spinal codes.Polar codes overcome the defects of insufficient reserved nodes of Spinal codes.Simulation results show that the performance of concatenated codes is improved compared to original Spinal codes.